IBM Power 770 and 780 Technical Overview and Introduction

ibm.com/redbooks Redpaper
Front cover
IBMPower770and780
Technical Overview
and Introduction
Alexandre Bicas Caldeira
Carlo Costantini
Steve Harnett
Volker Haug
Craig Watson
Fabien Willmann
Features the 9117-MMC and 9179-MHC based on
the latest POWER7 processor technology
Describes MaxCore and TurboCore for
redefining performance
Discusses Active Memory
Mirroring for Hypervisor

International Technical Support Organization
IBM Power 770 and 780 Technical Overview and
Introduction
December 2011
REDP-4798-00

© Copyright International Business Machines Corporation 2011. All rights reserved.
Note to U.S. Government Users Restricted Rights -- Use, duplication or disclosure restricted by GSA ADP Schedule
Contract with IBM Corp.
First Edition (December 2011)
This edition applies to the IBM Power 770 (9117-MMC) and Power 780 (9179-MHC) Power Systems servers.
Note: Before using this information and the product it supports, read the information in “Notices” on
page vii.

© Copyright IBM Corp. 2011. All rights reserved. iii
Contents
Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
The team who wrote this paper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Now you can become a published author, too! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Comments welcome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Stay connected to IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii
Chapter 1. General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Systems overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1 IBM Power 770 server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.2 IBM Power 780 server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Operating environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Physical package . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 System features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.4.1 Power 770 system features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.4.2 Power 780 system features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4.3 Minimum features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.4.4 Power supply features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4.5 Processor card features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4.6 Summary of processor features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.4.7 Memory features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.5 Disk and media features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.6 I/O drawers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.6.1 PCI-DDR 12X Expansion Drawers (#5796) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.6.2 12X I/O Drawer PCIe (#5802 and #5877). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.6.3 EXP 12S SAS Drawer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.6.4 EXP 24S SFF Gen2-bay Drawer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.6.5 I/O drawers and usable PCI slot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.7 Comparison between models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.8 Build to Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.9 IBM Editions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.10 Model upgrade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.11 Hardware Management Console models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
1.12 System racks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1.12.1 IBM 7014 model T00 rack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1.12.2 IBM 7014 model T42 rack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.12.3 IBM 7014 model S25 rack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.12.4 Feature number 0555 rack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.12.7 The AC power distribution unit and rack content . . . . . . . . . . . . . . . . . . . . . . . . 31
1.12.8 Rack-mounting rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.12.9 Useful rack additions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Chapter 2. Architecture and technical overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.1 The IBM POWER7 processor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.1.1 POWER7 processor overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

iv IBM Power 770 and 780 Technical Overview and Introduction
2.1.2 POWER7 processor core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.1.3 Simultaneous multithreading. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.1.4 Memory access. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.1.5 Flexible POWER7 processor packaging and offerings . . . . . . . . . . . . . . . . . . . . . 44
2.1.6 On-chip L3 cache innovation and Intelligent Cache . . . . . . . . . . . . . . . . . . . . . . . 46
2.1.7 POWER7 processor and Intelligent Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.1.8 Comparison of the POWER7 and POWER6 processors . . . . . . . . . . . . . . . . . . . 47
2.2 POWER7 processor cards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.2.1 Two-socket processor card. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.2.2 Four-socket processor card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.2.3 Processor comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2.3 Memory subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.3.1 Fully buffered DIMM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.3.2 Memory placement rules. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.3.3 Memory throughput. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
2.3.4 Active Memory Mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.4 Capacity on Demand. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.4.1 Capacity Upgrade on Demand (CUoD). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.4.2 On/Off Capacity on Demand (On/Off CoD). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.4.3 Utility Capacity on Demand (Utility CoD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.4.4 Trial Capacity On Demand (Trial CoD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.4.5 Software licensing and CoD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.5 CEC Enclosure interconnection cables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.6 System bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.6.1 I/O buses and GX++ card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.6.2 Flexible Service Processor bus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2.7 Internal I/O subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2.7.1 Blind-swap cassettes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.7.2 System ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.8 PCI adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.8.1 PCIe Gen1 and Gen2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.8.2 PCI-X adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.8.3 IBM i IOP adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.8.4 PCIe adapter form factors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
2.8.5 LAN adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2.8.6 Graphics accelerator adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
2.8.7 SCSI and SAS adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
2.8.8 iSCSI adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
2.8.9 Fibre Channel adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.8.10 Fibre Channel over Ethernet. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.8.11 InfiniBand Host Channel adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
2.8.12 Asynchronous adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
2.9 Internal storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
2.9.1 Dual split backplane mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
2.9.2 Triple split backplane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
2.9.3 Dual storage IOA configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
2.9.4 DVD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
2.10 External I/O subsystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2.10.1 PCI-DDR 12X Expansion drawer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2.10.2 12X I/O Drawer PCIe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
2.10.3 Dividing SFF drive bays in 12X I/O drawer PCIe . . . . . . . . . . . . . . . . . . . . . . . . 87
2.10.4 12X I/O Drawer PCIe and PCI-DDR 12X Expansion Drawer 12X cabling . . . . . 90
2.10.5 12X I/O Drawer PCIe and PCI-DDR 12X Expansion Drawer SPCN cabling . . . 91

Contents v
2.11 External disk subsystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
2.11.1 EXP 12S Expansion Drawer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
2.11.2 EXP24S SFF Gen2-bay Drawer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
2.11.3 TotalStorage EXP24 disk drawer and tower . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
2.11.4 IBM TotalStorage EXP24 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
2.11.5 IBM System Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
2.12 Hardware Management Console . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.12.1 HMC functional overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
2.12.2 HMC connectivity to the POWER7 processor-based systems . . . . . . . . . . . . . 102
2.12.3 High availability using the HMC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
2.13 IBM Systems Director Management Console . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
2.14 Operating system support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
2.14.1 Virtual I/O Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
2.14.2 IBM AIX operating system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
2.14.3 IBM i operating system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
2.14.4 Linux operating system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
2.14.5 Java supported versions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
2.14.6 Boosting performance and productivity with IBM compilers . . . . . . . . . . . . . . . 112
2.15 Energy management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
2.15.1 IBM EnergyScale technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
2.15.2 Thermal power management device card. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Chapter 3. Virtualization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
3.1 POWER Hypervisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
3.2 POWER processor modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
3.3 Active Memory Expansion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
3.4 PowerVM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
3.4.1 PowerVM editions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.4.2 Logical partitions (LPARs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.4.3 Multiple Shared Processor Pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
3.4.4 Virtual I/O Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
3.4.5 PowerVM Live Partition Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
3.4.6 Active Memory Sharing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
3.4.7 Active Memory Deduplication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
3.4.8 N_Port ID virtualization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
3.4.9 Operating system support for PowerVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
3.4.10 POWER7 Linux programming support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
3.5 System Planning Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Chapter 4. Continuous availability and manageability . . . . . . . . . . . . . . . . . . . . . . . . 153
4.1 Reliability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
4.1.1 Designed for reliability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
4.1.2 Placement of components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
4.1.3 Redundant components and concurrent repair. . . . . . . . . . . . . . . . . . . . . . . . . . 156
4.2 Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
4.2.1 Partition availability priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
4.2.2 General detection and deallocation of failing components . . . . . . . . . . . . . . . . . 157
4.2.3 Memory protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
4.2.4 Active Memory Mirroring for Hypervisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
4.2.5 Cache protection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
4.2.6 Special uncorrectable error handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
4.2.7 PCI enhanced error handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
4.2.8 POWER7 I/O chip freeze behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168

vi IBM Power 770 and 780 Technical Overview and Introduction
4.3 Serviceability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
4.3.1 Detecting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
4.3.2 Diagnosing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
4.3.3 Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
4.3.4 Notifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
4.3.5 Locating and servicing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
4.4 Manageability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
4.4.1 Service user interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
4.4.2 IBM Power Systems firmware maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
4.4.3 Electronic Services and Electronic Service Agent . . . . . . . . . . . . . . . . . . . . . . . 190
4.5 Operating system support for RAS features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Related publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Other publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Online resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Help from IBM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

© Copyright IBM Corp. 2011. All rights reserved. vii
Notices
This information was developed for products and services offered in the U.S.A.
IBM may not offer the products, services, or features discussed in this document in other countries. Consult
your local IBM representative for information on the products and services currently available in your area. Any
reference to an IBM product, program, or service is not intended to state or imply that only that IBM product,
program, or service may be used. Any functionally equivalent product, program, or service that does not
infringe any IBM intellectual property right may be used instead. However, it is the user's responsibility to
evaluate and verify the operation of any non-IBM product, program, or service.
IBM may have patents or pending patent applications covering subject matter described in this document. The
furnishing of this document does not give you any license to these patents. You can send license inquiries, in
writing, to:
IBM Director of Licensing, IBM Corporation, North Castle Drive, Armonk, NY 10504-1785 U.S.A.
The following paragraph does not apply to the United Kingdom or any other country where such
provisions are inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION
PROVIDES THIS PUBLICATION "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR
IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF NON-INFRINGEMENT,
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer of
express or implied warranties in certain transactions, therefore, this statement may not apply to you.
This information could include technical inaccuracies or typographical errors. Changes are periodically made
to the information herein; these changes will be incorporated in new editions of the publication. IBM may make
improvements and/or changes in the product(s) and/or the program(s) described in this publication at any time
without notice.
Any references in this information to non-IBM websites are provided for convenience only and do not in any
manner serve as an endorsement of those websites. The materials at those websites are not part of the
materials for this IBM product and use of those websites is at your own risk.
IBM may use or distribute any of the information you supply in any way it believes appropriate without incurring
any obligation to you.
Any performance data contained herein was determined in a controlled environment. Therefore, the results obtained in
other operating environments may vary significantly. Some measurements may have been made on development-level
systems and there is no guarantee that these measurements will be the same on generally available systems.
Furthermore, some measurement may have been estimated through extrapolation. Actual results may vary. Users of this
document should verify the applicable data for their specific environment.
Information concerning non-IBM products was obtained from the suppliers of those products, their published
announcements or other publicly available sources. IBM has not tested those products and cannot confirm the
accuracy of performance, compatibility or any other claims related to non-IBM products. Questions on the
capabilities of non-IBM products should be addressed to the suppliers of those products.
This information contains examples of data and reports used in daily business operations. To illustrate them
as completely as possible, the examples include the names of individuals, companies, brands, and products.
All of these names are fictitious and any similarity to the names and addresses used by an actual business
enterprise is entirely coincidental.
COPYRIGHT LICENSE:
This information contains sample application programs in source language, which illustrate programming
techniques on various operating platforms. You may copy, modify, and distribute these sample programs in
any form without payment to IBM, for the purposes of developing, using, marketing or distributing application
programs conforming to the application programming interface for the operating platform for which the sample
programs are written. These examples have not been thoroughly tested under all conditions. IBM, therefore,
cannot guarantee or imply reliability, serviceability, or function of these programs.

viii IBM Power 770 and 780 Technical Overview and Introduction
Trademarks
IBM, the IBM logo, and ibm.com are trademarks or registered trademarks of International Business Machines
Corporation in the United States, other countries, or both. These and other IBM trademarked terms are
marked on their first occurrence in this information with the appropriate symbol (® or ™), indicating US
registered or common law trademarks owned by IBM at the time this information was published. Such
trademarks may also be registered or common law trademarks in other countries. A current list of IBM
trademarks is available on the Web at http://www.ibm.com/legal/copytrade.shtml
The following terms are trademarks of the International Business Machines Corporation in the United States,
other countries, or both:
Active Memory™
AIX®
BladeCenter®
DS8000®
Electronic Service Agent™
EnergyScale™
Focal Point™
IBM Systems Director Active Energy
Manager™
IBM®
Micro-Partitioning®
POWER Hypervisor™
Power Systems™
Power Systems Software™
POWER6+™
POWER6®
POWER7®
PowerHA®
PowerPC®
PowerVM®
Power®
POWER®
pSeries®
Rational Team Concert™
Rational®
Redbooks®
Redpaper™
Redbooks (logo) ®
Storwize®
System Storage®
System x®
System z®
Tivoli®
XIV®
The following terms are trademarks of other companies:
Intel Xeon, Intel, Intel logo, Intel Inside logo, and Intel Centrino logo are trademarks or registered trademarks
of Intel Corporation or its subsidiaries in the United States and other countries.
LTO, Ultrium, the LTO Logo and the Ultrium logo are trademarks of HP, IBM Corp. and Quantum in the U.S.
and other countries.
Microsoft, and the Windows logo are trademarks of Microsoft Corporation in the United States, other
countries, or both.
Java, and all Java-based trademarks and logos are trademarks or registered trademarks of Oracle and/or its
affiliates.
Intel, Intel logo, Intel Inside, Intel Inside logo, Intel Centrino, Intel Centrino logo, Celeron, Intel Xeon, Intel
SpeedStep, Itanium, and Pentium are trademarks or registered trademarks of Intel Corporation or its
subsidiaries in the United States and other countries.
Linux is a trademark of Linus Torvalds in the United States, other countries, or both.
Other company, product, or service names may be trademarks or service marks of others.

© Copyright IBM Corp. 2011. All rights reserved. ix
Preface
This IBM® Redpaper™ publication is a comprehensive guide covering the IBM Power® 770
(9117-MMC) and Power 780 (9179-MHC) servers supporting IBM AIX®, IBM i, and Linux
operating systems. The goal of this paper is to introduce the major innovative Power 770 and
780 offerings and their prominent functions, including these:
The IBM POWER7® processor available at frequencies of 3.3 GHz, 3.44 GHz, 3.72 GHz,
and 3.92 GHz, and 4.14 GHz
The specialized IBM POWER7 Level 3 cache that provides greater bandwidth, capacity,
and reliability
The 1 Gb or 10 Gb Integrated Multifunction Card that provides two USB ports, one
serial port, and four Ethernet connectors for a processor enclosure and does not require a
PCI slot
The new Active Memory™ Mirroring (AMM) for Hypervisor feature that mirrors the main
memory used by the firmware
IBM PowerVM® virtualization, including PowerVM Live Partition Mobility and PowerVM
Active Memory Sharing
Active Memory Expansion that provides more usable memory than what is physically
installed on the system
IBM EnergyScale™ technology that provides features such as power trending,
power-saving, capping of power, and thermal measurement
Enterprise-ready reliability, serviceability, and availability
Professionals who want to acquire a better understanding of IBM Power Systems™ products
should read this Redpaper publication. The intended audience includes the following areas:
Clients
Sales and marketing professionals
Technical support professionals
IBM Business Partners
Independent software vendors
This Redpaper publication expands the current set of IBM Power Systems documentation by
providing a desktop reference that offers a detailed technical description of the Power 770
and Power 780 systems.
This paper does not replace the latest marketing materials and configuration tools. It is
intended as an additional source of information that, together with existing sources, can be
used to enhance your knowledge of IBM server solutions.
The team who wrote this paper
This paper was produced by a team of specialists from around the world working at the
International Technical Support Organization, Poughkeepsie Center.
Alexandre Bicas Caldeira works on the Power Systems Field Technical Sales Support
Team for IBM Brazil. He holds a degree in computer science from the Universidade Estadual
Paulista (UNESP). Alexandre has more than 11 years of experience working for IBM and IBM

x IBM Power 770 and 780 Technical Overview and Introduction
Business Partners on Power Systems hardware, AIX, and PowerVM virtualization products.
He is also skilled on IBM System Storage®, IBM Tivoli® Storage Manager, IBM System x®,
and VMware.
Carlo Costantini is a Certified IT Specialist for IBM and has over 33 years of experience
with IBM and IBM Business Partners. He currently works in Italy Power Systems Platforms
as Presales Field Technical Sales Support for IBM Sales Representatives and IBM
Business Partners. Carlo has broad marketing experience, and his current areas of focus
are competition, sales, and technical sales support. He is a Certified Specialist for
Power Systems servers. He holds a master’s degree in Electronic Engineering from
Rome University.
Steve Harnett is a Senior Accredited Professional, Chartered IT Professional, and a member
of the British Computing Society. He works as a pre-sales Technical Consultant in the IBM
Server and Technology Group in the UK. Steve has over 16 years of experience of working in
post sales supporting Power Systems. He is a product Topgun and a recognized SME in
Electronic Service Agent™, Hardware Management Console, and high-end Power Systems.
He also has several years of experience in developing and delivering education to clients,
IBM Business Partners, and IBMers.
Volker Haug is a certified Consulting IT Specialist within the IBM Systems and Technology
Group based in Ehningen, Germany. He holds a bachelor's degree in business management
from the University of Applied Studies in Stuttgart. His career has included more than 24
years working in the IBM PLM and Power Systems divisions as a RISC and AIX Systems
Engineer. Volker is an expert in Power Systems hardware, AIX, and PowerVM virtualization.
He is a POWER7 Champion and a member of the German Technical Expert Council, an
affiliate of the IBM Academy of Technology. He has written several books and white papers
about AIX, workstations, servers, and PowerVM virtualization.
Craig Watson has 15 years of experience working with UNIX-based systems in roles
including field support, systems administration, and technical sales. He has worked in the
IBM Systems and Technology group since 2003 and is currently working as a Systems
Architect, designing complex solutions for customers that include Power Systems, System x,
and Systems Storage. He holds a master’s degree in electrical and electronic engineering
from the University of Auckland.
Fabien Willmann is an IT Specialist working with Techline Power Europe in France. He
has 10 years of experience with Power Systems, AIX, and PowerVM virtualization. After
teaching hardware courses about Power Systems servers, he joined ITS as an AIX
consultant, where he developed his compentencies in AIX, HMC management, and
PowerVM virtualization. Building new Power Systems configurations for STG presales is his
major area of expertise today. He recently gave a workshop on the econfig configuration tool,
focused on POWER7 processor-based BladeCenters during the symposium for French
Business Partners in Montpellier.
The project that produced this publication was managed by:
Scott Vetter, IBM Certified Project Manager and PMP.

Preface xi
Thanks to the following people for their contributions to this project:
Larry Amy, Gary Anderson, Sue Beck, Terry Brennan, Pat Buckland, Paul D. Carey,
Pete Heyrman, John Hilburn, Dan Hurlimann, Kevin Kehne, James Keniston, Jay Kruemcke,
Robert Lowden, Hilary Melville, Thoi Nguyen, Denis C. Nizinski, Pat O’Rourke, Jan Palmer,
Ed Prosser, Robb Romans, Audrey Romonosky, Todd Rosedahl, Melanie Steckham,
Ken Trusits, Al Yanes
IBM U.S.A.
Stephen Lutz
IBM Germany
Tamikia Barrow
International Technical Support Organization, Poughkeepsie Center
Now you can become a published author, too!
Here’s an opportunity to spotlight your skills, grow your career, and become a published
author—all at the same time! Join an ITSO residency project and help write a book in your
area of expertise, while honing your experience using leading-edge technologies. Your efforts
will help to increase product acceptance and customer satisfaction, as you expand your
network of technical contacts and relationships. Residencies run from two to six weeks in
length, and you can participate either in person or as a remote resident working from your
home base.
Find out more about the residency program, browse the residency index, and apply online at:
ibm.com/redbooks/residencies.html
Comments welcome
Your comments are important to us!
We want our papers to be as helpful as possible. Send us your comments about this paper or
other IBM Redbooks® publications in one of the following ways:
Use the online Contact us review Redbooks form found at:
ibm.com/redbooks
Send your comments in an email to:
redbooks@us.ibm.com
Mail your comments to:
IBM Corporation, International Technical Support Organization
Dept. HYTD Mail Station P099
2455 South Road
Poughkeepsie, NY 12601-5400

xii IBM Power 770 and 780 Technical Overview and Introduction
Stay connected to IBM Redbooks
Find us on Facebook:
http://www.facebook.com/IBMRedbooks
Follow us on Twitter:
http://twitter.com/ibmredbooks
Look for us on LinkedIn:
http://www.linkedin.com/groups?home=&gid=2130806
Explore new Redbooks publications, residencies, and workshops with the IBM Redbooks
weekly newsletter:
https://www.redbooks.ibm.com/Redbooks.nsf/subscribe?OpenForm
Stay current on recent Redbooks publications with RSS Feeds:
http://www.redbooks.ibm.com/rss.html

© Copyright IBM Corp. 2011. All rights reserved. 1
Chapter 1. General description
The IBM Power 770 (9117-MMC) and IBM Power 780 servers (9179-MHC) utilize the latest
POWER7 processor technology designed to deliver unprecedented performance, scalability,
reliability, and manageability for demanding commercial workloads.
The innovative IBM Power 770 and Power 780 servers with POWER7 processors are
symmetric multiprocessing (SMP), rack-mounted servers. These modular-built system uses
one to four enclosures. Each enclosure is four EIA units (4U) tall and is housed in a 19-inch
rack.
New to the Power 770 and Power 780 models are two new, powerful POWER7 processor
cards. Each drawer contains a powerful POWER7 processor card and an enhanced
POWER7 I/O backplane.
1

2 IBM Power 770 and 780 Technical Overview and Introduction
1.1 Systems overview
You can find detailed information about the Power 770 and Power 780 systems within the
following sections.
1.1.1 IBM Power 770 server
Each Power 770 processor card features 64-bit architecture designed with two single-chip
module (SCM) POWER7 processors. Each POWER7 SCM enables up to either six or eight
active processor cores with 2 MB of L2 cache (256 KB per core), 24 MB of L3 cache (4 MB
per core) for the 6-core SCM, and 32 MB of L3 cache (4 MB per core) for the 8-core SCM.
A Power 770 server using 6-core SCM processors will enable up to 48 processor cores
running at frequencies of 3.72 GHz. A system configured with up to four CEC enclosures
using 8-core SCM processors will enable up to 64 processor cores running at frequencies up
to 3.30 GHz. The Power 770 server is available starting as low as four active cores and
incrementing one core at a time through built-in Capacity on Demand (CoD) functions to a
maximum of 64 active cores.
A single Power 770 CEC enclosure is equipped with 16 DIMM slots running at speeds up to
1066 MHz. A system configured with four drawers and 64 GB DDR3 DIMMs supports up to a
maximum of 4.0 TB of DDR3 memory. All POWER7 DDR3 memory uses memory
architecture that provides increased bandwidth and capacity. This enables operating at a
higher data rate for large memory configurations.
The Power 770 has two new integrated POWER7 I/O controllers that enhance I/O
performance while supporting a maximum of six internal PCIe adapters and six internal small
form-factor SAS DASD bays.
The Power 770 features Active Memory Mirroring (AMM) for Hypervisor, which is available as
an optional feature. AMM guards against system-wide outages due to any uncorrectable error
associated with firmware. Also available as an option is Active Memory Expansion, which
enhances memory capacity.

Chapter 1. General description 3
Figure 1-1 shows a Power 770 with the maximum four enclosures, and the front and rear
views of a single-enclosure Power 770.
Figure 1-1 Four-enclosure Power 770, a single-enclosure Power 770 front and rear views
1.1.2 IBM Power 780 server
Each Power 780 processor card comprises either two single-chip module (SCM) POWER7
processors or four SCM POWER7 processors, each designed with 64-bit architecture. Each
POWER7 SCM enables either up to six or eight active processor cores with 2 MB of L2 cache
(256 KB per core, 24 MB of L3 cache (4 MB per core) for the 6-core SCM, and 32 MB of L3
cache (4 MB per core) for the 8-core SCM.
For the Power 780, each POWER7 SCM processor is available at frequencies of 3.44 GHz
with six cores, 3.92 GHz with eight cores, or 4.14 GHz with four cores. The Power 780 server
is available starting as low as four active cores and incrementing one core at a time through
built-in Capacity on Demand (CoD) functions to a maximum of 96 active cores.
The Power 780 features the truly unique ability to switch between its standard throughput
optimized mode and its unique TurboCore mode. In TurboCore mode performance per core
is boosted with access to both additional cache and additional clock speed. Based on the
user's configuration option, any Power 780 system can be booted in standard mode, enabling
up to a maximum of 64 processor cores running at 3.92 GHz, or in TurboCore mode, enabling
up to 32 processor cores running at 4.14 GHz and twice the cache per core.
A single Power 770 CEC enclosure is equipped with 16 DIMM slots running at speeds up to
1066 MHz. A system configured with four drawers and 64 GB DDR3 DIMMs supports up to a
maximum of 4.0 TB of DDR3 memory. All POWER7 DDR3 memory uses memory
architecture that provides increased bandwidth and capacity. This enables operating at a
higher data rate for large memory configurations.
Power 770 single enclosure front viewPower 770 with 4 enclosures
Power 770 single enclosure rear view
Note: TurboCore mode is not supported on the 6-core processor card.

The Power 780 has two new integrated POWER7 I/O controllers that enhance I/O
performance while supporting a maximum of six internal PCIe adapters and six internal small
form-factor SAS DASD bays.
The Power 780 features AMM for Hypervisor, which is available as a standard feature.
AMM guards against system-wide outages due to any uncorrectable error associated with
firmware. Also available as an option is Active Memory Expansion, which enhances
memory capacity.
1.2 Operating environment
Table 1-1 lists the operating environment specifications for the servers.
Table 1-1 Operating environment for Power 770 and Power 780 (for one enclosure only)
Power 770 and Power 780 operating environment
Description Operating Non-operating
Temperature 5 - 35 degrees C
(41 to 95 degrees F)
5 - 45 degrees C
(41 - 113 degrees F)
Relative humidity 20 - 80% 8 - 80%
Maximum dew point 29 degrees C
(84 degrees F)
28 degrees C
(82 degrees F)
Operating voltage 200 - 240 V ac Not applicable
Operating frequency 50 - 60 +/- 3 Hz Not applicable
Power consumption Power 770:
1,600 watts maximum (per
enclosure with 16 cores active)
Power 780:
1,900 watts maximum (per
Not applicable
Power source loading Power 770:
1.649 kVA maximum (per
Power 780:
1.959 kVA maximum (per
Not applicable
Thermal output Power 770:
5,461 Btu/hr maximum (per
Power 780:
6,485 Btu/hr maximum (per
Not applicable
Maximum altitude 3048 m
(10,000 ft)
Not applicable

1.3 Physical package
Table 1-2 lists the physical dimensions of an individual enclosure. Both servers are available
only in a rack-mounted form factor. They are modular systems that can be constructed from
one to four building-block enclosures. Each of these enclosures can take 4U (EIA units) of
rack space. Thus, a two-enclosure system requires 8U, three enclosures require 12U, and
four enclosures require 16U.
Table 1-2 Physical dimensions of a Power 770 and Power 780 enclosure
Noise level for one enclosure Power 770 (one enclosure with 16 active cores):
7.1 bels (operating/idle)
6.6 bels (operating/idle) with acoustic rack doors
Power 780 (one enclosure with 24 active cores):
Noise level for four enclosures Power 770 (four enclosure with 64 active cores):
Power 780 (four enclosure with 96 active cores):
Power 770 and Power 780 operating environment
Description Operating Non-operating
Dimension Power 770 (Model 9117-MMC)
single enclosure
Power 780 (Model 9179-MHC)
single enclosure
Width 483 mm (19.0 in) 483 mm (19.0 in)
Depth 863 mm (32.0 in) 863 mm (32.0 in)
Height 174 mm (6.85 in), 4U (EIA units) 174 mm (6.85 in), 4U (EIA units)
Weight 70.3 kg (155 lb) 70.3 kg (155 lb)

Figure 1-2 shows the front and rear views of the Power 770 and Power 780.
Figure 1-2 Front and rear views of the Power 770 and Power 780
1.4 System features
The Power 770 processor card features 64-bit architecture designed with two single-chip
module (SCM) POWER7 processors. The Power 780 processor card comprises either two
single-chip module (SCM) POWER7 processors or four SCM POWER7 processors, each
designed with 64-bit architecture.
Each POWER7 SCM enables either up to six or eight active processor cores with 2 MB of L2
cache (256 KB per core), 24 MB of L3 cache (4 MB per core) for the 6-core SCM, and 32 MB
of L3 cache (4 MB per core) for the 8-core SCM.
1.4.1 Power 770 system features
The following features are available on the Power 770:
4U 19-inch rack-mount system enclosure
One to four system enclosures: 16U maximum system size
One processor card feature per enclosure (includes the voltage regulator):
– 0/12-core, 3.72 GHz processor card (#4983)
– 0/16-core, 3.3 GHz processor card (#4984)
POWER7 DDR3 Memory DIMMs (16 DIMM slots per CEC enclosure):
– 0/32 GB (4 X 8 GB), 1066 MHz (#5600)
– 0/64 GB (4 X 16 GB), 1066 MHz (#5601)
– 0/128 GB (4 X 32 GB), 1066 MHz (#5602)
– 0/256 GB (4 X 64 GB), 1066 MHz (#5564)
Six hot-swappable, 2.5-inch, small form factor, SAS disk or SSD bays per enclosure
GX++ Bus
Integrated
Ports
Power
Supplies
FSP
connectors
P
C
I
e
P
C
I
e
P
C
I
e
P
C
I
e
P
C
I
e
P
C
I
e
SPCN
Ports
Serial
Port
USB
Ports
Power 770 and Power 780 enclosure rear view
HMC
Ports

One hot-plug, slim-line, SATA media bay per enclosure (optional)
Redundant hot-swap AC power supplies in each enclosure
Choice of Integrated Multifunction Card options; maximum one per enclosure:
– Dual 10 Gb Copper and Dual 1 Gb Ethernet (#1768)
– Dual 10 Gb Optical and Dual 1 Gb Ethernet (#1769)
One serial port included on each Integrated Multifunction Card
Two USB ports included on each Integrated Multifunction Card, plus another USB port on
each enclosure (maximum nine usable per system)
Two HMC ports per enclosure (maximum four per system)
Eight I/O expansion slots per enclosure (maximum 32 per system)
– Six Gen2 PCIe 8x slots plus two GX++ slots per enclosure
Dynamic LPAR support, Processor and Memory CUoD
PowerVM (optional)
– Micro-Partitioning®
– Virtual I/O Server (VIOS)
– Automated CPU and memory reconfiguration support for dedicated and shared
processor logical partition groups (dynamic LPAR)
– Support for manual provisioning of resources, namely PowerVM Live Partition
Migration (PowerVM Enterprise Edition)
Optional PowerHA® for AIX, IBM i, and Linux
12X I/O drawer with PCI slots
– Up to 16 PCIe I/O drawers (#5802 or #5877)
– Up to 32 PCI-X DDR I/O drawers (7314-G30 or #5796)
Additional considerations: Note the following considerations:
The Ethernet port of the Integrated Multifunction Card cannot be used for an IBM i
console. Use separate Ethernet adapters that can be directly controlled by IBM i
without the Virtual I/O server for IBM i LAN consoles if desired. Alternatively, an
HMC can also be used for an IBM i console.
The first and second CEC enclosure must contain one Integrated Multifunction Card
(#1768 or #1769). The Integrated Multifunction Card is optional for the third or fourth
CEC enclosure.
Each Integrated Multifunction Card has four Ethernet ports, two USB ports, and one
serial port. Usage of the serial port by AIX/Linux is supported for MODEM call home,
TTY console, and snooping even if an HMC/SDMC is attached to the server. Usage
by the serial port to communicate with a UPS is not supported.
The first and second CEC enclosures each have two HMC/SDMC ports on the
service processor (#EU05). If there are two CEC enclosures, the HMC must be
connected to both service processor cards.

Disk-only I/O drawers
– Up to 56 EXP24S SFF SAS I/O drawers on external SAS controller (#5887)
– Up to 110 EXP12S SAS DASD/SSD I/O drawers on SAS PCI controllers (#5886)
– Up to 60 EXP24 SCSI DASD Expansion drawers on SCSI PCI controllers (7031-D24)
IBM Systems Director Active Energy Manager™
The Power 770 operator interface controls located on the front panel of the primary I/O
drawer consist of a power ON/OFF button with a POWER® indicator, an LCD display for
diagnostic feedback, a RESET button, and a disturbance or system attention LED.
1.4.2 Power 780 system features
The following features are available on the Power 780:
4U 19-inch rack-mount system enclosure
One to four system enclosures: 16U maximum system size
One processor card feature per enclosure (includes the voltage regulator):
– 0/16 core, 3.92 GHz or 0/8 core, 4.14 GHz (TurboCore) processor card (#5003)
– 0/24 core, 3.44 GHz processor card (#EP24)
POWER7 DDR3 Memory DIMMs (16 DIMM slots per processor card):
– 0/32 GB (4 X 8 GB), 1066 MHz (#5600)
– 0/64 GB (4 X 16 GB), 1066 MHz (#5601)
– 0/128 GB (4 X 32 GB), 1066 MHz (#5602)
– 0/256 GB (4 X 64 GB), 1066 MHz (#5564)
Six hot-swappable, 2.5-inch, small form factor, SAS disk or SSD bays per enclosure
One hot-plug, slim-line, SATA media bay per enclosure (optional)
Redundant hot-swap AC power supplies in each enclosure
Choice of Integrated Multifunction Card options; maximum one per enclosure:
One serial port included on each Integrated Multifunction Card
Two USB ports included on each Integrated Multifunction Card plus another USB port on
each enclosure (maximum nine usable per system)

Two HMC ports per enclosure (maximum four per system)
Eight I/O expansion slots per enclosure (maximum 32 per system)
– Six Gen2 PCIe 8x slots plus two GX++ slots per enclosure
Dynamic LPAR support, Processor and Memory CUoD
PowerVM (optional):
– Micro-Partitioning
– Virtual I/O Server (VIOS)
– Automated CPU and memory reconfiguration support for dedicated and shared
processor logical partition (LPAR) groups
– Support for manual provisioning of resources partition migration (PowerVM
Enterprise Edition)
Optional PowerHA for AIX, IBM i, and Linux
12X I/O drawer with PCI slots
– Up to 16 PCIe I/O drawers (#5802 or #5877)
– Up to 32 PCI-X DDR I/O drawers (7314-G30 or feature #5796)
Disk-only I/O drawers
– Up to 56 EXP24S SFF SAS I/O drawers on external SAS controller (#5887)
– Up to 110 EXP12S SAS DASD/SSD I/O drawers on SAS PCI controllers (#5886)
– Up to 60 EXP24 SCSI DASD Expansion drawers on SCSI PCI controllers (7031-D24)
IBM Systems Director Active Energy Manager
The Power 780 operator interface/controls located on the front panel of the primary I/O
drawer consist of a power ON/OFF button with a POWER indicator, an LCD display for
diagnostic feedback, a RESET button, and a disturbance or system attention LED.
Additional considerations: Note the following considerations:
The Ethernet ports of the Integrated Multifunction Card cannot be used for an IBM i
console. Separate Ethernet adapters that can be directly controlled by IBM i without
the Virtual I/O server should be used for IBM i LAN consoles if desired. Alternatively,
an HMC can also be used for an IBM i console.
The first and second CEC enclosure must contain one Integrated Multifunction Card
(#1768 or #1769). The Integrated Multifunction Card is optional for the third or fourth
CEC enclosure.
Each Integrated Multifunction Card has four Ethernet ports, two USB ports, and one
serial port. Usage of the serial port by AIX/Linux is supported for MODEM call home,
TTY console, and snooping even if an HMC/SDMC is attached to the server. Usage
by the serial port to communicate with a UPS is not supported.
The first and second CEC enclosures each have two HMC/SDMC ports on the
service processor (#EU05). If there are two CEC enclosures, the HMC must be
connected to both service processor cards.

1.4.3 Minimum features
Each system has a minimum feature set in order to be valid. Table 1-3 shows the minimum
system configuration for a Power 770.
Table 1-3 Minimum features for Power 770 system
Power 770 minimum features Additional notes
1x CEC enclosure (4U) 1x System Enclosure with IBM Bezel (#5585) or OEM Bezel
(#5586)
1x Service Processor (#5664)
1x DASD Backplane (#5652)
2x Power Cords (two selected by customer)
– 2x A/C Power Supply (#5632)
1x Operator Panel (#1853)
1x Integrated Multifunction Card options (one of these):
1x primary operating system (one
of these)
AIX (#2146)
Linux (#2147)
IBM i (#2145)
1x Processor Card 0/12-core, 3.72 GHz processor card (#4983)
0/16-core, 3.3 GHz processor card (#4984)
4x Processor Activations
(quantity of four for one of these)
One Processor Activation for processor feature #4983
(#5329)
One Processor Activation for processor feature #4984
(#5334)
2x DDR3 Memory DIMMs (one of
these)
0/32 GB (4 X 8 GB), 1066 MHz (#5600)
0/64 GB (4 X 16 GB), 1066 MHz (#5601)
0/128 GB (4 X 32 GB), 1066 MHz (#5602)
0/256 GB (4 X 64 GB), 1066 MHz (#5564)
32x Activations of 1 GB DDR3 -
POWER7 Memory (#8212)
-
For AIX and Linux: 1x disk drive
For IBM i: 2x disk drives
Formatted to match the system Primary O/S indicator selected,
or if using a Fibre Channel attached SAN (indicated by #0837) a
disk drive is not required.
1X Language Group (selected by
the customer)
-
1x Removable Media Device
(#5762)
Optionally orderable, a standalone system (not network
attached) would required this feature.
1x HMC Required for every Power 770 (9117-MMC)

Table 1-4 shows the minimum system configuration for a Power 780 system.
Table 1-4 Minimum features for Power 780 system
Note: Consider the following:
A minimum number of four processor activations must be ordered per system.
The minimum activations ordered with all initial orders of memory features #5600, #5601, and
#5602 must be 50% of their installed capacity.
The minimum activations ordered with MES orders of memory features #5600, #5601, and #5602
will depend on the total installed capacity of features #5600, #5601, and #5602. This allows
newly ordered memory to be purchased with less than 50% activations when the currently
installed capacity exceeds 50% of the existing features #5600, #5601, and #5602 capacity.
The minimum activations ordered with all initial orders of memory feature #5564 must be 192 GB
of 256 GB per each feature #5564 ordered (that is, 75% of the installed feature #5564 capacity).
The minimum activations purchased with MES orders of feature #5564 memory, 0/256 GB, will
depend on the total installed capacity of feature #5564. This allows MES orders of feature #5564
memory to be purchased with less than 192/256 GB per each feature #5564 ordered when the
system activations currently installed exceed 75% of the existing feature #5564 capacity.
1x CEC enclosure (4U) 1x System Enclosure with IBM Bezel (#5595) or OEM Bezel
(#5596)
1x Service Processor (#5664)
1x DASD Backplane (#5652)
2x Power Cords (two selected by customer)
– 2x A/C Power Supply (#5532)
1x Operator Panel (#1853)
1x Integrated Multifunction Card options (one of these):
1x primary operating system (one
of these)
AIX (#2146)
Linux (#2147)
IBM i (#2145)
1x Processor Card (one of these) 0/16 core, 3.92 GHz or 0/8 core, 4.14 GHz (TurboCore)
processor card (#5003)
0/24 core, 3.44 GHz processor card (#EP24)
4x Processor Activations for
Processor Feature #4982 (#5469)
-
2x DDR3 Memory DIMM (one of
these)
0/32 GB (4 X 8 GB), 1066 MHz (#5600)
0/64 GB (4 X 16 GB), 1066 MHz (#5601)
0/128 GB (4 X 32 GB), 1066 MHz (#5602)
0/256 GB (4 X 64 GB), 1066 MHz (#5564)
32x Activations of 1 GB DDR3 -
POWER7 Memory (#8212)
-
For AIX and Linux: 1x disk drive
For IBM i: 2x disk drives
Formatted to match the system Primary O/S indicator selected,
or if using a Fibre Channel attached SAN (indicated by #0837) a
disk drive is not required.
1X Language Group (selected by
the customer)
-

1.4.4 Power supply features
Two system AC power supplies are required for each CEC enclosure. The second power
supply provides redundant power for enhanced system availability. To provide full
redundancy, the two power supplies must be connected to separate power distribution
units (PDUs).
A CEC enclosure will continue to function with one working power supply. A failed power
supply can be hot-swapped but must remain in the system until the replacement power
supply is available for exchange. The system requires one functional power supply in each
CEC enclosure to remain operational.
Each Power 770 or Power 780 server with two or more CEC enclosures must have one
Power Control Cable (#6006 or similar) to connect the service interface card in the first
enclosure to the service interface card in the second enclosure.
1.4.5 Processor card features
Each of the four system enclosures contains one powerful POWER7 processor card feature,
consisting of two single-chip module processors. Each of the POWER7 processors in the
server has a 64-bit architecture, includes six or eight cores on a single-chip module, and
contains 2 MB of L2 cache (256 KB per core), 24 MB of L3 cache (4 MB per core) for the
6-core SCM, and 32 MB of L3 cache (4 MB per core) for the 8-core SCM.
There are two types of Power 770 processor cards, offering the following features:
Two 6-core POWER7 SCMs with 24 MB of L3 cache (12-cores per processor card, each
core with 4 MB of L3 cache) at 3.72 GHz (#4983)
Two 8-core POWER7 SCMs with 32 MB of L3 cache (16-cores per processor card, each
core with 4 MB of L3 cache) at 3.3 GHz (#4984)
The Power 780 has two types of processor cards. One of these has two different processing
modes (MaxCore and TurboCore).
1x Removable Media Device
(#5762)
Optionally orderable, a standalone system (not network
attached) requires this feature.
1x HMC Required for every Power 780 (9179-MHC)
Note the following considerations:
A minimum number of four processor activations must be ordered per system.
The minimum activations ordered with all initial orders of memory features #5600, #5601, and
#5602 must be 50% of their installed capacity.
The minimum activations ordered with MES orders of memory features #5600, #5601, and #5602
will depend on the total installed capacity of features #5600, #5601, and #5602. This allows
newly ordered memory to be purchased with less than 50% activations when the currently
installed capacity exceeds 50% of the existing features #5600, #5601, and #5602 capacity.
The minimum activations ordered with all initial orders of memory feature #5564 must be 192 GB
of 256 GB per each feature #5564 ordered (that is, 75% of the installed feature #5564 capacity).
The minimum activations purchased with MES orders of feature #5564 memory, 0/256 GB, will
depend on the total installed capacity of feature #5564. This allows MES orders of feature #5564
memory to be purchased with less than 192/256 GB per each feature #5564 ordered when the
system activations currently installed exceed 75% of the existing feature #5564 capacity.

The processor card houses the two or four POWER7 SCMs and the system memory. The
Power 780 processor card offers the following features:
Feature #5003 offers two 8-core POWER7 SCMs with 32 MB of L3 cache (16 cores per
processor card are activated in MaxCore mode and each core with 4 MB of L3 cache) at
3.92 GHz.
Feature #5003 also offers two 8-core POWER7 SCMs with 32 MB of L3 cache (8 cores
per processor card are activated in TurboCore mode and each core is able to use 8 MB of
L3 cache) at 4.14 GHz.
Feature #EP24 offers four 6-core POWER7 SCMs with 24 MB of L3 cache (24 cores per
processor card, each core with 4 MB of L3 cache) at 3.44 GHz.
Figure 1-3 shows the top view of the Power 770 and Power 780 system having two SCMs
installed. The two POWER7 SCMs and the system memory reside on a single processor card
feature.
Figure 1-3 Top view of Power 770 and Power 780 system with two SCMs
POWER7
Fans
PCIe Slot
PCIe Slot
PCIe Slot
PCIe Slot
PCIe Slot
PCIe Slot
Memory
Memory
Memory
TPMD
POWER7

Figure 1-4 shows the top view of the Power 780 system having four SCMs installed. The four
POWER7 SCMs and the system memory reside on a single processor card feature.
Figure 1-4 Top view of a Power 780 system with four SCMs
In standard or MaxCore mode, the Power 780 system uses all processor cores running at
3.92 GHz and has access to the full 32 MB of L3 cache. In TurboCore mode, only four of the
eight processor cores are available, but at a higher frequency (4.14 GHz), and these four
cores have access to the full 32 MB of L3 cache. Thus, in Turbo-core mode there are fewer
cores running at a higher frequency and a higher core-to-L3-cache ratio.
For a more detailed description of MaxCore and TurboCore modes, see 2.1.5, “Flexible
POWER7 processor packaging and offerings” on page 44.
Several types of Capacity on Demand (CoD) processors are optionally available on the
Power 770 and Power 780 servers to help meet changing resource requirements in an on
demand environment by using resources installed on the system but not activated. CoD
allows you to purchase additional permanent processor or memory capacity and dynamically
activate it when needed.
More detailed information about CoD can be found in 2.4, “Capacity on Demand” on page 60.
Note: TurboCore mode is supported on the Power 780, but is not supported on the
Power 770.
POWER7
POWER7
POWER7
POWER7
Memory
Memory
Memory
TPMD
PCIe Slot #1
PCIe Slot #2
PCIe Slot #3
PCIe Slot #4
PCIe Slot #5
PCIe Slot #6
Fans

1.4.6 Summary of processor features
Table 1-5 summarizes the processor feature codes for the Power 770.
Table 1-5 Summary of processor features for the Power 770
Feature
code
Description OS
support
#4983 0/12-core 3.72 GHz POWER7 processor card:
12-core 3.72 GHz POWER7 CUoD processor planar containing two
six-core processors. Each processor has 2 MB of L2 cache (256 KB per
core) and 32 MB of L3 cache (4 MB per core). There are 16 DDR3 DIMM
slots on the processor planar (8 DIMM slots per processor), which can
be used as Capacity on Demand (CoD) memory without activating the
processors. The voltage regulators are included in this feature code.
AIX
IBM i
Linux
#5329 One processor activation for processor #4983:
Each occurrence of this feature permanently activates one processor on
Processor Card #4983. One processor activation for processor feature
#4983 with inactive processors.
AIX
IBM i
Linux
#5330 Processor CoD utility billing for #4983, 100 processor-minutes:
Provides payment for temporary use of processor feature #4983 with
supported AIX or Linux operating systems. Each occurrence of this
feature will pay for 100 minutes of usage. The purchase of this feature
occurs after the customer has 100 minutes of use on processor cores in
the Shared Processor Pool that are not permanently active.
AIX
Linux
supported IBM i operating systems. Each occurrence of this feature will
pay for 100 minutes of usage. The purchase of this feature occurs after
the customer has 100 minutes of use on processor cores in the Shared
Processor Pool that are not permanently active.
IBM i
#5332 One processor-day on/off billing for #4983:
After an On/Off Processor Enablement feature is ordered and the
associated enablement code is entered into the system, you must report
your on/off usage to IBM at least monthly. This information, used to
compute your billing data, is then provided to your sales channel. The
sales channel will place an order for a quantity of On/Off Processor Core
Day Billing features and bill you. One #5332 must be ordered for each
billable processor core day of feature #4983 used by a supported AIX or
Linux operating system.
AIX
Linux
Day Billing features and the client will be charged. One #5333 must be
ordered for each billable processor core day of feature #4983 used by a
supported IBM i operating system.
IBM i

#4984 0/16-core 3.3 GHz POWER7 processor card:
16-core 3.3 GHz POWER7 CUoD processor planar containing two
eight-core processors. Each processor has 2 MB of L2 cache (256 KB
per core) and 32 MB of L3 cache (4 MB per core). There are 16 DDR3
DIMM slots on the processor planar (8 DIMM slots per processor), which
can be used as Capacity on Demand (CoD) memory without activating
the processors. The voltage regulators are included in this feature code.
AIX
IBM i
Linux
#5334 One processor activation for processor #4984:
Each occurrence of this feature will permanently activate one processor
on Processor Card #4984. One processor activation for processor
feature #4984 with inactive processors.
AIX
IBM i
Linux
supported AIX or Linux operating systems. Each occurrence of this
feature will pay for 100 minutes of usage. The purchase of this feature
occurs after the customer has 100 minutes of use on processor cores in
the Shared Processor Pool that are not permanently active.
AIX
Linux
supported IBM i operating systems. Each occurrence of this feature will
pay for 100 minutes of usage. The purchase of this feature occurs after
the customer has 100 minutes of use on processor cores in the Shared
Processor Pool that are not permanently active.
IBM i
supported AIX or Linux operating system.
AIX
Linux
IBM i
#7951 On/off processor enablement:
This feature can be ordered to enable your server for On/Off Capacity on
Demand. After it is enabled, you can request processors on a temporary
basis. You must sign an On/Off Capacity on Demand contract before you
order this feature.
Note: To renew this feature after the allowed 360 processor days have
been used, this feature must be removed from the system configuration
file and reordered by placing an MES order.
AIX
Linux
IBM i
Feature
code
Description OS
support

Table 1-6 summarizes the processor feature codes for the Power 780.
Table 1-6 Summary of processor features for the Power 780
Feature
code
Description OS
support
#5003 0/16 core 3.92 GHz / 4.14 GHz POWER7 Turbocore processor card:
This feature has two modes. Standard mode utilizes all 16 cores at
3.92 GHz and TurboCore mode utilizes eight cores at 4.14 GHz. This
feature is a POWER7 CUoD processor planar containing two 8-core
processors. TurboCore mode utilizes cores one through eight with
enhanced memory caching. TurboCore mode must be turned off when
you want to utilize more than eight cores. Switching between modes
requires a system reboot.
AIX
IBM i
Linux
#5333 1-core activation for processor feature #5003:
core on Processor Card #5003.
AIX
IBM i
Linux
#EP2L 100 on/off processor days of CoD billing for processor #5003:
After the ON/OFF Processor function is enabled in a system, you must
report your on/off usage to IBM at least monthly. This information, used
to compute your billing data, is provided to your sales channel. The sales
channel will place an order on your behalf for the quantity of this feature
that matches your reported use. One #EP2L provides 100 days of on/off
processor billing for POWER7 CoD Processor Book #5003 for AIX/Linux.
AIX
Linux
#EP2M 100 on/off processor days of CoD billing for processor #5003:
that matches your reported use. One #EP2M provides 100 days of on/off
processor billing for POWER7 CoD Processor Book #5003 for IBM i.
IBM i
#5342 One processor day on/off billing for #5003:
sales channel will place an order for a quantity of On/ Off Processor Core
AIX
Linux
#5343 One processor day on/off billing for #5003:
IBM i

#EP24 0/24 core 3.44 GHz POWER7 processor card:
24-core 3.44 GHz POWER7 CUoD processor planar containing four
6-core processors. Each processor has 2 MB of L2 cache (256 KB per
core) and 32 MB of L3 cache (4 MB per core). There are 16 DDR3 DIMM
slots on the processor planar (eight DIMM slots per processor), which
can be used as CoD memory without activating the processors. The
voltage regulators are included in this feature code.
AIX
IBM i
Linux
#EP25 1-core activation for processor feature #EP24:
core on Processor Card #EP24.
AIX
Linux
IBM i
#EP2N 100 on/off processor days of CoD billing for processor #EP24:
that matches your reported use. One #EP2N provides 100 days of on/off
processor billing for POWER7 CoD Processor Book #EP24 for
AIX/Linux.
AIX
Linux
#EP2P 100 on/off processor days of CoD billing for processor #EP24:
that matches your reported use. One #EP2P provides 100 days of on/off
processor billing for POWER7 CoD Processor Book #EP24 for IBM i.
IBM i
#EP28 One processor day on/off billing for #EP24:
Day Billing features and the client will be charged. One #EP27 must be
ordered for each billable processor core day of feature #EP24 used by a
AIX
Linux
#EP29 One processor day on/off billing for #EP24:
Day Billing features and the client will be charged. One #EP29 must be
ordered for each billable processor core day of feature #EP24 used by a
IBM i
#7951 On/Off Processor Enablement:
This feature can be ordered to enable your server for On/Off Capacity on
Demand. After it is enabled, you can request processors on a temporary
basis. You must sign an On/Off Capacity on Demand contract before you
order this feature.
Note: To renew this feature after the allowed 360 processor days have
been used, this feature must be removed from the system configuration
file and reordered by placing an MES order.
AIX
Linux
IBM i
Feature
code
Description OS
support

1.4.7 Memory features
In POWER7 systems, DDR3 memory is used throughout. The POWER7 DDR3 memory uses
a memory architecture to provide greater bandwidth and capacity. This enables operating at a
higher data rate for large memory configurations. All processor cards have 16 memory DIMM
slots (eight per processor) running at speeds up to 1066 MHz and must be populated with
POWER7 DDR3 Memory DIMMs.
Figure 1-5 outlines the general connectivity of an 8-core POWER7 processor and DDR3
memory DIMMS. The eight memory channels (four per memory controller) can be clearly
seen.
Figure 1-5 Outline of 8-core POWER7 processor connectivity to DDR3 DIMMs - Used in a 2-socket
Power 770 and Power 780
On each processor card for the Power 770 and Power 780 there is a total of 16 DDR3
memory DIMM slots to be connected. When using two SCMs per card, eight DIMM slots are
used per processor, and when using four SCMs per card in the Power 780 server, four DIMM
slots are used per processor.
The quad-high (96 mm) DIMM cards can have an 8 GB, 16 GB, 32 GB, or 64 GB capacity
and are connected to the POWER7 processor memory controller through an advanced
memory buffer ASIC. For each DIMM, there is a corresponding memory buffer. Each memory
channel into the POWER7 memory controllers is driven at 6.4 GHz.
Each DIMM (except the 64 GB DIMM) contains DDR3 x8 DRAMs in a configuration, with 10
DRAMs per rank, and plugs into a 276-pin DIMM slot connector. The 64 GB DIMM is an
8-rank DIMM using x4 parts (1024Kx4). The x4 DIMMs are 20 DRAMs per rank.
POWER7 Processor Chip
Memory Controller 0 Memory Controller 1
L3 Cache
Core Core Core Core
Core Core Core Core
AdvancedBuffer
DIMM5
A B C D D C B A
AdvancedBuffer
DIMM8
AdvancedBuffer
DIMM7
AdvancedBuffer
DIMM6
AdvancedBuffer
DIMM4
AdvancedBuffer
DIMM3
AdvancedBuffer
DIMM2
AdvancedBuffer
DIMM1

The Power 770 and Power 780 have memory features in 32 GB, 64 GB, 128 GB, and
256 GB capacities. Table 1-7 summarizes the capacities of the memory features and
highlights other characteristics.
Table 1-7 Summary of memory features
None of the memory in these features is active. Feature number #8212 or #8213 must be
purchased to activate the memory. Table 1-8 outlines the memory activation feature codes
and corresponding memory capacity activations.
Table 1-8 CoD system memory activation features
Note: DDR2 DIMMs (used in POWER6®-based systems) are not supported in
POWER7-based systems.
Feature
code
Memory
technology
Capacity Access
rate
DIMMs DIMM slots
used
#5600 DDR3 32 GB 1066 MHz 4 x 8 GB DIMMs 4
#5564 DDR3 256 GB 1066 MHz 4 x 64 GB DIMM 4
Feature
code
Activation
capacity
Additional information OS
support
#8212 1 GB Activation of 1 GB of DDR3 POWER7 memory. Each
occurrence of this feature permanently activates 1 GB
of DDR3 POWER7 memory.
AIX
IBM i
Linux
#8213 100 GB Activation of 100 GB of DDR3 POWER7 memory.
Each occurrence of this feature permanently activate
100 GB of DDR3 POWER7 memory.
AIX
IBM i
Linux
#7954 N/A On/Off Memory Enablement: This feature can be
ordered to enable your server for On/Off Capacity on
Demand. After it is enabled, you can request memory
on a temporary basis. You must sign an On/Off
Capacity on Demand contract before this feature is
ordered. To renew this feature after the allowed
999 GB Days have been used, this feature must be
removed from the system configuration file and
reordered by placing an MES order.
AIX
IBM i
Linux
#4710 N/A On/Off 999 GB-Days, Memory Billing POWER7:
After the ON/OFF Memory function is enabled in a
system, you must report your on/off usage to IBM at
least monthly. This information, used to compute your
billing data, is provided to your sales channel. The
sales channel will place an order on your behalf for the
quantity of this feature that matches your reported
use. One #4710 feature must be ordered for each 999
billable days for each 1 GB increment of POWER7
memory that was used.
AIX
IBM i
Linux

1.5 Disk and media features
Each system building block features two SAS DASD controllers with six hot-swappable
2.5-inch Small Form Factor (SFF) disk bays and one hot-plug, slim-line media bay per
enclosure. The SFF SAS disk drives and Solid State Drive (SSD) are supported internally. In
a full configuration with four connected building blocks, the combined system supports up to
24 disk bays. SAS drives and SSD drives can share the same backplane.
Table 1-9 shows the available disk drive feature codes that each bay can contain.
Table 1-9 Disk drive feature code description
#7377 N/A On/Off, 1 GB-1Day, Memory Billing POWER7:
After the ON/OFF Memory function is enabled in a
system you must report the client’s on/off usage to
IBM on a monthly basis. This information is used to
compute IBM billing data. One #7377 feature must be
ordered for each billable day for each 1 GB increment
of POWER7 memory that was used.
Note that inactive memory must be available in the
system for temporary use.
AIX
IBM i
Linux
Note:
All POWER7 memory features must be purchased with sufficient permanent memory activation
features so that the system memory is at least 50% active.
The minimum activations ordered with all initial orders of memory feature #5564 must be
192 GB of 256 GB per each feature #5564 ordered (that is, 75% of the installed feature
#5564 capacity).
Note: Memory CoD activations activate memory hardware only for the system serial
number for which they are purchased. If memory hardware is moved to another system,
the memory might not be functional in that system until arrangements are made to move
the memory activations or purchase additional memory activations.
Feature
code
Activation
capacity
Additional information OS
support
Feature code Description OS support
#1886 146 GB 15 K RPM SFF SAS Disk Drive AIX, Linux
#1917 146 GB 15 K RPM SAS SFF-2 Disk Drive AIX, Linux
#1995 177 GB SSD Module with eMLC AIX, Linux
#1775 177 GB SFF-1 SSD with eMLC AIX, Linux
#1793 177 GB SFF-2 SSD with eMLC AIX, Linux
#1885 300 GB 10 K RPM SFF SAS Disk Drive AIX, Linux
#1880 300 GB 15 K RPM SAS SFF Disk Drive AIX, Linux

Certain adapters are available for order in large quantities. Table 1-10 lists the disk drives
available in a quantity of 150.
Table 1-10 Available disk drives in quantity of 150
#1790 600 GB 10 K RPM SAS SFF Disk Drive AIX, Linux
#1947 139 GB 15 K RPM SAS SFF-2 Disk Drive IBM i
#1888 139 GB 15 K RPM SFF SAS Disk Drive IBM i
#1996 177 GB SSD Module with eMLC IBM i
#1787 177 GB SFF-1 SSD with eMLC IBM i
#1794 177 GB SFF-2 SSD with eMLC IBM i
#1911 283 GB 10 K RPM SFF SAS Disk Drive IBM i
#1879 283 GB 15 K RPM SAS SFF Disk Drive IBM i
#1916 571 GB 10 K RPM SAS SFF Disk Drive IBM i
#7550 Quantity 150 of #1790 (600 GB 10 K RPM SAS SFF Disk Drive) AIX, Linux
#1928 Quantity 150 of #1880 (300 GB 15 K RPM SAS SFF Disk Drive) AIX, Linux
#7547 Quantity 150 of #1885 (300 GB 10 K RPM SFF SAS Disk Drive) AIX, Linux
#7548 Quantity 150 of #1886 (146 GB 15 K RPM SFF SAS Disk Drive) AIX, Linux
#1866 Quantity 150 of #1917 (146 GB 15 K RPM SAS SFF-2 Disk Drive) AIX, Linux
#1926 Quantity 150 of #1879 (283 GB 15 K RPM SAS SFF Disk Drive) IBM i
#7544 Quantity 150 of #1888 (139 GB 15 K RPM SFF SAS Disk Drive) IBM i
#7557 Quantity 150 of #1911(283 GB 10 K RPM SFF SAS Disk Drive) IBM i
#7566 Quantity 150 of #1916 (571 GB 10 K RPM SAS SFF Disk Drive) IBM i
#1868 Quantity 150 of #1947 (139 GB 15 K RPM SAS SFF-2 Disk Drive) IBM i

The Power 770 and Power 780 support both 2.5-inch and 3.5-inch SAS SFF hard disks. The
3.5-inch DASD hard disk can be attached to the Power 770 and Power 780 but must be
located in a feature #5886 EXP12S I/O drawer, whereas 2.5-inch DASD hard files can be
either mounted internally or in the EXP24S SFF Gen2-bay Drawer (#5887).
If you need more disks than available with the internal disk bays, you can attach additional
external disk subsystems. For more detailed information about the available external disk
subsystems, see 2.11, “External disk subsystems” on page 92.
SCSI disks are not supported in the Power 770 and 780 disk bays. However, if you want to
use SCSI disks, you can attach existing SCSI disk subsystems.
The disk/media backplane feature #5652 provides six SFF disk slots and one SATA media
slot. In a full configuration with four connected building blocks, the combined system supports
up to four media devices with Media Enclosure and Backplane #5652. The SATA Slimline
DVD-RAM drive (#5762) is the only supported media device option.
1.6 I/O drawers
The system has eight I/O expansion slots per enclosure, including two dedicated GX++ slots.
If more PCI slots are needed, such as to extend the number of LPARs, up to 32 PCI-DDR
12X Expansion Drawers (#5796) and up to 16 12X I/O Drawer PCIe (#5802 and #5877) can
be attached.
The Power 770 and the Power 780 servers support the following 12X attached I/O drawers,
providing extensive capability to expand the overall server expandability and connectivity:
Feature #5802 provides PCIe slots and SFF SAS disk slots.
Feature #5877 provides PCIe slots.
Feature #5796 provides PCI-X slots.
The 7314-G30 drawer provides PCI-X slots (supported, but no longer orderable).
Disk-only I/O drawers are also supported, providing large storage capacity and multiple
partition support:
Feature #5886 EXP 12S holds a 3.5-inch SAS disk or SSD.
Feature #5887 EXP 24S SFF Gen2-bay Drawer for high-density storage holds SAS Hard
Disk drives.
The 7031-D24 holds a 3.5-inch SCSI disk (supported but no longer orderable).
The 7031-T24 holds a 3.5-inch SCSI disk (supported but no longer orderable).
1.6.1 PCI-DDR 12X Expansion Drawers (#5796)
The PCI-DDR 12X Expansion Drawer (#5796) is a 4U tall (EIA units) drawer and mounts in a
19-inch rack. Feature #5796 takes up half the width of the 4U (EIA units) rack space. Feature
#5796 requires the use of a #7314 drawer mounting enclosure. The 4U vertical enclosure can
hold up to two #5796 drawers mounted side by side in the enclosure. A maximum of four
#5796 drawers can be placed on the same 12X loop.

The I/O drawer has the following attributes:
A 4U (EIA units) rack-mount enclosure (#7314) holding one or two #5796 drawers
Six PCI-X DDR slots: 64-bit, 3.3 V, 266 MHz (blind-swap)
Redundant hot-swappable power and cooling units
1.6.2 12X I/O Drawer PCIe (#5802 and #5877)
The #5802 and #5877 expansion units are 19-inch, rack-mountable, I/O expansion drawers
that are designed to be attached to the system using 12X double data rate (DDR) cables. The
expansion units can accommodate 10 generation 3 cassettes. These cassettes can be
installed and removed without removing the drawer from the rack.
A maximum of two #5802 drawers can be placed on the same 12X loop. Feature #5877 is the
same as #5802, except it does not support disk bays. Feature #5877 can be on the same
loop as #5802. Feature #5877 cannot be upgraded to #5802.
The I/O drawer has the following attributes:
Eighteen SAS hot-swap SFF disk bays (only #5802)
Ten PCI Express (PCIe) based I/O adapter slots (blind-swap)
Redundant hot-swappable power and cooling units
1.6.3 EXP 12S SAS Drawer
The EXP 12S SAS drawer (#5886) is a 2 EIA drawer and mounts in a 19-inch rack. The
drawer can hold either SAS disk drives or SSD. The EXP 12S SAS drawer has twelve
3.5-inch SAS disk bays with redundant data paths to each bay. The SAS disk drives or SSDs
contained in the EXP 12S are controlled by one or two PCIe or PCI-X SAS adapters
connected to the EXP 12S via SAS cables.
1.6.4 EXP 24S SFF Gen2-bay Drawer
The EXP24S SFF Gen2-bay Drawer is an expansion drawer supporting up to twenty-four
2.5-inch hot-swap SFF SAS HDDs on POWER6 or POWER7 servers in 2U of 19-inch rack
space. The EXP24S bays are controlled by SAS adapters/controllers attached to the I/O
drawer by SAS X or Y cables.
The SFF bays of the EXP24S are different from the SFF bays of the POWER7 system units
or 12X PCIe I/O drawers (#5802 and #5803). The EXP24S uses Gen2 or SFF-2 SAS drives
that physically do not fit in the Gen1 or SFF-1 bays of the POWER7 system unit or 12X PCIe
I/O Drawers, or vice versa.
1.6.5 I/O drawers and usable PCI slot
The I/O drawer model types can be intermixed on a single server within the appropriate I/O
loop. Depending on the system configuration, the maximum number of I/O drawers that is
supported differs.
Note: Mixing #5802 or 5877 and #5796 on the same loop is not supported.

Table 1-11 summarizes the maximum number of I/O drawers supported and the total number
of PCI slots available when expansion consists of a single drawer type.
Table 1-11 Maximum number of I/O drawers supported and total number of PCI slots
Table 1-12 summarizes the maximum number of disk-only I/O drawers supported.
Table 1-12 Maximum number of disk only I/O drawers supported
1.7 Comparison between models
The Power 770 offers configuration options, where the POWER7 processor can have one of
two different processor cards installed. Both contain two Single Chip Module (SCM) cards.
These cards can contain the following processor configurations:
Two socket card: Eight cores at 3.3 GHz
Two socket card: Six cores at 3.72 GHz.
Both of these Power 770 models are available starting as low as four active cores and
incrementing one core at a time through built-in CoD functions to a maximum of 48 active
cores with the 3.72 GHz processor or 64 active cores with the 3.3 GHz processor.
The Power 780 offers configuration options where the POWER7 processor can have one of
two processor cards installed. These are either a two-socket SCM card or four SCM cards.
These processor cards contain the following processor configurations:
Two socket card: Eight cores at 3.92 GHz (MaxCore mode) or four cores at 4.14 GHz
(TurboCore mode)
Four socket card: Six cores at 3.44 GHz
Both of these Power 780 models are available starting as low as four active cores and
incrementing one core at a time through built-in CoD functions to a maximum of 64 active
cores with the 3.92 GHz processor or 96 active cores with the 3.44 GHz processor.
System drawers Max #5796
drawers
Max #5802 and
#5877 drawers
Total number of slots
#5796 #5802 and #5877
PCI-X PCIe PCI-X PCIe
1 drawer 8 4 48 6 0 46
2 drawers 16 8 96 12 0 92
3 drawers 24 12 144 18 0 138
4 drawers 32 16 192 24 0 184
Server Max #5886 drawers Max #5887 drawers
Power 770 110 56
Power 780 110 56

Table 1-13 summarizes the processor core options and frequencies and matches them to the
L3 cache sizes for the Power 770 and Power 780.
Table 1-13 Summary of processor core counts, core frequencies, and L3 cache sizes
1.8 Build to Order
You can perform a Build to Order (also called a la carte) configuration using the IBM
Configurator for e-business (e-config), where you specify each configuration feature that you
want on the system.
This is the only configuration method for the IBM Power 770 and Power 780 servers.
1.9 IBM Editions
IBM Edition offerings are not available for the IBM Power 770 and Power 780 servers.
1.10 Model upgrade
You can upgrade the 9117-MMA with IBM POWER6 or POWER6+™ processors to the IBM
Power 770 and Power 780 with POWER7 processors. For upgrades from POWER6 or
POWER6+ processor-based systems, IBM will install new CEC enclosures to replace the
enclosures that you currently have. Your current CEC enclosures will be returned to IBM in
exchange for the financial consideration identified under the applicable feature conversions
for each upgrade.
Clients taking advantage of the model upgrade offer from a POWER6 or POWER6+
processor-based system are required to return all components of the serialized MT-model
that were not ordered through feature codes. Any feature for which a feature conversion is
System Cores per
POWER7 SCM
Frequency
(GHz)
L3 cachea
a. The total L3 cache available on the POWER7 SCM, maintaining 4 MB per processor core.
Enclosure
summationb
b. The total number of processor cores and L3 cache within a populated enclosure.
System maximum
(cores)c
c. The maximum number of cores with four CEC enclosures and all cores activated.
Power 770 6 3.72 24 MB 12-cores and
48 MB L3 cache
48
64 MB L3 cache
64
96 MB L3 cache
96
Power 780 in MaxCore
moded
d. MaxCore mode applies to Power 780 only. Each POWER7 SCM has eight active cores and 32 MB L3 cache.
8 3.92 32 MB 16-cores and
64 MB L3 cache
64
Power 780 in TurboCore
modee
e. TurboCore mode applies to Power 780 only. Each POWER SCM uses four of the eight cores but at a higher
frequency and 32 MB L3 cache.
4 activated 4.14 32 MB 8-cores active and
64 MB L3 cache
32

used to obtain a new part must be returned to IBM also. Clients can keep and reuse any
features from the CEC enclosures that were not involved in a feature conversion transaction.
Upgrade considerations
Feature conversions have been set up for:
POWER6 and POWER6+ processors to POWER7 processors
DDR2 memory DIMMS to DDR3 memory DIMMS
Trim kits (A new trim kit is needed when upgrading to a 2-door, 3-door, or 4-door system.)
Enterprise enablement
The following features that are present on the current system can be moved to the
new system:
DDR3 memory DIMMs (#5600, #5601, and #5602)
Active Memory Expansion Enablement (#4791)
FSP/Clock Pass Through Card (#5665)
Service Processor (#5664) o 175 MB Cache RAID - Dual IOA Enablement Card (#5662)
Operator Panel (#1853)
Disk/Media Backplane (#5652)
PCIe adapters with cables, line cords, keyboards, and displays
PowerVM Standard edition (#7942) or PowerVm Enterprise edition (#7995)
I/O drawers (#5786, #5796, #5802, #5877, and #5886)
Racks (#0551, #0553, and #0555)
Doors (#6068 and #6069)
SATA DVD-RAM (#5762)
The Power 770 and Power 780 can support the following drawers:
#5802 and #5877 PCIe 12X I/O drawers
#5797 and #7413-G30 PCI-X (12X) I/O Drawer
#5786 and #7031-D24 TotalStorage EXP24 SCSI Disk Drawer
#5886 EXP12S SAS Disk Drawer
The Power 770 and Power 780 support only the SAS DASD SFF hard disks internally. The
existing 3.5-inch DASD hard disks can be attached to Power 770 and Power 780, but must
be located in an I/O drawer such as #5886.
For POWER6 or POWER6+ processor-based systems that have the On/Off CoD function
enabled, you must reorder the On/Off enablement features (#7951and #7954) when placing
the upgrade MES order for the new Power 770 or 780 system to keep the On/Off CoD
function active. To initiate the model upgrade, the On/Off enablement features should be
removed from the configuration file before the Miscellaneous Equipment Shipment (MES)
order is started. Any temporary use of processors or memory owed to IBM on the existing
system must be paid before installing the new Power 770 model MMC or Power 780
model MHC.
Features #8018 and #8030 are available to support migration of the PowerVM features #7942
or #7995 during the initial order and build of the MMC or MHC upgrade MES order.
Customers can add feature #8018 or #8030 to their upgrade orders in a quantity not to
exceed the quantity of feature #7942 or #7995 obtained for the system being upgraded.
Feature #7942 or #7995 must be migrated to the new configuration report in a quantity that
equals feature #8018 or #8030. Additional #7942 or #7995 features can be ordered during
the upgrade.
Clients can add feature #8018 to their upgrade orders in a quantity not to exceed the quantity
of feature #7942 obtained for the system being upgraded. Feature #7942 must be migrated to

the new configuration report in a quantity that equals feature #8018. Additional #7942
features can be ordered during the upgrade.
1.11 Hardware Management Console models
The Hardware Management Console (HMC) is required for managing the IBM
Power 770 and Power 780. It provides a set of functions that are necessary to manage
the system, including:
Creating and maintaining a multiple partition environment
Displaying a virtual operating system session terminal for each partition
Displaying a virtual operator panel of contents for each partition
Detecting, reporting, and storing changes in hardware conditions
Powering managed systems on and off
Acting as a service focal point for service representatives to determine an appropriate
service strategy
The IBM Power 770 and Power 780 are not supported by the Integrated Virtualization
Manager (IVM).
Several HMC models are supported to manage POWER7-based systems. Two models
(7042-C08 and 7042-CR6) are available for ordering at the time of writing, but you can also
use one of the withdrawn models listed in Table 1-14.
Table 1-14 HMC models supporting POWER7 processor technology-based servers
At the time of writing, base Licensed Machine Code Version 7 Revision 7.4.0 or later is
required to support the Power 770 and Power 780.
Existing HMC models 7310 can be upgraded to Licensed Machine Code Version 7 to support
environments that might include POWER5, POWER5+, POWER6, POWER6+, and
POWER7 processor-based servers. Licensed Machine Code Version 6 (#0961) is not
available for 7042 HMCs.
If you want to support more than 254 partitions in total, then the HMC might require a memory
upgrade to 4 GB.
Type-model Availability Description
7310-C05 Withdrawn IBM 7310 Model C05 Desktop Hardware Management Console
7310-C06 Withdrawn IBM 7310 Model C06 Deskside Hardware Management Console
7042-C08 Available IBM 7042 Model C08 Deskside Hardware Management Console
7310-CR3 Withdrawn IBM 7310 Model CR3 Rack-Mounted Hardware Management Console
7042-CR6 Available IBM 7042 Model CR6 Rack mounted Hardware Management Console

1.12 System racks
The Power 770 and its I/O drawers are designed to be mounted in the 7014-T00,
7014-T42, 7014-B42, 7014-S25, #0551, #0553, or #0555 rack. The Power 780 and I/O
drawers can be ordered only with the 7014-T00 and 7014-T42 racks. These are built to the
19-inch EIA standard. An existing 7014-T00, 7014-B42, 7014-S25, 7014-T42, #0551, #0553,
or #0555 rack can be used for the Power 770 and Power 780 if sufficient space and power
are available.
The 36U (1.8-meter) rack (#0551) and the 42U (2.0-meter) rack (#0553) are available for
order on MES upgrade orders only. For initial system orders, the racks must be ordered as
machine type 7014, models T00, B42, S25, or T42.
If a system is to be installed in a rack or cabinet that is not IBM, it must meet requirements.
1.12.1 IBM 7014 model T00 rack
The 1.8-meter (71-in.) model T00 is compatible with past and present IBM Power systems.
The features of the T00 rack are as follows:
It has 36U (EIA units) of usable space.
It has optional removable side panels.
It has an optional highly perforated front door.
It has optional side-to-side mounting hardware for joining multiple racks.
It has standard business black or optional white color in OEM format.
It has increased power distribution and weight capacity.
It supports both AC and DC configurations.
The rack height is increased to 1926 mm (75.8 in.) if a power distribution panel is fixed to
the top of the rack.
Up to four power distribution units (PDUs) can be mounted in the PDU bays (Figure 1-6 on
page 31), but others can fit inside the rack. See 1.12.7, “The AC power distribution unit
and rack content” on page 31.
Weights are:
– T00 base empty rack: 244 kg (535 lb)
– T00 full rack: 816 kg (1795 lb)
Note: The client is responsible for ensuring that the installation of the drawer in the
preferred rack or cabinet results in a configuration that is stable, serviceable, safe, and
compatible with the drawer requirements for power, cooling, cable management, weight,
and rail security.

1.12.2 IBM 7014 model T42 rack
The 2.0-meter (79.3-inch) Model T42 addresses the client requirement for a tall enclosure to
house the maximum amount of equipment in the smallest possible floor space. The features
that differ in the model T42 rack from the model T00 include:
It has 42U (EIA units) of usable space (6U of additional space).
The model T42 supports AC only.
Weights are:
– T42 base empty rack: 261 kg (575 lb)
– T42 full rack: 930 kg (2045 lb)
1.12.3 IBM 7014 model S25 rack
The 1.3-meter (49-inch) model S25 rack has the following features:
25U (EIA units)
Weights:
– Base empty rack: 100.2 kg (221 lb)
– Maximum load limit: 567.5 kg (1250 lb)
The S25 racks do not have vertical mounting space that accommodate feature number 7188
PDUs. All PDUs required for application in these racks must be installed horizontally in the
rear of the rack. Each horizontally mounted PDU occupies 1U of space in the rack, and
therefore reduces the space available for mounting servers and other components.
1.12.4 Feature number 0555 rack
The 1.3-meter rack (#0555) is a 25U (EIA units) rack. The rack that is delivered as #0555 is
the same rack delivered when you order the 7014-S25 rack. The included features might
differ. The #0555 is supported, but it is no longer orderable.
the same rack delivered when you order the 7014-T00 rack. The included features might
differ. Several features that are delivered as part of the 7014-T00 must be ordered separately
with the #0551.
the same rack delivered when you order the 7014-T42 or B42 rack. The included features
might differ. Several features that are delivered as part of the 7014-T42 or B42 must be
ordered separately with the #0553.
Note: A special door (#6250) and side panels (#6238) are available to make the rack
appear as a high-end server (but in a 19-inch rack format instead of a 24-inch rack).
Note: The Power 780 cannot be ordered with a S25 or B25 rack.

1.12.7 The AC power distribution unit and rack content
For rack models T00 and T42, 12-outlet PDUs are available. These include PDUs Universal
UTG0247 Connector (#9188 and #7188) and Intelligent PDU+ Universal UTG0247
Connector (#7109).
Four PDUs can be mounted vertically in the back of the T00 and T42 racks. Figure 1-6 shows
the placement of the four vertically mounted PDUs. In the rear of the rack, two additional
PDUs can be installed horizontally in the T00 rack and three in the T42 rack. The four vertical
mounting locations will be filled first in the T00 and T42 racks. Mounting PDUs horizontally
consumes 1U per PDU and reduces the space available for other racked components. When
mounting PDUs horizontally, use fillers in the EIA units occupied by these PDUs to facilitate
proper air flow and ventilation in the rack.
Figure 1-6 PDU placement and PDU view
For the Power 770 and Power 780 installed in IBM 7014 or #055x racks, the following PDU
rules apply:
For PDU #7188 and #7109 when using power cord #6654, #6655, #6656, #6657,
or #6658: Each pair of PDUs can power up to two Power 770 and Power 780
CEC enclosures.
For PDU #7188 and #7109 when using power cord #6489, 6491, #6492, or #6653: Each
pair of PDUs can power up to 4-5 Power 770 and Power 780 CEC enclosures.
For detailed power cord requirements and power cord feature codes, see the IBM Power
Systems Hardware Information Center website:
http://publib.boulder.ibm.com/infocenter/systems/scope/hw/index.jsp
Rack Rear View
43
21
Circuit breaker reset
Status LED

The Base/Side Mount Universal PDU (#9188) and the optional, additional, Universal PDU
(#7188) and the Intelligent PDU+ options (#7109) support a wide range of country
requirements and electrical power specifications. The PDU receives power through a
UTG0247 power line connector. Each PDU requires one PDU-to-wall power cord. Various
power cord features are available for different countries and applications by varying the
PDU-to-wall power cord, which must be ordered separately. Each power cord provides the
unique design characteristics for the specific power requirements. To match new power
requirements and save previous investments, these power cords can be requested with an
initial order of the rack or with a later upgrade of the rack features.
The PDU has 12 client-usable IEC 320-C13 outlets. There are six groups of two outlets fed by
six circuit breakers. Each outlet is rated up to 10 amps, but each group of two outlets is fed
from one 15 amp circuit breaker.
The Universal PDUs are compatible with previous models.
1.12.8 Rack-mounting rules
The system consists of one to four CEC enclosures. Each enclosure occupies 4U of
vertical rack space. The primary considerations to account for when mounting the system into
a rack are:
For configurations with two, three, or four drawers, all drawers must be installed together in
the same rack, in a contiguous space of 8 U, 12 U, or 16 U within the rack. The uppermost
enclosure in the system is the base enclosure. This enclosure will contain the active
service processor and the operator panel. If a second CEC enclosure is part of the
system, the backup service processor is contained in the second CEC enclosure.
The 7014-T42, -B42, or #0553 rack is constructed with a small flange at the bottom of EIA
location 37. When a system is installed near the top of a 7014-T42, -B42, or #0553 rack,
no system drawer can be installed in EIA positions 34, 35, or 36. This approach is to avoid
interference with the front bezel or with the front flex cable, depending on the system
configuration. A two-drawer system cannot be installed above position 29. A three-drawer
system cannot be installed above position 25. A four-drawer system cannot be installed
above position 21. (The position number refers to the bottom of the lowest drawer.)
When a system is installed in an 7014-T00, -T42, -B42, #0551, or #0553 rack that has no
front door, a Thin Profile Front Trim Kit must be ordered for the rack. The required trim kit
for the 7014-T00 or #0551 rack is #6263. The required trim kit for the 7014-T42, -B42, or
#0553 rack is #6272. When upgrading from a 9117-MMA, trim kits #6263 or #6272 can be
used for one drawer enclosures only.
Note: Ensure that the appropriate power cord feature is configured to support the power
being supplied.
Notes: Based on the power cord that is used, the PDU can supply from 4.8 - 19.2 kVA.
The total kilovolt ampere (kVA) of all the drawers that are plugged into the PDU must not
exceed the power cord limitation.
Each system drawer to be mounted in the rack requires two power cords, which are not
included in the base order. For maximum availability, be sure to connect power cords from
the same system to two separate PDUs in the rack, and to connect each PDU to
independent power sources.

The design of the Power 770 and Power 780 is optimized for use in a 7014-T00, -T42,
-B42, -S25, #0551, or #0553 rack. Both the front cover and the processor flex cables
occupy space on the front left side of an IBM 7014, #0551, and #0553 rack that might not
be available in typical non-IBM racks.
Acoustic door features are available with the 7014-T00, 7014-B42, 7014-T42, #0551,
and #0553 racks to meet the lower acoustic levels identified in the specification section
of this document. The acoustic door feature can be ordered on new T00, B42, T42, #0551,
and #0553 racks or ordered for the T00, B42, T42, #0551, and #0553 racks that you
already own.
1.12.9 Useful rack additions
This section highlights several available solutions for IBM Power Systems
rack-based systems.
IBM 7214 Model 1U2 SAS Storage Enclosure
The IBM System Storage 7214 Tape and DVD Enclosure Express is designed to mount in
one EIA unit of a standard IBM Power Systems 19-inch rack and can be configured with
one or two tape drives, or either one or two Slim DVD-RAM or DVD-ROM drives in the
right-side bay.
The two bays of the 7214 Express can accommodate the following tape or DVD drives for IBM
Power servers:
DAT72 36 GB Tape Drive: Up to two drives
Half-high LTO Ultrium 4 800 GB Tape Drive: Up to two drives
DVD-RAM Optical Drive: Up to two drives
DVD-ROM Optical Drive: Up to two drives
IBM System Storage 7214 Tape and DVD Enclosure
The IBM System Storage 7214 Tape and DVD Enclosure is designed to mount in one EIA unit
of a standard IBM Power Systems 19-inch rack and can be configured with one or two tape
drives, or either one or two Slim DVD-RAM or DVD-ROM drives in the right-side bay.
The two bays of the IBM System Storage 7214 Tape and DVD Enclosure can accommodate
the following tape or DVD drives for IBM Power servers:
Half-high LTO Ultrium 4 800 GB Tape Drive: Up to two drives
DVD-RAM Optical Drive: Up to two drives
DVD-ROM Optical Drive: Up to two drives
IBM System Storage 7216 Multi-Media Enclosure
The IBM System Storage 7216 Multi-Media Enclosure (Model 1U2) is designed to attach to
the Power 770 and the Power 780 through a USB port on the server or through a PCIe SAS
adapter. The 7216 has two bays to accommodate external tape, removable disk drive, or
DVD-RAM drive options.

The following optional drive technologies are available for the 7216-1U2:
DAT160 80 GB SAS Tape Drive (#5619
DAT320 160 GB SAS Tape Drive (#1402)
DAT320 160 GB USB Tape Drive (#5673)
Half-high LTO Ultrium 5 1.5 TB SAS Tape Drive (#8247)
DVD-RAM - 9.4 GB SAS Slim Optical Drive (#1420 and #1422)
RDX Removable Disk Drive Docking Station (#1103)
To attach a 7216 Multi-Media Enclosure to the Power 770 and Power 780, consider the
following cabling procedures:
Attachment by an SAS adapter
A PCIe Dual-X4 SAS adapter (#5901) or a PCIe LP 2-x4-port SAS Adapter 3 Gb (#5278)
must be installed in the Power 770 and Power 780 server to attach to a 7216 Model 1U2
Multi-Media Storage Enclosure. Attaching a 7216 to a Power 770 and Power 780 through
the integrated SAS adapter is not supported.
For each SAS tape drive and DVD-RAM drive feature installed in the 7216, the appropriate
external SAS cable will be included.
An optional Quad External SAS cable is available by specifying (#5544) with each 7216
order. The Quad External Cable allows up to four 7216 SAS tape or DVD-RAM features to
attach to a single System SAS adapter.
Up to two 7216 storage enclosure SAS features can be attached per PCIe Dual-X4 SAS
adapter (#5901) or the PCIe LP 2-x4-port SAS Adapter 3 Gb (#5278).
Attachment by a USB adapter
The Removable RDX HDD Docking Station features on 7216 only support the USB cable
that is provided as part of the feature code. Additional USB hubs, add-on USB cables, or
USB cable extenders are not supported.
For each RDX Docking Station feature installed in the 7216, the appropriate external USB
cable will be included. The 7216 RDX Docking Station feature can be connected to the
external, integrated USB ports on the Power 770 and Power 780 or to the USB ports on
4-Port USB PCI Express Adapter (# 2728).
The 7216 DAT320 USB tape drive or RDX Docking Station features can be connected to
the external, integrated USB ports on the Power 770 and Power 780.
The two drive slots of the 7216 enclosure can hold the following drive combinations:
One tape drive (DAT160 SAS or Half-high LTO Ultrium 5 SAS) with second bay empty
Two tape drives (DAT160 SAS or Half-high LTO Ultrium 5 SAS) in any combination
One tape drive (DAT160 SAS or Half-high LTO Ultrium 5 SAS) and one DVD-RAM SAS
drive sled with one or two DVD-RAM SAS drives
Up to four DVD-RAM drives
One tape drive (DAT160 SAS or Half-high LTO Ultrium 5 SAS) in one bay, and one RDX
Removable HDD Docking Station in the other drive bay
One RDX Removable HDD Docking Station and one DVD-RAM SAS drive sled with one
or two DVD-RAM SAS drives in the right bay
Two RDX Removable HDD Docking Stations
Note: The DAT320 160 GB SAS Tape Drive (#1402) and the DAT320 160 GB USB Tape
Drive (#5673) are no longer available as of July 15, 2011.

Figure 1-7 shows the 7216 Multi-Media Enclosure.
Figure 1-7 7216 Multi-Media Enclosure
In general, the 7216-1U2 is supported by the AIX, IBM i, and Linux operating system.
However, the RDX Removable Disk Drive Docking Station and the DAT320 USB Tape Drive
are not supported with IBM i.
Flat panel display options
The IBM 7316 Model TF3 is a rack-mountable flat panel console kit consisting of a 17-inch
337.9 mm x 270.3 mm flat panel color monitor, rack keyboard tray, IBM Travel Keyboard,
support for IBM keyboard/video/mouse (KVM) switches, and language support. The IBM
7316-TF3 Flat Panel Console Kit offers:
Slim, sleek, lightweight monitor design that occupies only 1U (1.75 inches) in a 19-inch
standard rack
A 17-inch, flat screen TFT monitor with truly accurate images and virtually no distortion
The ability to mount the IBM Travel Keyboard in the 7316-TF3 rack keyboard tray
Support for IBM keyboard/video/mouse (KVM) switches that provide control of as many as
128 servers, and support of both USB and PS/2 server-side keyboard and mouse
connections

Chapter 2. Architecture and technical
overview
The IBM Power 780 offers two versions of CEC enclosure. The first is a 2-socket CEC
enclosure, populated with 8-core POWER7 processor cards. This architecture (Figure 2-1 on
page 38) enables a maximum system configuration of 64 processors. The Power 780 also
offers a 4-socket CEC enclosure, populated with 6-core POWER7 processor cards
(Figure 2-2 on page 39), enabling a maximum system configuration of 96 cores.
The IBM Power 770 offers a 2-socket CEC enclosure, populated with 6-core or 8-core
POWER7 processors.
This chapter provides an overview of the system architecture and its major components. The
bandwidths that are provided are theoretical maximums used for reference.
The speeds shown are at an individual component level. Multiple components and application
implementation are key to achieving the best performance.
Always do the performance sizing at the application workload environment level and evaluate
performance using real-world performance measurements and production workloads.
2

Figure 2-1 shows the logical system diagram of the 2-socket Power 770 and Power 780.
Figure 2-1 Two-socket IBM Power 770 and Power 780 logical system diagram
P7-IOC
GX++ SLOT #2
GX++ SLOT #1
PCIe Gen2 x8 (FH/HL) SLOT #5
TPMD
SAS
Controller
Optional
RAID
Exp. Card
HDD1
HDD2
HDD3
HDD4
HDD5
HDD6
DVD
SMP
Connector
A
SMP
Connector
B
SMP
Connector
B
SMP
Connector
A
SAS
Controller
SAS
Controller
Optional
RAID
Exp. Card
2 x 10 Gbps + 2 x 1 Gbps Ethernet
P7-IOC
2.46 GHz
1.0 GHz
1.0 GHz
Buffer Buffer Buffer Buffer Buffer Buffer Buffer Buffer
DIMM#9
DIMM#10
DIMM#11
DIMM#12
DIMM#13
DIMM#14
DIMM#15
DIMM#16
Buffer
DIMM#1
DIMM#2
136.448 GBps
per socket
Buffer Buffer
DIMM#3
DIMM#4
Buffer Buffer
DIMM#5
DIMM#6
Buffer Buffer
DIMM#7
DIMM#8
Buffer
2 System Ports
2 HMC Ports
2 SPCN Ports
VPD Chip
Service
Processor
USB #1
USB #2
USB Controller
3.24GHz
3.24GHz
3.24GHz
POWER7 Chip 1
6-8 cores
Memory Controller
POWER7 Chip 2
6-8 cores
Memory Controller
136.448 GBps
per socket
2.46 GHz
2.46 GHz
2.46 GHz
2.46 GHz (2 * 4 Bytes) 19.712 GBps
2.46 GHz (2 * 4 Bytes) 19.712 GBps
2.46 GHz (2 * 4 Bytes) 19.712 GBps
2.46 GHz (2 * 4 Bytes) 19.712 GBps

Chapter 2. Architecture and technical overview 39
Figure 2-2 shows the logical system diagram of the 4-socket Power 780.
Figure 2-2 Four-socket IBM Power 780 logical system diagram
P7-IOC
GX++ SLOT #2
GX++ SLOT #1
Buffer
DIMM#1
DIMM#2
TPMD
SAS
Controller
Optional
RAID
Exp. Card
HDD1
HDD2
HDD3
HDD4
HDD5
HDD6
DVD
2.9GHz
SMP
Connector
A
SMP
Connector
B
SMP
Connector
B
SMP
Connector
A
SAS
Controller
SAS
Controller
Optional
RAID
Exp. Card
2 x 10 Gbps + 2 x 1 Gbps Ethernet
P7-IOC
Memory Controller
POWER7 Chip 3
6 cores
POWER7 Chip 4
6 cores
2.9Gbps
POWER7 Chip 1
6 cores
POWER7 Chip 2
6 cores
Memory Controller
Memory Controller
2.46GHz2.46GHz
2.46 GHz
2.46 GHz
1.0 GHz
1.0 GHz
2.9
G
Hz
2.9
G
Hz
Buffer BufferDIMM#3
DIMM#4
Buffer Buffer
DIMM#5
DIMM#6
Buffer Buffer
DIMM#7
DIMM#8
Buffer
Buffer Buffer Buffer Buffer Buffer Buffer Buffer Buffer
DIMM#9
DIMM#10
DIMM#11
DIMM#12
DIMM#13
DIMM#14
DIMM#15
DIMM#162.9 GHz
Memory Controller
Memory Controller
USB #1
USB #2
USB Controller
2 System Ports
2 HMC Ports
2 SPCN Ports
VPD Chip
Service
Processor
2.46 GHz (2 * 4 Bytes) 19.712 GBps
136.448 GBps
per socket
136.448 GBps
per socket
136.448 GBps
per socket
136.448 GBps
per socket
2.46 GHz (2 * 4 Bytes) 19.712 GBps
2.46 GHz (2 * 4 Bytes) 19.712 GBps
2.46 GHz (2 * 4 Bytes) 19.712 GBps

2.1 The IBM POWER7 processor
The IBM POWER7 processor represents a leap forward in technology achievement and
associated computing capability. The multi-core architecture of the POWER7 processor has
been matched with innovation across a wide range of related technologies to deliver leading
throughput, efficiency, scalability, and RAS.
Although the processor is an important component in delivering outstanding servers, many
elements and facilities have to be balanced on a server to deliver maximum throughput. As
with previous generations of systems based on POWER processors, the design philosophy
for POWER7 processor-based systems is one of system-wide balance in which the POWER7
processor plays an important role.
In many cases, IBM has been innovative in order to achieve required levels of throughput and
bandwidth. Areas of innovation for the POWER7 processor and POWER7 processor-based
systems include (but are not limited to) these:
On-chip L3 cache implemented in embedded dynamic random access memory (eDRAM)
Cache hierarchy and component innovation
Advances in memory subsystem
Advances in off-chip signaling
Exploitation of long-term investment in coherence innovation
The superscalar POWER7 processor design also provides a variety of other capabilities:
Binary compatibility with the prior generation of POWER processors
Support for PowerVM virtualization capabilities, including PowerVM Live Partition Mobility
to and from POWER6 and POWER6+ processor-based systems

Figure 2-3 shows the POWER7 processor die layout with the major areas identified:
Processor cores
L2 cache
L3 cache and chip interconnection
Simultaneous multiprocessing (SMP) links
Memory controllers.
Figure 2-3 POWER7 processor die with key areas indicated
2.1.1 POWER7 processor overview
The POWER7 processor chip is fabricated using the IBM 45 nm Silicon-On-Insulator (SOI)
technology using copper interconnect and implements an on-chip L3 cache using eDRAM.
The POWER7 processor chip is 567 mm2
and is built using 1.2 billion components
(transistors). Eight processor cores are on the chip, each with 12 execution units, 256 KB of
L2 cache, and access to up to 32 MB of shared on-chip L3 cache.
For memory access, the POWER7 processor includes two DDR3 (double data rate 3)
memory controllers, each with four memory channels. To be able to scale effectively, the
POWER7 processor uses a combination of local and global SMP links with very high
coherency bandwidth and takes advantage of the IBM dual-scope broadcast coherence
protocol.

Table 2-1 summarizes the technology characteristics of the POWER7 processor.
Table 2-1 Summary of POWER7 processor technology
2.1.2 POWER7 processor core
Each POWER7 processor core implements aggressive out-of-order (OoO) instruction
execution to drive high efficiency in the use of available execution paths. The POWER7
processor has an Instruction Sequence Unit that is capable of dispatching up to six
instructions per cycle to a set of queues. Up to eight instructions per cycle can be issued to
the instruction execution units. The POWER7 processor has a set of 12 execution units:
Two fixed point units
Two load store units
Four double precision floating point units
One vector unit
One branch unit
One condition register unit
One decimal floating point unit
These caches are tightly coupled to each POWER7 processor core:
Instruction cache: 32 KB
Data cache: 32 KB
L2 cache: 256 KB, implemented in fast SRAM
Technology POWER7 processor
Die size 567 mm2
Fabrication technology 45 nm lithography
Copper interconnect
Silicon-on-Insulator
eDRAM
Components 1.2 billion components/transistors offering the
equivalent function of 2.7 billion (For further details see
2.1.6, “On-chip L3 cache innovation and Intelligent
Cache” on page 46.)
Processor cores 4, 6, or 8
Max execution threads core/chip 4/32
L2 cache core/chip 256 KB/2 MB
On-chip L3 cache core/chip 4 MB/32 MB
DDR3 memory controllers 1 or 2
SMP design-point 32 sockets with IBM POWER7 processors
Compatibility With prior generation of POWER processor

2.1.3 Simultaneous multithreading
An enhancement in the POWER7 processor is the addition of the SMT4 mode to enable four
instruction threads to execute simultaneously in each POWER7 processor core. Thus, these
are the instruction thread execution modes of the POWER7 processor:
SMT1: Single instruction execution thread per core
SMT2: Two instruction execution threads per core
SMT4: Four instruction execution threads per core
SMT4 mode enables the POWER7 processor to maximize the throughput of the processor
core by offering an increase in processor-core efficiency. SMT4 mode is the latest step in an
evolution of multithreading technologies introduced by IBM. Figure 2-4 shows the evolution of
simultaneous multithreading in the industry.
Figure 2-4 Evolution of simultaneous multi-threading
The various SMT modes offered by the POWER7 processor allow flexibility, enabling users to
select the threading technology that meets an aggregation of objectives such as
performance, throughput, energy use, and workload enablement.
Intelligent Threads
The POWER7 processor features Intelligent Threads that can vary based on the workload
demand. The system either automatically selects (or the system administrator can manually
select) whether a workload benefits from dedicating as much capability as possible to a
single thread of work, or if the workload benefits more from having capability spread across
two or four threads of work. With more threads, the POWER7 processor can deliver more
total capacity as more tasks are accomplished in parallel. With fewer threads, those
workloads that need very fast individual tasks can get the performance that they need for
maximum benefit.
Multi-threading evolution
Thread 1 ExecutingThread 0 Executing No Thread Executing
FX0
FX1
FP0
FP1
LS0
LS1
BRX
CRL
1995 singlethread out of order
FX0
FX1
FP0
FP1
LS0
LS1
BRX
CRL
1997 hardwaremulti-thread
FX0
FX1
FP0
FP1
LS0
LS1
BRX
CRL
2004 2-way SMT
FX0
FX1
FP0
FP1
LS0
LS1
BRX
CRL
2010 4-way SMT
Thread 3 ExecutingThread 2 Executing

2.1.4 Memory access
Each POWER7 processor chip has two DDR3 memory controllers, each with four memory
channels (enabling eight memory channels per POWER7 processor). Each channel operates
at 6.4 GHz and can address up to 32 GB of memory. Thus, each POWER7 processor chip is
capable of addressing up to 256 GB of memory.
Figure 2-5 gives a simple overview of the POWER7 processor memory access structure.
Figure 2-5 Overview of POWER7 memory access structure
2.1.5 Flexible POWER7 processor packaging and offerings
POWER7 processors have the unique ability to optimize to various workload types. For
example, database workloads typically benefit from very fast processors that handle high
transaction rates at high speeds. Web workloads typically benefit more from processors with
many threads that allow the breaking down of web requests into many parts and handle them
in parallel. POWER7 processors uniquely have the ability to provide leadership performance
in either case.
TurboCore mode
Users can opt to run selected servers in TurboCore mode. It uses four cores per POWER7
processor chip with access to the full 32 MB of L3 cache (8 MB per core) and at a faster
processor core frequency, which might save on software costs for those applications that are
licensed per core.
Note: In certain POWER7 processor-based systems, one memory controller is active with
four memory channels being used.
POWER7 Processor Chip
Advanced
Buffer ASIC
Chip
Memory
Controller
Advanced
Buffer ASIC
Chip
Memory
Controller
Core Core Core Core
Core Core Core Core
Dual Integrated DDR3 memory controllers
• High channel and DIMM utilization
• Advanced energy management
• RAS advances
Eight high-speed 6.4 GHz channels
• New low-power differential signalling
New DDR3 buffer chip architecture
• Larger capacity support (32 GB/core)
• Energy management support
• RAS enablement
DDR3 DRAMs

MaxCore mode
MaxCore mode is for workloads that benefit from a higher number of cores and threads
handling multiple tasks simultaneously that take advantage of increased parallelism.
MaxCore mode provides up to eight cores and up to 32 threads per POWER7 processor.
POWER7 processor 4-core and 6-core offerings
The base design for the POWER7 processor is an 8-core processor with 32 MB of on-chip L3
cache (4 MB per core). However, the architecture allows for differing numbers of processor
cores to be active, 4 cores or 6 cores, as well as the full 8-core version.
In most cases (MaxCore mode), the L3 cache associated with the implementation is
dependant on the number of active cores. For a 6-core version, this typically means that
6 x 4 MB (24 MB) of L3 cache is available. Similarly, for a 4-core version, the L3 cache
available is 16 MB.
Optimized for servers
The POWER7 processor forms the basis of a flexible compute platform and can be offered in
a number of guises to address differing system requirements.
The POWER7 processor can be offered with a single active memory controller with four
channels for servers where higher degrees of memory parallelism are not required.
Similarly, the POWER7 processor can be offered with a variety of SMP bus capacities that
are appropriate to the scaling-point of particular server models.
Figure 2-6 outlines the physical packaging options that are supported with POWER7
processors.
Figure 2-6 Outline of the POWER7 processor physical packaging
Note: TurboCore is available on the Power 780 and Power 795.
Note: The 4-core processor is not available on the Power 770 and Power 780.
Single Chip Organic
1 x Memory Controller
Local broadcast SMP links active
Single Chip Glass Ceramic
2 x Memory Controllers
Local broadcast SMP links active
Global broadcast SMP links active

2.1.6 On-chip L3 cache innovation and Intelligent Cache
A breakthrough in material engineering and microprocessor fabrication has enabled IBM to
implement the L3 cache in eDRAM and place it on the POWER7 processor die. L3 cache is
critical to a balanced design, as is the ability to provide good signaling between the L3 cache
and other elements of the hierarchy, such as the L2 cache or SMP interconnect.
The on-chip L3 cache is organized into separate areas with differing latency characteristics.
Each processor core is associated with a Fast Local Region of L3 cache (FLR-L3) but also
has access to other L3 cache regions as shared L3 cache. Additionally, each core can
negotiate to use the FLR-L3 cache associated with another core, depending on reference
patterns. Data can also be cloned to be stored in more than one core's FLR-L3 cache, again
depending on reference patterns. This Intelligent Cache management enables the POWER7
processor to optimize the access to L3 cache lines and minimize overall cache latencies.
Figure 2-7 shows the FLR-L3 cache regions for each of the cores on the POWER7
processor die.
Figure 2-7 Fast local regions of L3 cache on the POWER7 processor
The innovation of using eDRAM on the POWER7 processor die is significant for
several reasons:
Latency improvement
A six-to-one latency improvement occurs by moving the L3 cache on-chip compared to L3
accesses on an external (on-ceramic) ASIC.
Bandwidth improvement
A 2x bandwidth improvement occurs with on-chip interconnect. Frequency and bus sizes
are increased to and from each core.

No off-chip driver or receivers
Removing drivers or receivers from the L3 access path lowers interface requirements,
conserves energy, and lowers latency.
Small physical footprint
The performance of eDRAM when implemented on-chip is similar to conventional SRAM
but requires far less physical space. IBM on-chip eDRAM uses only a third of the
components used in conventional SRAM, which has a minimum of six transistors to
implement a 1-bit memory cell.
Low energy consumption
The on-chip eDRAM uses only 20% of the standby power of SRAM.
2.1.7 POWER7 processor and Intelligent Energy
Energy consumption is an important area of focus for the design of the POWER7 processor,
which includes Intelligent Energy features that help to dynamically optimize energy usage
and performance so that the best possible balance is maintained. Intelligent Energy features
like EnergyScale work with IBM Systems Director Active Energy Manager to dynamically
optimize processor speed based on thermal conditions and system utilization.
2.1.8 Comparison of the POWER7 and POWER6 processors
Table 2-2 shows comparable characteristics between the generations of POWER7 and
POWER6 processors.
Table 2-2 Comparison of technology for the POWER7 processor and the prior generation
POWER7 POWER6+ POWER6
Technology 45 nm 65 nm 65 nm
Die size 567 mm2
341 mm2
341 mm2
Maximum cores 8 2 2
Maximum SMT
threads per core
4 threads 2 threads 2 threads
Maximum frequency 4.25 GHz 5.0 GHz 4.7 GHz
L2 Cache 256 KB per core 4 MB per core 4 MB per core
L3 Cache 4 MB of FLR-L3 cache
per core with each core
having access to the full
32 MB of L3 cache,
on-chip eDRAM
32 MB off-chip eDRAM
ASIC
32 MB off-chip eDRAM
ASIC
Memory support DDR3 DDR2 DDR2
I/O bus Two GX++ One GX++ One GX++
Enhanced cache
mode (TurboCore)
Yesa
a. Not supported on the Power 770 and Power 780 4-socket systems.
No No
Sleep and nap modeb
Both Nap only Nap only

2.2 POWER7 processor cards
IBM Power 770 and Power 780 servers are modular systems built using one to four CEC
enclosures. The processor and memory subsystem in each CEC enclosure is contained on a
single processor card. The processor card contains either two or four processor sockets and
16 fully buffered DDR3 memory DIMMs.
The IBM Power 770 supports the 2-socket processor cards, populated with 6-core or 8-core
POWER7 processors. This enables a maximum system configuration of 64-cores, built from
four CEC enclosures.
The IBM Power 780 supports both the 2-socket and 4-socket processor cards. The 4-socket
processor cards are populated with 6-core POWER7 processors, enabling a maximum
system configuration of 96 cores.
2.2.1 Two-socket processor card
The 2-socket processor card (Figure 2-8) is supported in both the Power 770 and the
Power 780 system. Each processor is connected to eight DIMMs via two memory controllers
(four DIMMS on each).
Figure 2-8 IBM Power 770 and Power 780 2-socket processor card
b. For more information about sleep and nap modes, see 2.15.1, “IBM EnergyScale technology”
on page 114.
Note: Mixing 2-socket and 4-socket CEC enclosures within a Power 780 is not supported.
POWER7
processors
DDR3
Memory
DIMMs

Power 770 systems
IBM Power 770 systems support two POWER7 processor options of varying clock speed and
core counts. Table 2-3 summarizes these options.
Table 2-3 Summary of POWER7 processor options for the Power 770 server
With two POWER7 processors in each enclosure, systems can be equipped as follows:
Using 6-core POWER7 processors:
– 12 cores
– 24 cores
– 36 cores
– 48 cores
Using 8-core POWER7 processors:
– 16-cores
– 32-cores
– 48-cores
– 64-cores
Power 780 systems
The IBM Power 780 2-socket CEC enclosures offer POWER7 processors with 8 cores.
However, the system can be booted on one of two modes:
MaxCore mode
TurboCore mode
In MaxCore mode, all eight cores of each POWER7 processor are active, run at 3.92 GHz,
and have full access to the 32 MB of L3 cache. In TurboCore mode the system uses just four
of the POWER7 processor cores, but runs at the higher frequency of 4.14 GHz and has
access to the full 32 MB of L3 cache.
Table 2-4 summarizes the POWER7 processor and mode options for the Power 780 system.
Table 2-4 Summary of POWER7 processor options and modes for the Power 780 server
Cores per POWER7
processor
Frequency L3 cache size available per
POWER7 processor
6 3.72 GHz 24 MB
8 3.30 GHz 32 MB
Active cores per
POWER7 processor
System mode Frequency L3 cache size available per
POWER7 processor
8 MaxCore 3.92 GHz 32 MB
4 TurboCore 4.14 GHz 32 MB

With two POWER7 processors in each enclosure, systems can be equipped as follows:
MaxCore mode:
– 16 cores
– 32 cores
– 48 cores
– 64 cores
TurboCore mode:
– 8 cores
– 16 cores
– 24 cores
– 32 cores
2.2.2 Four-socket processor card
A 4-socket processor card is supported on the Power 780 (Figure 2-9), enabling a maximum
system configuration of 96 cores (6-core processors). Each POWER7 processor is connected
to four memory DIMMs through a single memory controller.
Figure 2-9 IBM Power 780 4-socket processor card
Power 780 Systems
Table 2-5 summarizes the POWER7 processor options for the Power 780 4-socket system.
Table 2-5 Summary of POWER7 processor options and modes for the Power 780 server
The TurboCore option is not supported with the 4-socket processor cards.
Cores per POWER7
processor
Frequency L3 cache size available per
POWER7 processor
6 3.44 GHz 24 MB
POWER7
processors
DDR3
Memory
DIMMs

2.2.3 Processor comparison
The 2-socket and 4-socket processor cards available for the Power 780 utilize slightly different
POWER7 processors. Table 2-6 shows a comparison.
Table 2-6 Comparison of processors used with 2-socket and 4-socket processor cards
The most significant difference between the processors are the interconnects. On the
2-socket processor card, the POWER7 processor has two memory controllers, each
connected to four memory DIMMs (Figure 2-10).
Figure 2-10 Processor interconnects on 2-socket processor card
Area POWER7 processor used on
2-socket CPU card
POWER7 processor used on
4-socket CPU card
Technology 45 nm 45 nm
Die size 567 mm2
567 mm2
Power 250 W 150 W
Cores 8 6
Max frequency 3.92 GHz
(4.14 GHz with TurboCore)
3.44 GHz
L2/L3 256 K/4 MB per core 256 K/4 MB per core
Memory support DDR3 DDR3
Fabric Bus Star Fabric Bus Star Fabric Bus
I/O Bus Two GX++ Two GX++
TurboCore mode Yes No
Sleep/nap mode Yes Yes
P7
XYZ
A
B
MC0
MC1
GX1
GX0
DIMM 1
DIMM 2
DIMM 3
DIMM 4
DIMM 1
DIMM 2
DIMM 3
DIMM 4
GX1 Bus (4 Byte – 2.46Gb/s)
Z Y X Buses
8 byte
3.248 Gb/s
A B Buses
8 byte
2.464 Gb/s

The POWER7 processor used on the 4-socket processor card also has two memory
controllers, but only one is used. This results in four DIMMs per memory controller, the same
as the processor used on the 2-socket processor card.
Similarly, the processor used on the 4-socket CPU card has two GX++ buses, but only one is
used (Figure 2-11).
Figure 2-11 Processor interconnects on 4-socket processor card
2.3 Memory subsystem
On the Power 770 and Power 780 servers, independently of using two or four Single Chip
Modules (SCMs), each enclosure houses 16 DDR3 DIMM slots. The DIMM cards for the
Power 770 and Power 780 are 96 mm tall, fully buffered, and placed in one of the 16 DIMM
slots on the processor card.
2.3.1 Fully buffered DIMM
Fully buffered DIMM technology is used to increase reliability, speed, and density of
memory subsystems. Conventionally, data lines from the memory controllers have to be
connected to the data lines in every DRAM module. As memory width and access speed
increases, the signal decays at the interface of the bus and the device. This effect traditionally
degrades either the memory access times or memory density. Fully buffered DIMMs
overcome this effect by implementing an advanced buffer between the memory controllers
and the DRAMs with two independent signaling interfaces. This technique decouples the
DRAMs from the bus and memory controller interfaces, allowing efficient signaling between
the buffer and the DRAM.
2.3.2 Memory placement rules
The minimum DDR3 memory capacity for the Power 770 and Power 780 systems is 64 GB of
installed memory.
P7
XYW
A
B
MC0
MC1
GX1
GX0
DIMM 1
DIMM 2
DIMM 3
DIMM 4
W Y X Buses
8 byte
2.8 Gb/s
B Buses
8 byte
2.464 Gb/s

All the memory DIMMs for the Power 770 and Power 780 are Capacity Upgrade on Demand
capable and must have a minimum of 50% of its physical capacity activated. For example, the
minimum installed memory for both servers is 64 GB RAM, whereas they can have a
minimum of 32 GB RAM active.
Figure 2-12 shows the physical memory DIMM topology for the Power 770 and Power 780
with two single-chip-modules (SCMs).
Figure 2-12 Physical memory DIMM topology for the Power 770 and Power 780 with two SCMs
For the POWER 770 and POWER 780 server models with two SCMs, there are 16 buffered
DIMM slots:
DIMM slots J1A to J8A are connected to the memory controllers on POWER7 processor 0.
DIMM slots J1B to J8B are connected to the memory controllers on POWER7 processor 1.
These DIMMs slots are divided into four Quads, each Quad having four DIMMs.
Note: DDR2 memory (used in POWER6 processor-based systems) is not supported in
POWER7 processor-based systems.
P7
SCM
A
B
A
B
P7
SCM
Regulator #1
Regulator #4
A
B
A
B
Regulator #5
DDR3 DIMM #5
DDR3 DIMM #6
DDR3 DIMM #7
DDR3 DIMM #8
MC0
MC1
DDR3 DIMM #1
DDR3 DIMM #2
DDR3 DIMM #3
DDR3 DIMM #4
DDR3 DIMM #13
DDR3 DIMM #14
DDR3 DIMM #15
DDR3 DIMM #16
DDR3 DIMM #9
DDR3 DIMM #10
DDR3 DIMM #11
DDR3 DIMM #12
MC0
MC1
Regulator #6
Regulator #7
TPMD Slot
Regulator #3
Regulator #2
I/OConnectors
P3-C1
P3-C2
P3-C3
P3-C4
P3-C6
P3-C7
P3-C8
P3-C9
P3-C10
P3-C11
P3-C12
P3-C13
P3-C17
P3-C18
P3-C19
P3-C20
J1A
J2A
J3A
J4A
J5A
J6A
J7A
J8A
J1B
J2B
J3B
J4B
J5B
J6B
J7B
J8B
Quad #1
Quad #1
Quad #2
Quad #2
Quad #1
Quad #1
Quad #2
Quad #2
Quad #3
Quad #3
Quad #4
Quad #4
Quad #3
Quad #3
Quad #4
Quad #4

Figure 2-13 shows the physical memory DIMM topology for the Power 780 with four
single-chip-modules (SCMs).
Figure 2-13 Physical memory DIMM topology for the Power 780 with four SCMs
For the POWER 780 with the four SCMs, there are 16 buffered DIMM slots available:
DIMM slots J1A to J4A are connected to the memory controller on POWER7 processor 0.
DIMM slots J5A to J8A are connected to the memory controller on POWER7 processor 1.
DIMM slots J1B to J4B are connected to the memory controller on POWER7 processor 2.
DIMM slots J5B to J8B are connected to the memory controller on POWER7 processor 3.
The memory plugging rules are as follows:
Plug sequence will always allow for memory mirroring (for example, no Feature Code
needs to be specified for memory mirroring). The green cells in the following tables
indicate the AMM (Active Memory Mirroring) base configuration.
DIMMs must be installed 4x DIMMs at a time, referred to as a DIMM-quad and identified in
Table 2-7 on page 55 and Table 2-8 on page 55 by colors.
DIMM-quads must be homogeneous. (Only DIMMs of the same capacity are allowed on
the same Quad.)
A DIMM-quad is the minimum installable unit for subsequent upgrades.
Although each drawer can have different capacity memory DIMMs, for maximum memory
performance, the total memory capacity on each memory controller should be equivalent.
The DIMM-quad placement rules for a single enclosure are as follows (see Figure 2-12 on
page 53 for the physical memory topology):
– Quad 1: J1A, J2A, J5A, J6A (mandatory minimum for each enclosure)
B
B
B
B
Regulator #5
DDR3 DIMM #5
DDR3 DIMM #6
DDR3 DIMM #7
DDR3 DIMM #8
DDR3 DIMM #1
DDR3 DIMM #2
DDR3 DIMM #3
DDR3 DIMM #4
DDR3 DIMM #13
DDR3 DIMM #14
DDR3 DIMM #15
DDR3 DIMM #16
DDR3 DIMM #9
DDR3 DIMM #10
DDR3 DIMM #11
DDR3 DIMM #12
Regulator #6
Regulator #7
TPMD Slot
I/OConnectors
P3-C1
P3-C2
P3-C3
P3-C4
P3-C6
P3-C7
P3-C8
P3-C9
P3-C10
P3-C11
P3-C12
P3-C13
P3-C17
P3-C18
P3-C19
P3-C20
J1A
J2A
J3A
J4A
J5A
J6A
J7A
J8A
J1B
J2B
J3B
J4B
J5B
J6B
J7B
J8B
Quad #1
Quad #1
Quad #2
Quad #2
Quad #1
Quad #1
Quad #2
Quad #2
Quad #3
Quad #3
Quad #4
Quad #4
Quad #3
Quad #3
Quad #4
Quad #4
P7
SCM
P7
SCM
Regulator #1
Regulator #4
B
B
MC0
MC0
Regulator #3
Regulator #2
P7
SCM
B
MC0
P7
SCM
B
MC0

– Quad 2: J3A, J4A, J7A, J8A (mandatory minimum for each enclosure)
– Quad 3: J1B, J2B, J5B, J6B
– Quad 4: J3B, J4B, J7B, J8B
Table 2-7 shows the optimal placement of each DIMM-quad within a single enclosure system.
Each enclosure must have at least DIMM-quads installed in slots J1A, J2A, J5A, J6A, and
J5A, J6A, J7A, and J8A, as shown with the highlighted color.
Table 2-7 Optimum DIMM-quad placement for a single enclosure system
When populating a multi-enclosure system with DIMM-quads, each enclosure must have at
least two DIMM-quads installed in slots J1A, J2A, J5A, J6A, J3A, J4A, J7A, and J8A.. After
the mandatory requirements and memory-plugging rules are followed, there is an optimal
approach to populating the systems.
Table 2-8 shows the optimal placement of each DIMM-quad within a dual-enclosure system.
Each enclosure must have at least two DIMM-quads installed.
Table 2-8 Optimum DIMM-quad placement for a dual enclosure system
Enclosure 0
POWER7 processor 0 POWER7 processor 1
Memory controller 1 Memory controller 0 Memory controller 1 Memory controller 0
J1A J2A J3A J4A J5A J6A J7A J8A J1B J2B J3B J4B J5B J6B J7B J8B
Q1 Q1 Q2 Q2 Q1 Q1 Q2 Q2 Q3 Q3 Q4 Q4 Q3 Q3 Q4 Q4
Quads Q1 and Q2 must be identical to each other.
Note: For maximum memory performance, the total memory capacity on each memory controller
should be equivalent.
Enclosure 0
Enclosure 1
POWER7 Processor 0 POWER7 Processor 1
Memory Controller 1 Memory Controller 0 Memory Controller 1 Memory Controller 0
Quads Q1 and Q2 must be identical to each other. Quads Q3 and Q4 must be identical to each other.

Table 2-9 shows the optimal placement of each DIMM-quad within a three-enclosure system.
Table 2-9 Optimum DIMM-quad placement for a three-enclosure system
Enclosure 0
Memory controller 1 Memory Controller 0 Memory controller 1 Memory controller 0
Enclosure 1
Enclosure 2
Quads Q5 and Q6 must be identical to each other.

Table 2-10 shows the optimal placement of each DIMM-quad within a four-enclosure system.
Table 2-10 Optimum DIMM-quad placement for a four-enclosure system
Enclosure 0
Enclosure 1
Enclosure 2
Enclosure 3

2.3.3 Memory throughput
POWER7 has exceptional cache, memory, and interconnect bandwidths. Table 2-11 shows
the bandwidth estimate for the Power 770 system running at 3.3 GHz.
Table 2-11 Power 770 memory bandwidth estimates for POWER7 cores running at 3.3 GHz
With an increase in frequency, the Power 780 running at 3.92 GHz generates higher cache
bandwidth (Table 2-12).
In TurboCore mode, the Power 780 will have its cores running at 4.14 GHz generating even
higher cache bandwidth (Table 2-13).
Memory Bandwidth
L1 (data) cache 158.4 Gbps
L2 cache 158.4 Gbps
L3 cache 105.6 Gbps
System memory:
4x enclosures:
136.44 Gbps per socket
1091.58 Gbps
Inter-node buses (four enclosures) 158.02 Gbps
Intra-node buses (four enclosures) 415.74 Gbps
Memory Bandwidth
L2 cache 188.16 Gbps
System memory:
4x enclosures:
1091.58 Gbps
Inter-node buses (4 enclosures) 158.02 Gbps
Intra-node buses (4 enclosures) 415.74 Gbps
Memory Bandwidth
System memory:
4x enclosures:
1091.58 Gbps
Inter-node buses (4 enclosures) 158.02 Gbps
Intra-node buses (4 enclosures) 415.74 Gbps

2.3.4 Active Memory Mirroring
Power 770 and Power 780 servers have the ability to provide mirroring of the hypervisor
code among different memory DIMMs. This feature will enhance the availability of a server
and keep it operable in case a DIMM failure occurs in one of the DIMMs that hold the
hypervisor code.
The hypervisor code, which resides on the initial DIMMs (J1A to J8A), will be mirrored on the
same group of DIMMs to allow for more usable memory. Table 2-9 on page 56 shows the
DIMMs involved on the Memory Mirroring operation.
Figure 2-14 shows how Active Memory Mirroring uses different DIMMs.
Figure 2-14 Memory Mirroring among different DIMMs
To enable the Active Memory Mirroring (AMM) feature, the server must have 8x DIMMs of the
same size populated on slots J1A to J8A. It is also mandatory that the server has enough free
memory to accommodate the mirrored memory pages. Active Memory Mirroring is required
and automatically enabled on the Power 780. However, on the Power 770 AMM is optional
and is ordered and enabled via feature #4797.
Besides the hypervisor code itself, other components that are vital to the server operation are
also mirrored:
Hardware page tables (HPTs), responsible for tracking the state of the memory pages
assigned to partitions
Translation control entities (TCEs), responsible for providing I/O buffers for the
partition’s communications
Memory used by the hypervisor to maintain partition configuration, I/O states, Virtual I/O
information, and partition state
There are components that are not mirrored after they are not vital to the regular server
operations and would require a larger amount of memory to accommodate its data:
Advanced Memory Sharing Pool
Memory used to hold the contents of platform dumps
Note: Active Memory Mirroring will not mirror partition data. It was designed to mirror
only the hypervisor code and its components, allowing this data to be protected against a
DIMM failure.
POWER7
Processor Chip
MC0
MC1
Hypervisor Data J1A
Hypervisor Data J2A
J3A
J4A
Mirrored Hypervisor Data J5A
Mirrored Hypervisor Data J6A
J7A
J8A

It is possible to check whether the Memory Mirroring option is enabled and change its current
status via HMC, under the Advanced Tab on the CEC Properties Panel (Figure 2-15).
Figure 2-15 CEC Properties window on an HMC
After a failure on one of the DIMMs containing hypervisor data occurs, all the server
operations remain active and the service processor will isolate the failing DIMMs. Because
there are no longer eight functional DIMMs behind a memory controller, Active Memory
Mirroring will not be available until this DIMM is replaced. Systems will stay in the partially
mirrored state until the failing DIMM is replaced.
2.4 Capacity on Demand
Several types of Capacity on Demand (CoD) are optionally available on the Power 770 and
Power 780 servers to help meet changing resource requirements in an on demand
environment by using resources that are installed on the system but that are not activated.
2.4.1 Capacity Upgrade on Demand (CUoD)
CUoD allows you to purchase additional permanent processor or memory capacity and
dynamically activate them when needed.
2.4.2 On/Off Capacity on Demand (On/Off CoD)
On/Off Capacity on Demand allows you to temporarily activate and deactivate processor
cores and memory units to help meet the demands of business peaks such as seasonal
activity, period-end, or special promotions. When you order an On/Off CoD feature, you
receive an enablement code that allows a system operator to make requests for additional
processor and memory capacity in increments of one processor day or 1 GB memory day.
The system monitors the amount and duration of the activations. Both prepaid and post-pay
options are available.

On the post-pay options, charges are based on usage reporting collected monthly.
Processors and memory can be activated and turned off an unlimited number of times when
additional processing resources are needed.
This offering provides a system administrator an interface at the HMC to manage the
activation and deactivation of resources. A monitor that resides on the server records the
usage activity. This usage data must be sent to IBM on a monthly basis. A bill is then
generated based on the total amount of processor and memory resources utilized, in
increments of processor and memory (1 GB) days.
Before using temporary capacity on your server, you must enable your server. To do this, an
enablement feature (MES only) must be ordered, and the required contracts must be in place.
If a Power 770 or Power 780 server uses the IBM i operating system in addition to any other
supported operating system on the same server, the client must inform IBM which operating
system caused the temporary On/Off CoD processor usage so that the correct feature can be
used for billing.
The features that are used to order enablement codes and support billing charges on the
Power 770 and Power 780 can be seen in 1.4.6, “Summary of processor features” on
page 15, and 1.4.7, “Memory features” on page 19.
The On/Off CoD process consists of three steps:
1. Enablement
Before requesting temporary capacity on a server, you must enable it for On/Off CoD. To
do this, order an enablement feature and sign the required contracts. IBM will generate an
enablement code, mail it to you, and post it on the web for you to retrieve and enter on the
target server.
A processor enablement code allows you to request up to 360 processor days of
temporary capacity. If the 360 processor-day limit is reached, place an order for another
processor enablement code to reset the number of days that you can request back to 360.
A memory enablement code lets you request up to 999 memory days of temporary
capacity. If you have reached the limit of 999 memory days, place an order for another
memory enablement code to reset the number of allowable days that you can request
back to 999.
2. Activation requests
When On/Off CoD temporary capacity is needed, simply use the HMC menu for
On/Off CoD. Specify how many of the inactive processors or GB of memory are required
to be temporarily activated for a certain number of days. You will be billed for the days
requested, whether the capacity is assigned to partitions or left in the Shared
Processor Pool.
At the end of the temporary period (days that were requested), you must ensure that the
temporarily activated capacity is available to be reclaimed by the server (not assigned to
partitions), or you are billed for any unreturned processor days.
3. Billing
The contract, signed by the client before receiving the enablement code, requires the
On/Off CoD user to report billing data at least once a month (whether or not activity
occurs). This data is used to determine the proper amount to bill at the end of each billing
period (calendar quarter). Failure to report billing data for use of temporary processor or
memory capacity during a billing quarter can result in default billing equivalent to 90
processor days of temporary capacity.

For more information regarding registration, enablement, and usage of On/Off CoD, visit:
http://www.ibm.com/systems/power/hardware/cod
2.4.3 Utility Capacity on Demand (Utility CoD)
Utility CoD automatically provides additional processor performance on a temporary basis
within the Shared Processor Pool.
Utility CoD enables you to place a quantity of inactive processors into the server's Shared
Processor Pool, which then becomes available to the pool's resource manager. When the
server recognizes that the combined processor utilization within the Shared Processor Pool
exceeds 100% of the level of base (purchased and active) processors assigned across
uncapped partitions, then a Utility CoD Processor Minute is charged and this level of
performance is available for the next minute of use.
If additional workload requires a higher level of performance, the system automatically
allows the additional Utility CoD processors to be used, and the system automatically and
continuously monitors and charges for the performance needed above the base
(permanent) level.
Registration and usage reporting for Utility CoD is made using a public website, and payment
is based on reported usage. Utility CoD requires PowerVM Standard Edition or PowerVM
Enterprise Edition to be active.
If a Power 770 or Power 780 server uses the IBM i operating system in addition to any other
supported operating system on the same server, the client must inform IBM which operating
system caused the temporary Utility CoD processor usage so that the correct feature can be
used for billing.
For more information regarding registration, enablement, and use of Utility CoD, visit:
http://www.ibm.com/systems/support/planning/capacity/index.html
2.4.4 Trial Capacity On Demand (Trial CoD)
A standard request for Trial CoD requires you to complete a form including contact
information and vital product data (VPD) from your Power 770 or Power 780 system with
inactive CoD resources.
A standard request activates two processors or 4 GB of memory (or both two processors
and 4 GB of memory) for 30 days. Subsequent standard requests can be made after each
purchase of a permanent processor activation. An HMC is required to manage Trial
CoD activations.
An exception request for Trial CoD requires you to complete a form including contact
information and VPD from your Power 770 or Power 780 system with inactive CoD
resources. An exception request will activate all inactive processors or all inactive memory (or
all inactive processor and memory) for 30 days. An exception request can be made only one
time over the life of the machine. An HMC is required to manage Trial CoD activations.
To request either a Standard or an Exception Trial, visit:
https://www-912.ibm.com/tcod_reg.nsf/TrialCod?OpenForm

2.4.5 Software licensing and CoD
For software licensing considerations with the various CoD offerings, see the most recent
revision of the Capacity on Demand User’s Guide at:
http://www.ibm.com/systems/power/hardware/cod
2.5 CEC Enclosure interconnection cables
IBM Power 770 or 780 systems can be configured with more than one system enclosure.
The connection between the processor cards in the separate system enclosures requires
a set of processor drawer interconnect cables. Each system enclosure must be connected
to each other through a flat flexible SMP cable. These cables are connected on the front of
the drawers.
Furthermore, service processor cables are needed to connect the components in each
system enclosure to the active service processor for system functions monitoring.
These cables connect at the rear of each enclosure and are required for two-drawer,
three-drawer, and four-drawer configurations.
The star fabric bus topology that connects the processors together in separate drawers is
contained on SMP Flex cables that are routed external to the drawers. These flex cables
attach directly to the CPU cards at the front of the drawer and are routed behind the front
covers of the drawers.
For performance reasons, there will be multiple link connections between the CEC
enclosures, and there will be differences in SMP Flex cabling between the two SCM
processor card configurations (#4983, #4984, #5003) and the four SCM processor cards
configurations (#EP24).
The SMP and FSP cable features described in Table 2-14 are required to connect
the processors together when system configuration is made of two, three, or four
CEC enclosures.
Table 2-14 Required flex cables feature codes
The cables are designed to support hot-addition of a system enclosure up to the maximum
scalability. When adding a new drawer, existing cables remain in place and new cables are
added. The only exception is for cable #3711, which is replaced when growing from a
2-drawer to a 3-drawer configuration.
CEC enclosures SMP cables for the
0-12 core (#4983) and
0-16 core (#4984)
processor cards
SMP cables for the
0-24 core (#EP24)
processor cards
FSP cables
Two #3711 and #3712 #3715 and #3716 #3671
Three #3712 and #3713 #3715, #3716, #3717 #3671 and #3672
Four #3712, #3713, #3714 #3716, #3717, #3718 #3671, #3672, #3673
Note: The #3712 and #3716 provide two SMP cable sets, and #3714 and #3717 provide
three SMP cable sets.

The cables are also designed to allow the concurrent maintenance of the Power 770 or
Power 780 in case the IBM service representative needs to extract a system enclosure from
the rack. The design of the flexible cables allows each system enclosure to be disconnected
without any impact on the other drawers.
To allow such concurrent maintenance operation, plugging in the SMP Flex cables in the
order of their numbering is extremely important. Each cable is numbered (Figure 2-16).
Figure 2-16 SMP cable installation order
Table 2-15 reports the SMP cable usage for two-enclosure or three-enclosure scenarios.
Table 2-15 SMP cable two and three enclosures scenario
Processor type SMP cables Cable number From
connector
To
connector
Two CEC enclosures
#4983, #4984, #5003,
(0-12 core, 0-16 core)
#3711 1 U1-P3-T1 U2-P3-T1
#3712 2 U1-P3-T4 U2-P3-T4
#EP24 (0-24 core) #3715 1 U1-P3-T1 U2-P3-T1
#3716 4 U1-P3-T2 U3-P3-T2
2 U1-P3-T4 U2-P3-T4
Three CEC enclosures
#4983, #4984, #5003,
(0-12 core, 0-16 core)
#3712 2 U1-P3-T4 U2-P3-T4
#3713 3 U1-P3-T2 U3-P3-T1
4 U2-P3-T2 U3-P3-T2

Table 2-16 reports the SMP cable usage for the four-enclosure scenario.
Table 2-16 SMP cable four-enclosures scenario
When adding CEC enclosures in an MES upgrade, SMP cables will likely have to be added
depending on how many enclosures are being added.
Adding more than one CEC enclosure to an existing configuration can be accomplished with
a hot-add if the enclosures are added one at a time. The other option is to bring the entire
system down and add all the additional enclosures at the same time. Depending on whether
the hot-add option is desired, certain SMP cable features might or might not be necessary.
#EP24 (0-24 core) #3715 1 U1-P3-T1 U2-P3-T1
#3716 4 U2-P3-T2 U3-P3-T2
2 U1-P3-T4 U2-P3-T4
#3717 7 U2-P3-T3 U3-P3-T3
3 U1-P3-T2 U3-P3-T1
6 U1-P3-T3 U3-P3-T4
connector
To
connector
#4983, #4984, #5003
(0-12 core, 0-16 core)
#3712 2 U1-P3-T4 U2-P3-T4
#3713 3 U1-P3-T2 U3-P3-T1
4 U2-P3-T2 U3-P3-T2
#3714 5 U1-P3-T1 U4-P3-T4
6 U2-P3-T3 U4-P3-T3
7 U3-P3-T3 U4-P3-T3
#EP24 (0-24 core) #3716 4 U2-P3-T2 U3-P3-T2
2 U1-P3-T4 U2-P3-T4
#3717 7 U3-P3-T3 U4-P3-T3
3 U1-P3-T2 U3-P3-T1
6 U2-P3-T3 U4-P3-T4
#3718 5 U1-P3-T1 U4-P3-T1
connector
To
connector

Similarly, the Flexible Service Processor (FSP) flex cables must be installed in the correct
order (Figure 2-17), as follows:
1. Install a second node flex cable from node 1 to node 2.
2. Add a third node flex cable from node 1 and node 2 to node 3.
3. Add a fourth node flex cable from node 1 and node 2 to node 4.
Figure 2-17 FSP flex cables
The design of the Power 770 and Power 780 is optimized for use in an IBM 7014-T00 or
7014-T42 rack. Both the front cover and the external processor fabric cables occupy space on
the front left and right side of an IBM 7014 rack. Racks that are not from IBM might not offer
the same room. When a Power 770 or Power 780 is configured with two or more system
enclosures in a 7014-T42 or 7014-B42 rack, the CEC enclosures must be located in EIA 36 or
below to allow space for the flex cables.

The total width of the server, with cables installed, is 21 inches (Figure 2-18).
Figure 2-18 Front view of the rack with SMP cables overlapping the rack rails
SMP
Cable
Drawer 1 - A Left
to
Drawer 4 - A Left
SMP
Cable
Drawer 2 – B Left
to
Drawer 3 - B Left
SMP
Cable
Drawer 1 – B Left
to
Drawer 3 - A Left
SMP
Cable
Drawer 1 - B Right
to
Drawer 2 - B Right
SMP
Cable
Drawer 2 - A Right
to
Drawer 4 - B Right
SMP
Cable
Drawer 3 - A Right
to
Drawer 4 - A Right
SMP
Cable
Drawer 1 - A Left
to
Drawer 4 - A Left
SMP
Cable
Drawer 2 – B Left
to
Drawer 3 - B Left
SMP
Cable
Drawer 1 – B Left
to
Drawer 3 - A Left
SMP
Cable
Drawer 1 - B Right
to
Drawer 2 - B Right
SMP
Cable
Drawer 2 - A Right
to
Drawer 4 - B Right
SMP
Cable
Drawer 3 - A Right
to
Drawer 4 - A Right

In the rear of the rack, the FSP cables require only some room in the left side of the racks
(Figure 2-19).
Figure 2-19 Rear view of rack with detail of FSP flex cables
2.6 System bus
This section provides additional information related to the internal buses.
2.6.1 I/O buses and GX++ card
Each CEC enclosure of the Power 770 and Power 780 contains one POWER7 processor
card, where as each processor card comprises either two single-chip module POWER7
processors (#4983, #4984, or #5003) or, new to the Power 780, four single-chip module
POWER7 processors (#EP24).
Within a CEC enclosure a total of four GX++ buses are available for I/O connectivity and
expansion. Two GX++ buses off one of the two sockets are routed through the midplane to
the I/O backplane and drive two POWER7 I/O chips (POWER7 IOC) on the I/O backplane.
The two remaining GX++ buses from either of the two sockets are routed to the midplane and
feed two GX++ adapter slots.
The optional GX++ 12X DDR Adapter Dual-port (#1808), which is installed in the GX++
adapter slot, enables the attachment of a 12X loop, which runs at either SDR or DDR speed
depending upon the 12X I/O drawers attached. These GX++ adapter slots are hot-pluggable
and do not share space with any of the PCIe slots.
Two Drawer Cable
Three Drawer Cable
Four Drawer Cable
Two Drawer Cable
Three Drawer Cable
Four Drawer Cable

Table 2-17 shows the I/O bandwidth for available processors cards.
Table 2-17 I/O bandwidth
2.6.2 Flexible Service Processor bus
The Flexible Service Processor (FSP) flex cable, which is located at the rear of the system, is
used for FSP communication between the system drawers. A FSP card (#5664) is installed in
system drawer 1 and system drawer 2, and the FSP/Clock Pass-Through card (#5665) is
installed in system drawer 3 and system drawer 4 for connecting FSP flex cable. The FSP
cable has been changed to point-to-point cabling similar to the processor drawer interconnect
cable. When a system drawer is added, another FSP flex cable is added. A detailed cable
configuration is discussed in 2.5, “CEC Enclosure interconnection cables” on page 63.
2.7 Internal I/O subsystem
The internal I/O subsystem resides on the I/O planar, which supports eight PCIe slots. All
PCIe slots are hot-pluggable and enabled with enhanced error handling (EEH). In the unlikely
event of a problem, EEH-enabled adapters respond to a special data packet that is generated
from the affected PCIe slot hardware by calling system firmware, which examines the
affected bus, allows the device driver to reset it, and continues without a system reboot.
Processor card Slot description Frequency Device Bandwidth
(maximum
theoretical)
#4983, #4984
or #5003
CPU Socket 0
(CP0) GX bus 1
2.464 Gbps P7IOC-B 19.712 GBps
CPU Socket 0
(CP0) GX bus 0
2.464 Gbps P7IOC-A 19.712 GBps
CPU Socket 1
(CP1) GX bus 1
2.464 Gbps GX slot 2 19.712 GBps
CPU Socket 1
(CP1) GX bus 0
2.464 Gbps GX slot 1 19.712 GBps
Single enclosure 78.848 GBps
Total (4x enclosures) 315.392 GBps
#EP24 CPU Socket 0
(CP0) GX bus 1
2.464 Gbps P7IOC-B 19.712 GBps
CPU Socket 2
(CP2) GX bus 0
2.464 Gbps P7IOC-A 19.712 GBps
CPU Socket 3
(CP3) GX bus 0
2.464 Gbps GX slot - lower 19.712 GBps
CPU Socket 1
(CP1) GX bus 1
2.464 Gbps GX slot - upper 19.712 GBps
Single enclosure 78.848 GBps
Total (4x enclosures) 315.392 GBps

Table 2-18 lists the slot configuration of the Power 770 and Power 780.
Table 2-18 Slot configuration of the Power 770 and 780
2.7.1 Blind-swap cassettes
The Power 770 and Power 780 uses new fourth-generation blind-swap cassettes to manage
the installation and removal of adapters. This new mechanism requires an interposer card
that allows the PCIe adapters to plug in vertically to the system, allows more airflow through
the cassette, and provides more room under the PCIe cards to accommodate the GX+
multifunctional host bridge chip heat sink height. Cassettes can be installed and removed
without removing the CEC enclosure from the rack.
2.7.2 System ports
Each CEC enclosure is equipped with a Integrated Multifunction Card (#1768 or #1769). This
integrated card provides two USB ports, one serial port, and four Ethernet connectors for a
processor enclosure and does not require a PCIe slot. When ordering a Power 770 or
Power 780, the following options can be selected:
Dual 10 Gb Copper and Dual 1 Gb Ethernet (#1768)
Dual 10 Gb Optical and Dual 1 Gb Ethernet (#1769)
All of the connectors are on the rear bulkhead of the CEC, and one Integrated Multifunction
Card can be placed in an individual CEC enclosure. An Integrated Multifunction Card is
required in CEC enclosures one and two, but it is not required in CEC enclosures three or
four. On a multi-enclosure system, the Integrated Multifunction Card features can differ. The
copper twinax ports support up to 5 m cabling distances. The RJ-45 ports support up to
100 m cabling distance using a Cat5e cable. The optical ports only support the 850 Nm optic
cable (multi mode cable) and support up to 300 m cabling distances.
The Power 770 and Power 780 each support one serial port in the rear of the system. This
connector is a standard 9-pin male D-shell, and it supports the RS232 interface. Because the
Power 770 and Power 780 are managed by an HMC/SDMC, this serial port is always
controlled by the operating system, and therefore is available in any system configuration. It
is driven by the integrated PLX Serial chip, and it will support any serial device that has an
operating system device driver. The FSP virtual console will be on the HMC /SDMC.
Slot number Description Location code PCI host bridge (PHB) Max. card size
Slot 1 PCIe Gen2 x8 P2-C1 P7IOC A PCIe PHB5 Full length
Slot 5 PCIe Gen2 x8 P2-C5 P7IOC B PCIe PHB5 Full length
Slot 6 PCIe Gen2 x8 P2-C6 P7IOC B PCIe PHB4 Full length
Slot 7 GX++ P1-C2 - -
Slot 8 GX++ P1-C3 - -

2.8 PCI adapters
This section covers the different types and functionalities of the PCI cards supported by IBM
Power 770 and Power 780 systems.
2.8.1 PCIe Gen1 and Gen2
Peripheral Component Interconnect Express (PCIe) uses a serial interface and allows for
point-to-point interconnections between devices (using a directly wired interface between
these connection points). A single PCIe serial link is a dual-simplex connection that uses two
pairs of wires, one pair for transmit and one pair for receive, and can transmit only one bit per
cycle. These two pairs of wires are called a lane. A PCIe link can consist of multiple lanes. In
such configurations, the connection is labeled as x1, x2, x8, x12, x16, or x32, where the
number is effectively the number of lanes.
Two generations of PCIe interface are supported in Power 770 and Power 780 models:
Gen1: Capable of transmitting at the extremely high speed of 2.5 Gbps, which gives a
capacity of a peak bandwidth of 2 GBps simplex on an x8 interface
Gen2: Double the speed of the Gen1 interface, which gives a capacity of a peak
bandwidth of 4 GBps simplex on an x8 interface
PCIe Gen1 slots support Gen1 adapter cards and also most of the Gen2 adapters. In this
case, where a Gen2 adapter is used in a Gen1 slot, the adapter will operate at PCIe Gen1
speed. PCIe Gen2 slots support both Gen1 and Gen2 adapters. In this case, where a Gen1
card is installed into a Gen2 slot, it will operate at PCIe Gen1 speed with a slight performance
enhancement. When a Gen2 adapter is installed into a Gen2 slot, it will operate at the full
PCIe Gen2 speed.
The IBM Power 770 and Power 780 CEC enclosure is equipped with six PCIe x8 Gen2 slots.
2.8.2 PCI-X adapters
IBM offers PCIe adapter options for the Power 770 and Power 780 CEC enclosure. If a
PCI-extended (PCI-X) adapter is required, a PCI-X DDR 12X I/O Drawer (#5796) can be
attached to the system by using a GX++ adapter loop. PCIe adapters use a different type
of slot than PCI and PCI-X adapters. If you attempt to force an adapter into the wrong type
of slot, you might damage the adapter or the slot. All adapters support Extended Error
Handling (EEH).
2.8.3 IBM i IOP adapters
IBM i IOP adapters are not supported with the Power 770 and Power 780, which has
these results:
Existing PCI adapters that require an IOP are affected.
Existing I/O devices are affected, such as certain tape libraries or optical drive libraries, or
any HVD SCSI device.
Twinax displays or printers cannot be attached except through an OEM protocol converter.
Before adding or rearranging adapters, use the System Planning Tool to validate the new
adapter configuration. See the System Planning Tool website:
http://www.ibm.com/systems/support/tools/systemplanningtool/

If you are installing a new feature, ensure that you have the software required to support the
new feature, and determine whether there are any existing PTF prerequisites to install. See
the IBM Prerequisite website for information:
https://www-912.ibm.com/e_dir/eServerPreReq.nsf
2.8.4 PCIe adapter form factors
IBM POWER7 processor-based servers are able to support two form factors of
PCIe adapters:
PCIe low profile (LP) cards, which are used with the Power 710 and Power 730 PCIe
slots. Low profile adapters are also used in the PCIe riser card slots of the Power 720 and
Power 740 servers.
PCIe full-height and full-high cards, which are plugged into the following servers slots:
– Power 720 and Power 740 (Within the base system, five PCIe half-length slots
are supported.)
– Power 750
– Power 755
– Power 770
– Power 780
– Power 795
– PCIe slots of the #5802 and #5877 drawers
Low-profile PCIe adapter cards are only supported in low-profile PCIe slots, and full-height
and full-high cards are only supported in full-high slots.
Figure 2-20 lists the PCIe adapter form factors.
Figure 2-20 PCIe adapter form factors
Many of the full-height card features are available in low-profile format. For example, the
#5273 8 Gb dual port Fibre Channel adapter is the low-profile adapter equivalent of the
#5735 adapter full-height. They have equivalent functional characteristics.
• Low Profile PCIe Slots
– Power 710 / 730
– Power 720 / 740
• PCIe Expansion riser
Low Profile
Full Height Full High
• Full Height / High PCIe Slots
– Power 720 / 740 / 750 / 770 / 780
– 12X PCIe I/O Drawer
• #5802 / 5877
• #5803 / 5873

Table 2-19 is a list of low-profile adapter cards and their equivalent in full height.
Table 2-19 Equivalent adapter cards
Before adding or rearranging adapters, use the System Planning Tool to validate the new
adapter configuration. See the System Planning Tool website:
If you are installing a new feature, ensure that you have the software required to support the
new feature and determine whether there are any existing update prerequisites to install. To
do this, use the IBM Prerequisite website:
https://www-912.ibm.com/e_dir/eServerPreReq.nsf
The following sections discuss the supported adapters and provide tables of orderable
feature numbers. The tables indicate operating support, AIX (A), IBM i (i), and Linux (L), for
each of the adapters.
2.8.5 LAN adapters
To connect a Power 770 and Power 780 to a local area network (LAN), you can use the
Integrated Multifunction Card. For more information see 2.7.2, “System ports” on page 70.
Low profile Adapter description Full height
Feature
code
CCIN Feature
code
CCIN
#2053 57CD PCIe RAID and SSD SAS Adapter 3 Gb Low
Profile
#2054 or
#2055
57CD or
57CD
#5269 POWER GXT145 PCI Express Graphics
Accelerator (LP)
#5748 5748
#5270 10 Gb FCoE PCIe Dual Port adapter (LP) #5708 2BCB
#5271 4-Port 10/100/1000 Base-TX PCI-Express
adapter
#5717 5717
#5272 10 Gigabit Ethernet-CX4 PCI Express adapter
(LP)
#5732 5732
#5273 8 Gigabit PCI Express Dual Port Fibre Channel
adapter (LP)
#5735 577D
#5274 2-Port Gigabit Ethernet-SX PCI Express
adapter (LP)
#5768 5768
#5275 10 Gb ENet Fibre RNIC PCIe 8x #5769 5769
#5276 4 Gigabit PCI Express Dual Port Fibre Channel
adapter (LP)
#5774 5774
#5277 4-Port Async EIA-232 PCIe adapter (LP) #5785
#5278 SAS Controller PCIe 8x #5901 57B3
Note: The Integrated Multifunction Card can be shared to LPARS using VIOS.

Other LAN adapters are supported in the CEC enclosure PCIe slots or in I/O enclosures that
are attached to the system using a 12X technology loop. Table 2-20 lists the additional LAN
adapters that are available.
Table 2-20 Available LAN adapters
Feature
code
CCIN Adapter description Slot Size OS
support
#5269 10 Gigabit Ethernet-SR PCI Express
adapter
PCIe Short,LP A, L
#5287 PCIe2 2-port 10 GbE SR adapter PCIe Low profile
Short
A, L
#5288 PCIe2 2-Port 10 GbE SFP+Copper
adapter
PCIe Full height
Short
A, L
#5706 5706 IBM 2-Port 10/100/1000 Base-TX
Ethernet PCI-X adapter
PCI-X Full height
Short
A, i, L
#5717 5717 4-Port 10/100/1000 Base-TX PCI
Express adapter
PCIe Full height
Short
A, L
#5732 5732 10 Gigabit Ethernet-CX4 PCI Express
adapter
PCIe Full height
Short
A, L
#5740 4-Port 10/100/1000 Base-TX PCI-X
adapter
PCI-X Full height
Short
A, L
#5744 2B44 PCIe2 4-Port 10 GbE&1 GbE
SR&RJ45 adapter
PCIe Full high L
#5745 2B43 PCIe2 4-Port 10 GbE&1 GbE
SFP+Copper&RJ45 adapter
PCIe Short L
#5767 5767 2-Port 10/100/1000 Base-TX Ethernet
PCI Express adapter
PCIe Full height
Short
A, i, L
#5768 5768 2-Port Gigabit Ethernet-SX PCI
Express adapter
PCIe Full height
Short
A, i, L
#5769 5769 10 Gigabit Ethernet-SR PCI Express
adapter
PCIe Full height
Short
A, L
#5772 576E 10 Gigabit Ethernet-LR PCI Express
adapter
PCIe Full height
Short
A, i, L

2.8.6 Graphics accelerator adapters
The IBM Power 770 and Power 780 support up to eight graphics adapters (Table 2-21). They
can be configured to operate in either 8-bit or 24-bit color modes. These adapters support
both analog and digital monitors, and do not support hot-plug. The total number of graphics
accelerator adapters in any one partition cannot exceed four.
Table 2-21 Available graphics accelerator adapters
2.8.7 SCSI and SAS adapters
To connect to external SCSI or SAS devices, the adapters that are listed in Table 2-22 are
available to be configured.
Table 2-22 Available SCSI and SAS adapters
Feature
code
support
#2849a
a. Supported, but no longer orderable.
2849 POWER GXT135P Graphics
Accelerator with Digital Support
PCI-X Short A, L
#5748 POWER GXT145 PCI Express
Graphics Accelerator
PCIe Short A, L
Feature
code
support
#1912a
a. Supported, but no longer orderable.
571A PCI-X DDR Dual Channel Ultra320
SCSI adapter
PCI-X Short A, i, L
#2055 57CD PCIe RAID and SSD SAS Adapter
3 Gb with Blind Swap Cassette
PCIe Short A, i, L
#5646 Blind Swap Type III Cassette- PCIe,
Short Slot
PCIe Short N/A
#5647 Blind Swap Type III Cassette- PCI-X or
PCIe, Standard Slot
PCI-X
or PCIe
Short N/A
#5736 571A PCI-X DDR Dual Channel Ultra320
SCSI adapter
PCI-X Short A, i, L
#5901 57B3 PCIe Dual-x4 SAS adapter PCIe Short A, i, L
#5903a b
574E PCIe 380MB Cache Dual - x4 3 Gb
SAS RAID adapter
PCIe Short A, i, L
#5908 575C PCI-X DDR 1.5 GB Cache SAS RAID
adapter (BSC)
PCI-X Long A, i, L
#5912 572A PCI-X DDR Dual - x4 SAS adapter PCI-X Short A, i, L
#5913b
b. A pair of adapters is required to provide mirrored write cache data and adapter redundancy
57B5 PCIe2 1.8 GB Cache RAID SAS
adapter Tri-port 6 Gb
PCIe A, i, L
#7863 PCI Blind Swap Cassette Kit, Double
Wide Adapters, Type III
PCI Short N/A

Table 2-23 compares Parallel SCSI to SAS attributes.
Table 2-23 Comparing Parallel SCSI to SAS
2.8.8 iSCSI adapters
iSCSI adapters in Power Systems provide the advantage of increased bandwidth through the
hardware support of the iSCSI protocol. The 1 Gigabit iSCSI TOE (TCP/IP Offload Engine)
PCI-X adapters support hardware encapsulation of SCSI commands and data into TCP, and
transports them over the Ethernet using IP packets. The adapter operates as an iSCSI TOE.
This offload function eliminates host protocol processing and reduces CPU interrupts. The
adapter uses a small form factor LC type fiber optic connector or a copper RJ45 connector.
Table 2-24 lists the orderable iSCSI adapters.
Table 2-24 Available iSCSI adapters
Items to compare Parallel SCSI SAS
Architecture Parallel, all devices connected to
shared bus
Serial, point-to-point, discrete
signal paths
Performance 320 MBps (Ultra320 SCSI),
performance degrades as devices
are added to shared bus
3 Gbps, scalable to 12 Gbps,
performance maintained as more
devices are added
Scalability 15 drives Over 16,000 drives
Compatibility Incompatible with all other drive
interfaces
Compatible with Serial ATA (SATA)
Max. cable length 12 meters total (must sum lengths
of all cables used on bus)
8 meters per discrete connection,
total domain cabling hundreds of
meters
Cable from factor Multitude of conductors adds bulk,
cost
Compact connectors and cabling
save space, cost
Hot pluggability No Yes
Device identification Manually set, user must ensure no
ID number conflicts on bus
Worldwide unique ID set at time of
manufacture
Termination Manually set, user must ensure
proper installation and
functionality of terminators
Discrete signal paths enable
device to include termination by
default
Feature
code
support
#5713 573B 1 Gigabit iSCSI TOE PCI-X on Copper
Media Adapter
PCI-X Short A, i, L
#5714a
a. Supported, but no longer orderable
573C 1 Gigabit iSCSI TOE PCI-X on Optical
Media Adapter
PCI-X A, i, L

2.8.9 Fibre Channel adapter
The IBM Power 770 and Power 780 servers support direct or SAN connection to devices that
use Fibre Channel adapters. Table 2-25 summarizes the available Fibre Channel adapters.
All of these adapters except #5735 have LC connectors. If you attach a device or switch with
an SC type fibre connector, an LC-SC 50 Micron Fiber Converter Cable (#2456) or an LC-SC
62.5 Micron Fiber Converter Cable (#2459) is required.
Table 2-25 Available Fibre Channel adapters
2.8.10 Fibre Channel over Ethernet
Fibre Channel over Ethernet (FCoE) allows for the convergence of Fibre Channel and
Ethernet traffic onto a single adapter and converged fabric.
Figure 2-21 shows a comparison between an existing FC and network connection and a
FCoE connection.
Figure 2-21 Comparison between existing FC and network connection and FCoE connection
Feature
code
support
5729 a b
a. A Gen2 PCIe slot is required to provide the bandwidth for all four ports to operate at full speed.
PCIe2 8 Gb 4-port Fibre Channel
Adapter
PCIe A, L
5735b
b. N_Port ID Virtualization (NPIV) capability is supported through VIOS.
577D 8 Gigabit PCI Express Dual Port Fibre
Channel Adapter
PCIe Short A, i, L
5749 576B 4 Gbps Fibre Channel (2-Port) PCI-X Short i
5758 1910
280D
280E
4 Gb Single-Port Fibre Channel PCI-X
2.0 DDR Adapter
PCI-X Short A, L
5759 1910
5759
4 Gb Dual-Port Fibre Channel PCI-X
2.0 DDR Adapter
PCI-X Short A, L
5774 5774 4 Gigabit PCI Express Dual Port Fibre
Channel Adapter
PCIe Short A, i, L
Ethernet
and Fibre
Channel
Cables
Ethernet
Cable
Fibre Channel
Cable
FC Switch
Ethernet Switch
CEC or I/O Drawer
Ethernet
CEC or I/O Drawer
FC
Rack
Fibre Channel (FC)
Device or FC Switch
Ethernet
Cables
Ethernet
Cable
Fibre Channel
Cable
FCoE Switch
CEC or I/O Drawer
Rack
Fibre Channel (FC)
Device or FC Switch
FCoE
Ethernet Device/
Switch
Ethernet Device/
Switch or FCoE
Device/Switch

For more information about FCoE, read An Introduction to Fibre Channel over Ethernet, and
Fibre Channel over Convergence Enhanced Ethernet, REDP-4493.
IBM offers a 10 Gb FCoE PCIe Dual Port adapter (#5708). This is a high-performance 10 Gb
dual port PCIe Converged Network Adapter (CNA) utilizing SR optics. Each port can provide
Network Interface Card (NIC) traffic and Fibre Channel functions simultaneously. It is
supported on AIX and Linux for FC and Ethernet.
2.8.11 InfiniBand Host Channel adapter
The InfiniBand Architecture (IBA) is an industry-standard architecture for server I/O and
inter-server communication. It was developed by the InfiniBand Trade Association (IBTA) to
provide the levels of reliability, availability, performance, and scalability necessary for present
and future server systems with levels significantly better than can be achieved by using
bus-oriented I/O structures.
InfiniBand (IB) is an open set of interconnect standards and specifications. The main IB
specification has been published by the InfiniBand Trade Association and is available at:
http://www.infinibandta.org/
InfiniBand is based on a switched fabric architecture of serial point-to-point links, where these
IB links can be connected to either host channel adapters (HCAs), used primarily in servers,
or to target channel adapters (TCAs), used primarily in storage subsystems.
The InfiniBand physical connection consists of multiple byte lanes. Each individual byte lane
is a four-wire, 2.5, 5.0, or 10.0 Gbps bidirectional connection. Combinations of link width and
byte-lane speed allow for overall link speeds from 2.5 Gbps to 120 Gbps. The architecture
defines a layered hardware protocol, as well as a software layer to manage initialization and
the communication between devices. Each link can support multiple transport services for
reliability and multiple prioritized virtual communication channels.
For more information about InfiniBand, read HPC Clusters Using InfiniBand on IBM Power
Systems Servers, SG24-7767.
IBM offers the GX++ 12X DDR Adapter (#1808) that plugs into the system backplane (GX++
slot). There are two GX++ slots in each CEC enclosure. By attaching a 12X to 4X converter
cable (#1828), an IB switch can be attached.
Table 2-26 lists the available InfiniBand adapters.
Table 2-26 Available Fibre Channel adapters
2.8.12 Asynchronous adapter
Asynchronous PCI adapters provide connection of asynchronous EIA-232 or RS-422
devices.
Recent PowerHA releases no longer support heartbeats over serial connections.
Feature
code
support
#1808 GX++ 12X DDR adapter, Dual-port GX++ A, L
#5285a
a. Requires PCIe Gen2 full-high slot
2-Port 4X IB QDR adapter 40 Gb PCIe Full high A, L

Table 2-27 lists the available InfiniBand adapters.
Table 2-27 Available asynchronous adapters
2.9 Internal storage
Serial Attached SCSI (SAS) drives the Power 770 and Power 780 internal disk subsystem.
SAS provides enhancements over parallel SCSI with its point-to-point high frequency
connections. SAS physical links are a set of four wires used as two differential signal pairs.
One differential signal transmits in one direction. The other differential signal transmits in the
opposite direction. Data can be transmitted in both directions simultaneously.
The Power 770 and Power 780 CEC enclosures have an extremely flexible and powerful
backplane for supporting hard disk drives (HDD) or solid-state drives (SSD). The six small
form factor (SFF) bays can be configured in three ways to match your business needs. There
are two integrated SAS controllers that can be optionally augmented with a 175 MB Cache
RAID - Dual IOA Enablement card (Figure 2-22 on page 81). These two controllers provide
redundancy and additional flexibility. The optional 175 MB Cache RAID - Dual IOA
Enablement Card (#5662) enables dual 175 MB write cache and provides dual batteries for
protection of that write cache.
There are two PCIe integrated SAS controllers under the POWER7 I/O chip and also the SAS
controller that is directly connected to the DVD media bay (Figure 2-22 on page 81).
Power 770 and Power 780 supports various internal storage configurations:
Dual split backplane mode: The backplane is configured as two sets of three bays (3/3).
Triple split backplane mode: The backplane is configured as three sets of two
bays (2/2/2).
Dual storage IOA configuration using internal disk drives (Dual RAID of internal drives
only): The backplane is configured as one set of six bays.
Dual storage IOA configuration using internal disk drives and external enclosure (Dual
RAID of internal drives and external drives).
Configuration options will vary depending on the controller options and the operating system
selected. The controllers for the dual split backplane configurations are always the two
embedded controllers. But if the triple split backplane configuration is used, the two integrated
SAS controllers run the first two sets of bays and require a #5901 PCIe SAS adapter located
in a PCIe slot in a CEC enclosure. This adapter controls the third set of bays. By having three
controllers, you can have three boot drives supporting three partitions.
Feature
code
support
#2728 57D1 4-port USB PCIe adapter PCIe Short A, L
#5785 4-Port Asynchronous EIA-232 PCIe
adapter
PCIe Short A, L
#5289 2B42 2-Port Async EIA-232 PCIe adapter PCIe Short A, L

You can configure the two embedded controllers together as a pair for higher redundancy or
you can configure them separately. If you configure them separately, they can be owned by
different partitions or they could be treated independently within the same partition. If
configured as a pair, they provide controller redundancy and can automatically switch over to
the other controller if one has problems. Also, if configured as a pair, both can be active at the
same time (active/active) assuming that there are two or more arrays configured, providing
additional performance capability as well as redundancy. The pair controls all six small form
factor (SFF) bays and both see all six drives. The dual split (3/3) and triple split (2/2/2)
configurations are not used with the paired controllers. RAID 0 and RAID 10 are supported,
and you can also mirror two sets of controller/drives using the operating system.
Power 770 and Power 780, with more than one CEC enclosure, support enclosures with
different internal storage configurations.
Adding the optional 175 MB Cache RAID - Dual IOA Enablement Card (#5662) causes the
pair of embedded controllers in that CEC drawer to be configured as dual controllers,
accessing all six SAS drive bays. With this feature you can get controller redundancy,
additional RAID protection options, and additional I/O performance. RAID 5 (a minimum of
three drives required) and RAID 6 (a minimum of four drives required) are available when
configured as dual controllers with one set of six bays. Feature #5662 plugs in to the disk or
media backplane and enables a 175 MB write cache on each of the two embedded RAID
adapters by providing two rechargeable batteries with associated charger circuitry.
The write cache can provide additional I/O performance for attached disk or solid-state drives,
particularly for RAID 5 and RAID 6. The write cache contents are mirrored for redundancy
between the two RAID adapters, resulting in an effective write cache size of 175 MB. The
batteries provide power to maintain both copies of write-cache information in the event that
power is lost.
Without feature #5662, each controller can access only two or three SAS drive bays.
Another expansion option is an SAS expansion port (#1819). The SAS expansion port can
add more SAS bays to the six bays in the system unit. A # 5886 EXP 12S SAS disk drawer is
attached using a SAS port on the rear of the processor drawer, and its two SAS bays are run
by the pair of embedded controllers. The pair of embedded controllers is now running 18 SAS
bays (six SFF bays in the system unit and twelve 3.5-inch bays in the drawer). The disk
drawer is attached to the SAS port with a SAS YI cable, and the embedded controllers are
connected to the port using a feature #1819 cable assembly. In this 18-bay configuration, all
drives must be HDDs.
IBM i supports configurations using one set of six bays but does not support logically splitting
the backplane into split (dual or triple). Thus, the 175 MB Cache RAID - Dual IOA Enablement
Card (#5662) is required if IBM is to access any of the SAS bays in that CEC enclosure. AIX
and Linux support configurations using two sets of three bays (3/3) or three sets of two bays
Note: These solid-state drives (SSD) or hard disk drive (HDD) configuration rules apply:
You can mix SSD and HDD drives when configured as one set of six bays.
If you want to have both SSDs and HDDs within a dual split configuration, you must use
the same type of drive within each set of three. You cannot mix SSDs and HDDs within
a subset of three bays.
If you want to have both SSDs and HDDs within a triple split configuration, you must
use the same type of drive within each set of two. You cannot mix SSDs and HDDs
within a subset of two bays. The #5901 PCIe SAS adapter that controls the remaining
two bays in a triple split configuration does not support SSDs.

(2/2/2) without feature 5662. With feature #5662, they support dual controllers running one
set of six bays.
Figure 2-22 shows the internal SAS topology overview.
Figure 2-22 Internal SAS topology overview
The system backplane also includes a third embedded controller for running the DVD-RAM
drive in the CEC enclosure. Because the controller is independent from the two other SAS
disk/SSD controllers, it allows the DVD to be switched between multiple partitions without
impacting the assignment of disks or SSDs in the CEC drawer.
Table 2-28 summarizes the internal storage combination and the feature codes required for
any combination.
Table 2-28 SAS configurations summary
SAS
subsystem
configuration
#5662 External SAS
components
SAS port
cables
SAS cables Notes
Two-way split
backplane
No None None N/A IBM i does not support
this combination.
Connecting to an
external disk
enclosure is not
supported.
Three-way
split backplane
No Dual x4 SAS
adapter
(#5901)
Internal SAS
port (#1815)
SAS cable
for three-way
split
backplane
AI cable
(#3679) -
Adapter to
internal drive
(1 meter)
IBM i does not support
this combination.
An I/O adapter can be
located in another
enclosure of the
system.
P7IOC
Integrated
SAS Adapter
P7IOC
DVD
DASD
DASD
DASD
VSES
Redriver
Integrated
SAS Adapter
Integrated
SAS Adapter
Redriver
SAS Port Exp.
A
SAS Port Exp.
B
Optional
Battery
Optional
Battery
External Port

2.9.1 Dual split backplane mode
Dual split backplane mode offers two set of three disks and is the standard configuration. If
desired, one of the sets can be connected to an external SAS PCIe or PCI-X adapter if #1819
is selected. Figure 2-23 shows how the six disk bays are shared with the dual split backplane
mode. Although solid-state drives (SSDs) are supported with a dual split backplane
configuration, mixing SSDs and hard disk drives (HDDs) in the same split domain is not
supported. Also, mirroring SSDs with HDDs is not possible, or vice versa.
Figure 2-23 Dual split backplane overview
Dual storage
IOA with
internal disk
Yes None None N/A Internal SAS port
cable (#1815) cannot
be used with this or
HA RAID
configuration.
Dual storage
IOA with
internal disk
and external
disk enclosure
Yes Requires an
external disk
enclosure
(#5886)
Internal SAS
port (#1819)
SAS cable
assembly for
connecting
to an
externalSAS
drive
enclosure
#3686 or
#3687
#3686 is a 1-meter
cable.
#3687 is a 3-meter
cable.
SAS
subsystem
configuration
#5662 External SAS
components
SAS port
cables
SAS cables Notes
P7IOC
Integrated
SAS Adapter
P7IOC
DVD
DASD
DASD
DASD
VSES
Redriver
Integrated
SAS Adapter
Integrated
SAS Adapter
Redriver
SAS Port Exp.
A
SAS Port Exp.
B
External Port

2.9.2 Triple split backplane
The triple split backplane mode offers three sets of two disk drives each. This mode requires
#1815 internal SAS cable, a SAS cable #3679, and a SAS controller, such as #5901.
Figure 2-24 shows how the six disk bays are shared with the triple split backplane mode. The
PCI adapter that drives two of the six disks can be located in the same Power 770 (or
Power 780) CEC enclosure as the disk drives or adapter, even in a different system enclosure
or external I/O drawer.
Figure 2-24 Triple split backplane overview
Although SSDs are supported with a triple split backplane configuration, mixing SSDs
and HDDs in the same split domain is not supported. Also, mirroring SSDs with HDDs is
not possible.
2.9.3 Dual storage IOA configurations
The dual storage IOA configurations are available with internal or internal with external disk
drives from another I/O drawer. Solid-state drive (SSD) are not supported with this mode.
If #1819 is selected for an enclosure, selecting SAS cable #3686 or #3687 to support RAID
internal and external drives is necessary (Figure 2-25 on page 84). If #1819 is not selected
for the enclosure, the RAID supports only enclosure internal disks.
This configuration increases availability using dual storage IOA or high availability (HA) to
connect multiple adapters to a common set of internal disk drives. It also increases the
performance of RAID arrays. These rules apply to this configuration:
This configuration uses the 175 MB Cache RAID - Dual IOA enablement card.
Using the dual IOA enablement card, the two embedded adapters can connect to each
other and to all six disk drives, as well as the 12 disk drives in an external disk drive
enclosure if one is used.
P7IOC
Integrated
SAS Adapter
P7IOC
DVD
DASD
DASD
DASD
VSES
Redriver
Integrated
SAS Adapter
Integrated
SAS Adapter
Redriver
SAS Port Exp.
A
SAS Port Exp.
B
External Port
4-Port SAS Adapter

The disk drives are required to be in RAID arrays.
There are no separate SAS cables required to connect the two embedded SAS RAID
adapters to each other. The connection is contained within the backplane.
RAID 0, 10, 5, and 6 support up to six drives.
Solid-state drives (SSD) and HDDs can be used, but can never be mixed in the same
disk enclosure.
To connect to the external storage, you need to connect to the #5886 disk drive enclosure.
Figure 2-25 shows the topology of the RAID mode.
Figure 2-25 RAID mode (external disk drives option)
2.9.4 DVD
The DVD media bay is directly connected to the integrated SAS controller on the I/O
backplane and has a specific chip (VSES) for controlling the DVD LED and power. The VSES
appears as a separate device to the device driver and operating systems (Figure 2-22 on
page 81).
Because the integrated SAS controller is independent from the two SAS disk/SSD controllers,
it allows the DVD to be switched between multiple partitions without impacting the
assignment of disks or SSDs in the CEC enclosure.
P7IOC
Integrated
SAS Adapter
P7IOC
DVD
DASD
DASD
DASD
VSES
Redriver
Integrated
SAS Adapter
Integrated
SAS Adapter
Redriver
SAS Port Exp.
A
SAS Port Exp.
B
Optional
Battery
Optional
Battery
External Port

2.10 External I/O subsystems
This section describes the external 12X I/O subsystems that can be attached to the
Power 770 and Power 780, listed as follows:
PCI-DDR 12X Expansion Drawer (#5796)
12X I/O Drawer PCIe, small form factor (SFF) disk (#5802)
12X I/O Drawer PCIe, No Disk (#5877)
Table 2-29 provides an overview of all the supported I/O drawers.
Table 2-29 I/O drawer capabilities
The two GX++ buses from the second processor card feed two GX++ Adapter slots. An
optional GX++ 12X DDR Adapter, Dual-port (#1808), which is installed in GX++ Adapter slot,
enables the attachment of a 12X loop, which runs at either SDR or DDR speed depending on
the 12X I/ O drawers that are attached.
2.10.1 PCI-DDR 12X Expansion drawer
The PCI-DDR 12X Expansion Drawer (#5796) is a 4U (EIA units) drawer and mounts in a
19-inch rack. Feature #5796 is 224 mm (8.8 in.) wide and takes up half the width of the 4U
(EIA units) rack space. The 4U enclosure can hold up to two #5796 drawers mounted
side-by-side in the enclosure. The drawer is 800 mm (31.5 in.) deep and can weigh up to
20 kg (44 lb).
The PCI-DDR 12X Expansion Drawer has six 64-bit, 3.3 V, PCI-X DDR slots, running at
266 MHz, that use blind-swap cassettes and support hot-plugging of adapter cards. The
drawer includes redundant hot-plug power and cooling.
Two interface adapters are available for use in the #5796 drawer:
Dual-Port 12X Channel Attach Adapter Long Run (#6457)
Dual-Port 12X Channel Attach Adapter Short Run (#6446)
The adapter selection is based on how close the host system or the next I/O drawer in the
loop is physically located. Feature #5796 attaches to a host system CEC enclosure with a
12X adapter in a GX++ slot through SDR or DDR cables (or both SDR and DDR cables). A
maximum of four #5796 drawers can be placed on the same 12X loop. Mixing #5802/5877
and #5796 on the same loop is not supported.
A minimum configuration of two 12X cables (either SDR or DDR), two AC power cables, and
two SPCN cables is required to ensure proper redundancy.
Drawer
feature code
DASD PCI slots Requirements for the Power 770
and Power 780
#5796 - 6 x PCI-X GX++ adapter card #1808
#5802 18 x SAS disk drive bays 10 x PCIe GX++ adapter card #1808
#5877 - 10 x PCIe GX++ adapter card #1808

Figure 2-26 shows the back view of the expansion unit.
Figure 2-26 PCI-X DDR 12X Expansion Drawer rear side
2.10.2 12X I/O Drawer PCIe
The 12X I/O Drawer PCIe is a 19-inch I/O and storage drawer. It provides a 4U-tall (EIA units)
drawer containing 10 PCIe-based I/O adapter slots and 18 SAS hot-swap Small Form Factor
disk bays, which can be used for either disk drives or SSD (#5802). The adapter slots use
blind-swap cassettes and supports hot-plugging of adapter cards.
A maximum of two #5802 drawers can be placed on the same 12X loop. Feature #5877 is the
same as #5802 except that it does not support any disk bays. Feature #5877 can be on the
same loop as #5802. Feature #5877 cannot be upgraded to #5802.
The physical dimensions of the drawer are 444.5 mm (17.5 in.) wide by 177.8 mm (7.0 in.)
high by 711.2 mm (28.0 in.) deep for use in a 19-inch rack.
A minimum configuration of two 12X DDR cables, two AC power cables, and two SPCN
cables is required to ensure proper redundancy. The drawer attaches to the host CEC
enclosure with a 12X adapter in a GX++ slot through 12X DDR cables that are available in
various cable lengths:
0.6 m (#1861)
1.5 m (#1862)
3.0 m (#1865)
8 m (#1864)
The 12X SDR cables are not supported.
12X Port 1
(P1-C7-T2)
P1-C8-T3 P1-C1
P1-C2
P1-C3 P1-C4
P1-C5
P1-C6
12X Port 0
(P1-C7-T1)E1
E2
SPCN 0
(P1-C8-T1)
SPCN 1
(P1-C8-T2)

Figure 2-27 shows the front view of the 12X I/O Drawer PCIe (#5802).
Figure 2-27 Front view of the 12X I/O Drawer PCIe
Figure 2-28 shows the rear view of the 12X I/O Drawer PCIe (#5802).
Figure 2-28 Rear view of the 12X I/O Drawer PCIe
2.10.3 Dividing SFF drive bays in 12X I/O drawer PCIe
Disk drive bays in the 12X I/O drawer PCIe can be configured as one, two, or four sets. This
allows for partitioning of disk bays. Disk bay partitioning configuration can be done via the
physical mode switch on the I/O drawer.
Power cables
Disk drives
Port cardsService card
10 PCIe cards X2 SAS connectors
SPCN Connectors
12X Connectors Mode Switch
Note: A mode change using the physical mode switch requires power-off/on of the drawer.

Figure 2-29 indicates the mode switch in the rear view of the #5802 I/O Drawer.
Figure 2-29 Disk bay partitioning in #5802 PCIe 12X I/O drawer
Each disk bay set can be attached to its own controller or adapter. The #5802 PCIe 12X I/O
Drawer has four SAS connections to drive bays. It can connect to a PCIe SAS adapter or to
controllers on the host system.
Figure 2-29 shows the configuration rule of disk bay partitioning in the #5802 PCIe 12X I/O
Drawer. There is no specific feature code for mode switch setting.
The SAS ports, as associated with the mode selector switch map to the disk bays, have the
mappings shown in Table 2-30.
Table 2-30 SAS connection mappings
Note: The IBM System Planning Tool supports disk bay partitioning. Also, the IBM
configuration tool accepts this configuration from IBM System Planning Tool and passes it
through IBM manufacturing using the Customer Specified Placement (CSP) option.
Location code Mappings Number of bays
P4-T1 P3-D1 to P3-D5 5 bays
MODE
SWITCH
1
2
4
#5802 12X I/O Drawer
AIX/Linux
• One set: 18 bays
• Two sets: 9 + 9 bays
• Four sets: 5 + 4 + 5 + 4 bays
IBMi
• Two sets: 9 + 9 bays
PCIe 12X I/O Drawer – SFF Drive Bays

The location codes for the front and rear views of the #5802 I/O drawer are provided in
Figure 2-30 and Figure 2-31.
Figure 2-30 #5802 I/O drawer from view location codes
Figure 2-31 #5802 I/O drawer rear view location codes
Configuring the #5802 disk drive subsystem
The #5802 SAS disk drive enclosure can hold up 18 disk drives. The disks in this enclosure
can be organized in several configurations depending on the operating system used, the type
of SAS adapter card, and the position of the mode switch.
P3-D1
P3-D2
P3-D3
P3-D4
P3-D5
P3-D6
P3-D7
P3-C1
P3-C2
P3-D8
P3-D9
P3-D10
P3-D11
P3-C3
P3-C4
P3-D12
P3-D13
P3-D14
P3-D15
P3-D16
P3-D17
P3-D18
E1
E2
ARECW500-0
P1-C1
P1-C2
P1-C3
P1-C4
P1-T2
P1-C5
P1-C6
P1-C7
P1-C8
P1-C9
P1-C10
P4-T5
P2-T1
P2-T2
P2-T3
ARECW501-0
P4-T1
P4-T2
P4-T3
P4-T4
P1-T1

Each disk bay set can be attached to its own controller or adapter. The feature #5802 PCIe
12X I/O Drawer has four SAS connections to drive bays. It connects to PCIe SAS adapters or
controllers on the host systems.
For detailed information about how to configure, see the IBM Power Systems Hardware
Information Center:
http://publib.boulder.ibm.com/infocenter/powersys/v3r1m5/index.jsp
2.10.4 12X I/O Drawer PCIe and PCI-DDR 12X Expansion Drawer 12X cabling
I/O Drawers are connected to the adapters in the CEC enclosure with data transfer cables:
12X DDR cables for the #5802 and #5877 I/O drawers
12X SDR and/or DDR cables for the #5796 I/O drawers
The first 12X I/O Drawer that is attached in any I/O drawer loop requires two data transfer
cables. Each additional drawer, up to the maximum allowed in the loop, requires one
additional data transfer cable. Note the following information:
A 12X I/O loop starts at a CEC bus adapter port 0 and attaches to port 0 of an I/O drawer.
The I/O drawer attaches from port 1 of the current unit to port 0 of the next I/O drawer.
Port 1 of the last I/O drawer on the 12X I/O loop connects to port 1 of the same CEC bus
adapter to complete the loop.
Figure 2-32 shows typical 12X I/O loop port connections.
Figure 2-32 Typical 12X I/O loop port connections
Table 2-31 shows various 12X cables to satisfy the various length requirements.
Table 2-31 12X connection cables
Feature code Description
#1861 0.6-meter 12X DDR cable
I/O I/O
1 0
I/O
1 0
I/O
10
I/O
10
C
0
1

General rule for the 12X IO Drawer configuration
To optimize performance and distribute workload, use as many multiple GX++ buses as
possible. Figure 2-33 shows several examples of a 12X IO Drawer configuration.
Figure 2-33 12X IO Drawer configuration
Supported 12X cable length for PCI-DDR 12X Expansion Drawer
Each #5796 drawer requires one Dual Port PCI DDR 12X Channel Adapter, either Short Run
(#6446) or Long Run (#6457). The choice of adapters is dependent on the distance to the
next 12X Channel connection in the loop, either to another I/O drawer or to the system unit.
Table 2-32 identifies the supported cable lengths for each 12X channel adapter. I/O drawers
containing the short range adapter can be mixed in a single loop with I/O drawers containing
the long range adapter. In Table 2-32, a Yes indicates that the 12X cable identified in that
column can be used to connect the drawer configuration identified to the left. A No means that
it cannot be used.
Table 2-32 Supported 12X cable lengths
2.10.5 12X I/O Drawer PCIe and PCI-DDR 12X Expansion Drawer SPCN cabling
System Power Control Network (SPCN) is used to control and monitor the status of power
and cooling within the I/O drawer.
SPCN cables connect all ac-powered expansion units (Figure 2-34 on page 92):
1. Start at SPCN 0 (T1) of the first (top) CEC enclosure to J15 (T1) of the first expansion unit.
2. Cable all units from J16 (T2) of the previous unit to J15 (T1) of the next unit.
Connection type 12X cable options
0.6 m 1.5 m 3.0 m 8.0 m
#5796 to #5796 with #6446 in both drawers Yes Yes No No
#5796 with #6446 adapter to #5796 with #6457 adapter Yes Yes Yes No
#5796 to #5796 with #6457 adapter in both drawers Yes Yes Yes Yes
#5796 with #6446 adapter to system unit No Yes Yes No
#5796 with #6457 adapter to system unit No Yes Yes Yes
PCIe
770
780
PCIe
770
780
PCIe
770
780
770
780
770
780 PCI-X
PCIe
770
780 PCIe
PCIe
770
780 PCIe
PCIe
770
780 PCIe
One PCIe I/O Drawer
Two PCIe I/O Drawers
Three PCIe I/O Drawer
Four PCIe I/O Drawer

3. From J16 (T2) of the final expansion unit, connect to the second CEC enclosure,
SPCN 1 (T2).
4. To complete the cabling loop, connect SPCN 1 (T2) of the topmost (first) CEC enclosure
to the SPCN 0 (T1) of the next (second) CEC.
5. Ensure that a complete loop exists from the topmost CEC enclosure, through all attached
expansions and back to the next lower (second) CEC enclosure.
Figure 2-34 SPCN cabling examples
Table 2-33 shows the SPCN cables to satisfy various length requirements.
Table 2-33 SPCN cables
2.11 External disk subsystems
This section describes the following external disk subsystems that can be attached to the
Power 770 and Power 780:
EXP 12S SAS Expansion Drawer (#5886) (supported, but no longer orderable)
EXP24S SFF Gen2-bay Drawer for high-density storage (#5887)
TotalStorage EXP24 Disk Drawer (#5786)
Note: Only the first two CEC enclosures of a multi-CEC system are included in
SPCN cabling with I/O expansion units. CEC enclosures number three and four are
not connected.
Feature code Description
#6006 SPCN cable drawer-to-drawer, 3 m
#6007 SPCN cable rack-to-rack, 15 m
SPCN Connections
I/O Units
I/O
Expansion
Unit
*J15(T1)
*J16(T2)
I/O
Expansion
Unit
*J15(T1)
*J16(T2)
I/O
Expansion
Unit
*J15(T1)
*J16(T2)
I/O
Expansion
Unit
*J15(T1)
*J16(T2)
Secondary 3 CEC
No connection to third
or fourth CEC
Secondary 2 CEC
No connection to third
or fourth CEC
P1-C1-T8 (SPCN0)
Secondary 1 CEC
P1-C1-T9 (SPCN1)
P1-C1-T8 (SPCN0)
Primary CEC
P1-C1-T9 (SPCN1)

IBM 7031 TotalStorage EXP24 Ultra320 SCSI Expandable Storage Disk Enclosure
(no longer orderable)
IBM System Storage
Table 2-29 on page 85 provides an overview of SAS external disks subsystems.
Table 2-34 I/O drawer capabilities
2.11.1 EXP 12S Expansion Drawer
The EXP 12S (#5886) is an expansion drawer with twelve 3.5-inch form factor SAS bays.
#5886 supports up to 12 hot-swap SAS HDDs or up to eight hot-swap SSDs. The EXP 12S
includes redundant ac power supplies and two power cords. Though the drawer is one set of
12 drives, which is run by one SAS controller or one pair of SAS controllers, it has two SAS
attachment ports and two service managers for redundancy. The EXP 12S takes up a 2 EIA
space in a 19-inch rack. The SAS controller can be a SAS PCI-X or PCIe adapter or pair
of adapters.
The drawer can either be attached using the backplane, providing an external SAS port, or
using one of the following adapters:
PCIe 380 MB Cache Dual -x4 3 Gb SAS RAID adapter (#5805)
PCI-X DDR Dual -x4 SAS adapter (#5900 is supported but no longer orderable)
PCIe Dual -x4 SAS adapter (#5901)
PCIe 380 MB Cache Dual -x4 3 Gb SAS RAID adapter (#5903 is supported but no longer
orderable)
PCI-X DDR 1.5 GB Cache SAS RAID adapter (#5904)
PCI-X DDR Dual -x4 SAS adapter (#5912)
PCIe2 1.8 GB Cache RAID SAS Adapter (#5913)
The SAS disk drives or SSD contained in the EXP 12S Expansion Drawer are controlled by
one or two PCIe or PCI-X SAS adapters connected to the EXP 12S Expansion Drawer
through SAS cables. The SAS cable varies, depending on the adapter being used, the
operating system being used, and the protection desired.
The large cache PCI-X DDR 1.5 GB Cache SAS RAID Adapter (#5904) and PCI-X DDR
1.5 GB Cache SAS RAID Adapter (BSC) (#5908) uses a SAS Y cable when a single port is
running the EXP 12S Expansion Drawer. A SAS X cable is used when a pair of adapters is
used for controller redundancy.
The medium cache PCIe 380 MB Cache Dual - x4 3 Gb SAS RAID Adapter (#5903) is always
paired and uses a SAS X cable to attach the feature #5886 I/O drawer.
The zero cache PCI-X DDR Dual - x4 SAS Adapter (#5912) and PCIe Dual-x4 SAS Adapter
(#5901) use a SAS Y cable when a single port is running the EXP 12S Expansion Drawer. A
SAS X cable is used for AIX or Linux environments when a pair of adapters is used for
controller redundancy.
Drawer
feature code
DASD PCI slots Requirements for a
Power 770 and Power 780
#5886 12 x SAS disk drive bays - Any supported SAS adapter
#5887 24x - Any supported SAS adapter

The following SAS X cables are available for usage with a PCIe2 1.8 GB Cache RAID SAS
adapter (#5913):
3 meters (#3454)
6 meters (#3455)
10 meters (#3456)
In all of these configurations, all 12 SAS bays are controlled by a single controller or a single
pair of controllers.
A second EXP 12S Expansion Drawer can be attached to another drawer by using two SAS
EE cables, providing 24 SAS bays instead of 12 bays for the same SAS controller port. This
configuration is called cascading. In this configuration, all 24 SAS bays are controlled by a
single controller or a single pair of controllers.
There is a maximum of up to 110 EXP 12S Expansion Drawer on SAS PCI controllers.
The #5886 can be directly attached to the SAS port on the rear of the Power 770 and 780,
providing a very low-cost disk storage solution.
Adding the optional 175 MB Cache RAID - Dual IOA Enablement Card (#5662) to the
Power 770 and Power 780 causes the pair of embedded controllers in that processor
enclosure to be configured as dual controllers, accessing all six SAS bays. Using the internal
SAS Cable Assembly for SAS Port (#1819) connected to the rear port, the pair of embedded
controllers is now running 18 SAS bays (six SFF bays in the system unit and twelve 3.5-inch
bays in the drawer). The disk drawer is attached to the SAS port with a SAS YI cable. In this
18-bay configuration, all drives must be HDD.
A second unit cannot be cascaded to a EXP 12S Expansion Drawer attached in this way.
Various disk options are available to be installed in the EXP 12S Expansion Drawer.
Table 2-35 shows the available disk drive feature codes.
Table 2-35 Disk options for the EXP 12S drawer
For detailed information about the SAS cabling, see the Serial-attached SCSI cable
planning documentation:
http://publib.boulder.ibm.com/infocenter/powersys/v3r1m5/index.jsp?topic=/p7had/p7
hadsascabling.htm
#3586 69 GB 3.5" SAS Solid State Drive AIX, Linux
#3646 73.4 GB 15K RPM SAS Disk Drive AIX. Linux
#3647 146.8 GB 15K RPM SAS Disk Drive AIX, Linux
#3648 300 GB 15 K RPM SAS Disk Drive AIX, Linux
#3649 450 GB 15 K RPM SAS Disk Drive AIX, Linux
#3587 69 GB 3.5” SAS Solid State Drive IBM i
#3676 69.7 GB 15 K RPM SAS Disk Drive IBM i

2.11.2 EXP24S SFF Gen2-bay Drawer
The EXP24S SFF Gen2-bay Drawer (#5887) is an expansion drawer supporting up to 24
hot-swap 2.5-inch SFF SAS HDDs on POWER6 or POWER7 servers in 2U of 19-inch
rack space.
The SFF bays of the EXP24S are different from the SFF bays of the POWER7 system units
or 12X PCIe I/O Drawers (#5802, #5803). The EXP 24S uses Gen-2 or SFF-2 SAS drives
that physically do not fit in the Gen-1 or SFF-1 bays of the POWER7 system unit or 12X PCIe
I/O Drawers, or vice versa.
The EXP24S SAS ports are attached to SAS controllers, which can be a SAS PCI-X or PCIe
adapter or pair of adapters. The EXP24S SFF Gen2-bay Drawer can also be attached to an
imbedded SAS controller in a server with an imbedded SAS port. Attachment between the
SAS controller and the EXP24S SAS ports is via the appropriate SAS Y or X cables.
The SAS adapters/controllers that support the EXP24S are:
PCI-X 1.5 GB Cache SAS RAID Adapter 3 Gb (#5904, #5906, #5908)
PCIe 380 MB Cache SAS RAID Adapter 3 Gb (#5805, #5903)
PCIe Dual-x4 SAS Adapter 3 Gb (#5901, #5278)
PCIe2 1.8GB Cache RAID SAS Adapter (#5913)
The SAS disk drives contained in the EXP24S SFF Gen2-bay Drawer are controlled by one
or two PCIe or PCI-X SAS adapters connected to the EXP24S through SAS cables. The SAS
cable varies, depending on the adapter being used, the operating system being used, and the
protection desired.
Adding the optional 175 MB Cache RAID - Dual IOA Enablement Card (#5662) to the
Power 770 and Power 780 causes the pair of embedded controllers in that processor
enclosure to be configured as dual controllers, accessing all six SAS bays. Using the internal
SAS Cable Assembly for SAS Port (#1819) connected to the rear port, the pair of embedded
Note: The following consideration should be applied:
The large cache PCI-X DDR 1.5 GB Cache SAS RAID Adapter (#5904) and PCI-X
DDR 1.5 GB Cache SAS RAID Adapter (BSC) (#5908) use an SAS Y cable when a
single port is running the EXP24S. A SAS X cable is used when a pair of adapters is
used for controller redundancy.
The medium cache PCIe 380 MB Cache Dual - x4 3 Gb SAS RAID Adapter (# 5903) is
always paired and uses a SAS X cable to attach the feature #5887 I/O drawer.
The zero cache PCI-X DDR Dual - x4 SAS Adapter (#5912) and PCIe Dual-x4 SAS
Adapter (#5901) use a SAS Y cable when a single port is running the EXP24S. A
SAS X cable is used for AIX or Linux environments when a pair of adapters is used for
controller redundancy.
The PCIe Gen2 1.8GB Cache RAID SAS Adapter (#5913) uses SAS YO cables.
In all of these configurations, all 24 SAS bays are controlled by a single controller or a
single pair of controllers.
A second EXP24S drawer can be attached to another drawer by using two SAS EE
cables, providing 48 SAS bays instead of 24 bays for the same SAS controller port. This
configuration is called cascading. In this configuration, all 48 SAS bays are controlled
by a single controller or a single pair of controllers.
The EXP24S SFF Gen2-bay Drawer can be directly attached to the SAS port on the
rear of the Power 770 and Power 780, providing a very low-cost disk storage solution.

controllers is now running 30 SAS bays (six SFF bays in the system unit and twenty-four
2.5-inch bays in the drawer). The disk drawer is attached to the SAS port with a SAS YI cable.
In this 30-bay configuration, all drives must be HDD.
A second unit cannot be cascaded to a EXP24S SFF Gen2-bay Drawer attached in this way.
The EXP24S SFF Gen2-bay Drawer can be ordered in one of three possible
manufacturing-configured MODE settings (not customer set-up) of 1, 2 or 4 sets of disk bays.
With IBM AIX/Linux/VIOS, the EXP 24S can be ordered with four sets of six bays (mode4),
two sets of 12 bays (mode 2), or one set of 24 bays (mode 1). With IBM i, the EXP24S can be
ordered as one set of 24 bays (mode 1).
There are six SAS connectors on the rear of the XP24S SFF Gen2-bay Drawer to which SAS
adapters/controllers are attached. They are labeled T1, T2, and T3, and there are two T1, two
T2, and two T3 (Figure 2-35).
In mode 1, two or four of the six ports are used. Two T2 are used for a single SAS
adapter, and two T2 and two T3 are used with a paired set of two adapters or dual
adapters configuration.
In mode 2 or mode 4, four ports will be used, two T2 and two T3, to access all SAS bays.
Figure 2-35 EXP24S SFF Gen2-bay Drawer rear connectors
An EXP24S SFF Gen2-bay Drawer in mode 4 can be attached to two or four SAS
controllers and provide a great deal of configuration flexibility. An EXP24S in mode 2 has
similar flexibility. Up to 24 HDDs can be supported with any of the supported SAS
adapters/controllers.
Note: Note the following information:
The modes for the EXP24S SFF Gen2-bay Drawer are set by IBM Manufacturing.
There is no option to reset after the drawer has been shipped.
If you order multiple EXP24S, avoid mixing modes within that order. There is no
externally visible indicator as to the drawer's mode.
Several EXP24S cannot be cascaded on the external SAS connector. Only one #5887
is supported.
The Power 770 or Power 780 supports up to 56 XP24S SFF Gen2-bay Drawers.

Include the EXP24S SFF Gen2-bay Drawer no-charge specify codes with EXP24S
orders to indicate to IBM Manufacturing the mode to which the drawer should be set and
the adapter/controller/cable configuration that will be used. Table 2-36 lists the no-charge
specify codes and the physical adapters/controllers/cables with their own chargeable
feature numbers.
Table 2-36 EXP 24S Cabling
These cabling options for the EXP 24S Drawer are available:
X cables for #5278
– 3 m (#3661)
– 6 m (#3662)’
– 15 m (#3663)
X cables for #5913 (all 6 Gb except for 15 m cable)
– 3 m (#3454)
– 6 m (#3455)
– 10 m (#3456)
YO cables for #5278
– 1.5 m (#3691)
– 3 m (#3692)
– 6 m (#3693)
– 15 m (#3694)
YO cables for #5913 (all 6 Gb except for 15 m cable)
– 1.5 m (#3450)
– 3 m (#3451)
– 6 m (#3452)
– 10 m (#3453)
YI cables for system unit SAS port (3 Gb)
– 1.5 m (#3686)
– 3 m (#3687)
Feature code Mode Adapter/controller Cable to drawer Environment
#9360 1 Pair #5901 2 YO Cables A, L, VIOS
#9361 2 Two #5901 2 YO Cables A, L, VIOS
#9365 4 Four #5901 2 X Cable A, L, VIOS
#9366 2 Two pair #5901 2 X Cables A, L, VIOS
#9367 1 Pair #5903, #5805 2 YO Cables A, i, L, VIOS
#9368 2 Four #5903, #5805 2 X Cables A, L, VIOS
#9382 1 One #5904/06/08 1 YO Cable A, i, L, VIOS
#9383 1 Pair #5904/06/08 2 YO Cables A, i, L, VIOS
#9384 1 CEC SAS port 1 YI Cable A, i, L, VIOS
#9385 1 Two #5913 2 YO Cables A, i, L, VIOS
#9386 2 Four #5913 4 X Cables A, L, VIOS

For detailed information about the SAS cabling, see the serial-attached SCSI cable
planning documentation:
http://publib.boulder.ibm.com/infocenter/powersys/v3r1m5/index.jsp?topic=/p7had/p7
hadsascabling.htm
2.11.3 TotalStorage EXP24 disk drawer and tower
The TotalStorage EXP24 is available as a 4 EIA unit drawer and mounts in a 19-inch rack
(#5786). The front of the IBM TotalStorage EXP24 Ultra320 SCSI Expandable Storage Disk
Enclosure has bays for up to 12 disk drives organized in two SCSI groups of up to six drives.
The rear also has bays for up to 12 disk drives organized in two additional SCSI groups of up
to six drives, plus slots for the four SCSI interface cards. Each SCSI drive group can be
connected by either a Single Bus Ultra320 SCSI Repeater Card (#5741) or a Dual Bus
Ultra320 SCSI Repeater Card (#5742). This allows the EXP24 to be configured as four sets
of six bays, two sets of 12 bays, or two sets of six bays plus one set of 12 bays.
The EXP24 features #5786 and #5787 have three cooling fans and two power supplies to
provide redundant power and cooling. The SCSI disk drives contained in the EXP24 are
controlled by PCI-X SCSI adapters connected to the EXP24 SCSI repeater cards by SCSI
cables. The PCI-X adapters are located in the Power 740 system unit or in an attached I/O
drawer with PCI-X slots.
The 336 system maximum is achieved with a maximum of 24 disks in a maximum of 14
TotalStorage EXP24 disk drawers (#5786) or 14 TotalStorage EXP24 disk towers (#5787).
2.11.4 IBM TotalStorage EXP24
The IBM 7031 TotalStorage EXP24 Ultra320 SCSI Expandable Storage Disk Enclosure
supports up to 24 Ultra320 SCSI Disk Drives arranged in four independent SCSI groups of up
to six drives or in two groups of up to 12 drives. Each SCSI drive group can be connected by
either a Single Bus Ultra320 SCSI Repeater Card or a Dual Bus Ultra320 SCSI Repeater
Card, allowing a maximum of eight SCSI connections per TotalStorage EXP24.
The IBM 7031 Model D24 (7031-D24) is an Expandable Disk Storage Enclosure that is a
horizontal 4 EIA by 19-inch rack drawer for mounting in equipment racks.
The IBM 7031 Model T24 (7031-T24) is an Expandable Disk Storage Enclosure that is a
vertical tower for floor-standing applications.
Note: IBM plans to offer a 15-meter, 3 Gb bandwidth SAS cable for the #5913 PCIe2
1.8 GB Cache RAID SAS Adapter when attaching the EXP24S Drawer (#5887) for large
configurations where the 10 meter cable is a distance limitation.
The EXP24S Drawer rails are fixed length and designed to fit Power Systems provided
racks of 28 inches (711 mm) deep. EXP24S uses 2 EIA of space in a 19-inch wide rack.
Other racks might have different depths, and these rails will not adjust. No adjustable
depth rails are orderable at this time.
Note: The EXP24S SCSI disk drawer is an earlier technology drawer compared to the later
SAS EXP12S drawer. It is used to house existing SCSI disk drives that are supported, but
that are no longer orderable.

2.11.5 IBM System Storage
The IBM System Storage Disk Systems products and offerings provide compelling storage
solutions with superior value for all levels of business, from entry-level up to high-end
storage systems.
IBM System Storage N series
The IBM System Storage N series is a Network Attached Storage (NAS) solution and
provides the latest technology to customers to help them improve performance, virtualization
manageability, and system efficiency at a reduced total cost of ownership. For more
information about the IBM System Storage N series hardware and software, see:
http://www.ibm.com/systems/storage/network
IBM System Storage DS3000 family
The IBM System Storage DS3000 is an entry-level storage system designed to meet the
availability and consolidation needs for a wide range of users. New features, including larger
capacity 450 GB SAS drives, increased data protection features such as RAID 6, and more
FlashCopies per volume, provide a reliable virtualization platform. For more information about
the DS3000 family, see:
http://www.ibm.com/systems/storage/disk/ds3000/index.html
IBM System Storage DS5000
New DS5000 enhancements help reduce cost by introducing SSD drives. Also with the new
EXP5060 expansion unit supporting 60 1 TB SATA drives in a 4U package, customers can
see up to a one-third reduction in floor space over standard enclosures. With the addition of
1 Gbps iSCSI host attach, customers can reduce cost for their less demanding applications
while continuing to provide high performance where necessary, utilizing the 8 Gbps FC host
ports. With the DS5000 family, you get consistent performance from a smarter design that
simplifies your infrastructure, improves your TCO, and reduces your cost. For more
information about the DS5000 family, see:
IBM Storwize V7000 Midrange Disk System
IBM Storwize® V7000 is a virtualized storage system to complement virtualized server
environments that provides unmatched performance, availability, advanced functions, and
highly scalable capacity never seen before in midrange disk systems. Storwize V7000 is a
powerful midrange disk system that has been designed to be easy to use and enable rapid
deployment without additional resources. Storwize V7000 is virtual storage that offers greater
efficiency and flexibility through built-in solid state drive (SSD) optimization and thin
provisioning technologies. Storwize V7000 advanced functions also enable non-disruptive
migration of data from existing storage, simplifying implementation and minimizing disruption
to users. Storwize V7000 also enables you to virtualize and reuse existing disk systems,
Note: A new IBM 7031 TotalStorage EXP24 Ultra320 SCSI Expandable Storage Disk
Enclosure cannot be ordered for the Power 720 and Power 740, and thus only existing
7031-D24 drawers or 7031-T24 towers can be moved to the Power 720 and 740 servers.
AIX and Linux partitions are supported along with the usage of a IBM 7031 TotalStorage
EXP24 Ultra320 SCSI Expandable Storage Disk Enclosure.

supporting a greater potential return on investment (ROI). For more information about
Storwize V7000, see:
http://www.ibm.com/systems/storage/disk/storwize_v7000/index.html
IBM XIV Storage System
IBM offers a mid-sized configuration of its self-optimizing, self-healing, resilient disk solution,
the IBM XIV® Storage System, storage reinvented for a new era. Now, organizations with
mid-size capacity requirements can take advantage of the latest IBM technology for their
most demanding applications with as little as 27 TB of usable capacity and incremental
upgrades. For more information about XIV, see:
http://www.ibm.com/systems/storage/disk/xiv/index.html
IBM System Storage DS8000
The IBM System Storage DS8000® family is designed to offer high availability, multiplatform
support, and simplified management tools. With its high capacity, scalability, broad server
support, and virtualization features, the DS8000 family is well suited for simplifying the
storage environment by consolidating data from multiple storage systems on a single system.
The high-end model DS8800 is the most advanced model in the IBM DS8000 family lineup
and introduces new dual IBM POWER6-based controllers that usher in a new level of
performance for the company’s flagship enterprise disk platform. The DS8800 offers twice the
maximum physical storage capacity than the previous model. For more information about the
DS8000 family, see:
2.12 Hardware Management Console
The Hardware Management Console (HMC) is a dedicated workstation that provides a
graphical user interface (GUI) for configuring, operating, and performing basic system tasks
for the POWER7 processor-based systems (and the POWER5, POWER5+, POWER6, and
POWER6+ processor-based systems) that function in either non-partitioned or clustered
environments. In addition, the HMC is used to configure and manage partitions. One HMC is
capable of controlling multiple POWER5, POWER5+, POWER6, and POWER6+, and
POWER7 processor-based systems.
Several HMC models are supported to manage POWER7 processor-based systems. Two
models (7042-C08, 7042-CR6) are available for ordering at the time of writing, but you can
also use one of the withdrawn models listed in Table 2-37.
Table 2-37 HMC models supporting POWER7 processor technology-based servers
7310-C05 Withdrawn IBM 7310 Model C05 Desktop Hardware Management Console
7042-C08 Available IBM 7042 Model C08 Deskside Hardware Management Console
7310-CR3 Withdrawn IBM 7310 Model CR3 Rack-mounted Hardware Management Console

At the time of writing, the HMC must be running V7R7.4.0. It can also support up to 48
Power7 systems. Updates of the machine code, HMC functions, and hardware prerequisites,
can be found on the Fix Central website:
http://www-933.ibm.com/support/fixcentral/
2.12.1 HMC functional overview
The HMC provides three groups of functions:
Server
Virtualization
HMC management
Server management
The first group contains all functions related to the management of the physical servers under
the control of the HMC:
System password
Status Bar
Power On/Off
Capacity on Demand
Error management
– System indicators
– Error and event collection reporting
– Dump collection reporting
– Call Home
– Customer notification
– Hardware replacement (Guided Repair)
– SNMP events
Concurrent Add/Repair/Upgrade
Redundant Service Processor
Firmware Updates
Virtualization management
The second group contains all of the functions related to virtualization features, such as a
partition configuration or the dynamic reconfiguration of resources:
System Plans
System Profiles
Partitions (create, activate, shutdown)
Profiles
Partition Mobility
DLPAR (processors, memory, I/O, and so on)
Custom Groups
7042-CR6 Available IBM 7042 Model CR6 Rack-mounted Hardware Management Console
Note: An HMC is a mandatory requirement for the both the Power7 770 and 780 systems,
but it is possible to share an HMC with other Power systems.

HMC Console management
The last group relates to the management of the HMC itself, its maintenance, security, and
configuration, for example:
Guided set-up wizard
Electronic Service Agent set up wizard
User Management
– User IDs
– Authorization levels
– Customizable authorization
Disconnect and reconnect
Network Security
– Remote operation enable and disable
– User definable SSL certificates
Console logging
HMC Redundancy
Scheduled Operations
Back-up and Restore
Updates, Upgrades
Customizable Message of the day
The HMC provides both a graphical interface and command-line interface (CLI) for all
management tasks. Remote connection to the HMC using a web browser (as of HMC
Version 7; previous versions required a special client program called WebSM) is possible.
The CLI is also available by using the Secure Shell (SSH) connection to the HMC. It can
be used by an external management system or a partition to remotely perform many
HMC operations.
2.12.2 HMC connectivity to the POWER7 processor-based systems
POWER5, POWER5+, POWER6, POWER6+, and POWER7 processor technology-based
servers that are managed by an HMC require Ethernet connectivity between the HMC and
the server’s Service Processor. In addition, if Dynamic LPAR, Live Partition Mobility, or
PowerVM Active Memory Sharing operations are required on the managed partitions,
Ethernet connectivity is needed between these partitions and the HMC. A minimum of two
Ethernet ports are needed on the HMC to provide such connectivity. The rack-mounted
7042-CR5 HMC default configuration provides four Ethernet ports. The deskside 7042-C07
HMC standard configuration offers only one Ethernet port. Be sure to order an optional PCIe
adapter to provide additional Ethernet ports.
For any logical partition in a server it is possible to use a Shared Ethernet Adapter that is
configured via a Virtual I/O Server. Therefore, a partition does not require its own physical
adapter to communicate with an HMC.

For the HMC to communicate properly with the managed server, eth0 of the HMC must be
connected to either the HMC1 or HMC2 ports of the managed server, although other network
configurations are possible. You can attach a second HMC to HMC Port 2 of the server for
redundancy (or vice versa). These must be addressed by two separate subnets. Figure 2-36
shows a simple network configuration to enable the connection from HMC to server and to
enable Dynamic LPAR operations. For more details about HMC and the possible network
connections, see Hardware Management Console V7 Handbook, SG24-7491.
Figure 2-36 HMC to service processor and LPARs network connection
The default mechanism for allocation of the IP addresses for the service processor HMC
ports is dynamic. The HMC can be configured as a DHCP server, providing the IP address at
the time the managed server is powered on. In this case, the FSPs are allocated IP address
from a set of address ranges predefined in the HMC software. These predefined ranges are
identical for Version 710 of the HMC code and for previous versions.
If the service processor of the managed server does not receive a DHCP reply before time
out, predefined IP addresses will be set up on both ports. Static IP address allocation is also
an option. You can configure the IP address of the service processor ports with a static IP
address by using the Advanced System Management Interface (ASMI) menus.
Power System
LPAR
n
LPAR
...
LPAR
2
LPAR
1
entxentxentxent1
Service
Processor
HMC 2
eth0
HMC 1eth1
Management LAN

2.12.3 High availability using the HMC
The HMC is an important hardware component. When in operation, POWER7
processor-based servers and their hosted partitions can continue to operate when no HMC is
available. However, in such conditions, certain operations cannot be performed, such as a
DLPAR reconfiguration, a partition migration using PowerVM Live Partition Mobility, or the
creation of a new partition. You might therefore decide to install two HMCs in a redundant
configuration so that one HMC is always operational, even when performing maintenance of
the other one, for example.
If redundant HMC function is desired, a server can be attached to two independent HMCs to
address availability requirements. Both HMCs must have the same level of Hardware
Management Console Licensed Machine Code Version 7 and installed fixes to manage
POWER7 processor-based servers or an environment with a mixture of POWER5,
POWER5+, POWER6, POWER6+, and POWER7 processor-based servers. The HMCs
provide a locking mechanism so that only one HMC at a time has write access to the service
processor. It is recommended that both HMCs are available on a public subnet to allow full
syncronization of functionality. Depending on your environment, you have multiple options to
configure the network.
Note: The service processor is used to monitor and manage the system hardware
resources and devices. The service processor offers two Ethernet 10/100 Mbps ports as
connections. Note the following information:
Both Ethernet ports are visible only to the service processor and can be used to attach
the server to an HMC or to access the ASMI options from a client web browser using
the HTTP server integrated into the service processor internal operating system.
When not configured otherwise (DHCP or from a previous ASMI setting), both Ethernet
ports of the first FSP have predefined IP addresses:
– Service processor Eth0 or HMC1 port is configured as 169.254.2.147 with netmask
255.255.255.0.
255.255.255.0.
For the second FSP of IBM Power 770 and 780, these default addresses are:
255.255.255.0.
255.255.255.0.
For more information about the service processor, see “Service processor” on page 169.

Figure 2-37 shows one possible highly available HMC configuration managing two servers.
These servers have only one CEC and therefore only one FSP. Each HMC is connected to
one FSP port of all managed servers.
Figure 2-37 Highly available HMC and network architecture
Note that only hardware management networks (LAN1 and LAN2) are highly available
(Figure 2-37) for simplicity. However, the management network (LAN3) can be made
highly available by using a similar concept and adding more Ethernet adapters to LPARs
and HMCs.
Both HMCs must be on a separate VLAN to protect from any network contention. Each HMC
can be a DHCP server for its VLAN.
Redundant service processor connectivity
For the Power 770 and Power 780 with two or more CECs, two redundant service processors
are installed in CEC enclosures 1 and 2. Redundant service processor function requires that
each HMC must be attached to one Ethernet port in CEC enclosure 1 and one Ethernet port
in CEC enclosure 2.
HMC1 HMC2
System A System B
1 2
FSP
1 2
FSP
LAN 1 LAN 2
LPAR A1
LPAR A2
LPAR A3
LPAR B1
LPAR B2
LPAR B3
eth0eth1
eth0eth1
LAN1 – Hardware management network for
first FSP ports (private)
second FSP ports (private), separate
network hardware from LAN1
LAN3 – Open network for HMC access and
dLPAR operations
LAN3 – Open network

Figure 2-38 shows a redundant HMC and redundant service processor connectivity
configuration.
Figure 2-38 Redundant HMC connection and redundant service processor configuration
In a configuration with multiple systems or HMC, the customer is required to provide switches
or hubs to connect each HMC to the server FSP Ethernet ports in each system:
One HMC should connect to the port labeled HMC Port 1 on the first two CEC drawers
of each system.
A second HMC must be attached to HMC Port 2 on the first two CEC drawers of
each system.
This solution provides redundancy for both the HMC and the service processors.
HMC1 HMC2
CEC 1 CEC 2
1 2
FSP
1 2
FSP
LAN 1 LAN 2
LPAR 1
LPAR 2
LPAR 3
eth0eth1
eth0eth1
first FSP ports (private)
second FSP ports (private), separate
network hardware from LAN1
LAN3 - Open network for HMC access and
dLPAR operations
LAN3 – Open network

Figure 2-39 describes the four possible Ethernet connectivity options between the HMC and
service processors.
Figure 2-39 Summary of HMC to FSP configuration option depending on number of CEC
For details about redundant HMC, see Hardware Management Console V7 Handbook,
SG24-7491.
HMC code level
The HMC code must be at V7R7.4.0 to support the Power 770 and Power 780 systems.
In a dual HMC configuration, both must be at the same version and release of the HMC.
Drawer 1
FSP Card
Enet 1
Enet 2
HUB 0
Configuration #1 – Single drawer and one HMC
Note: HUB is optional.
Customer can have a direct connection to the FSP card.
HMC #1
Enet
Drawer 1
FSP Card
Enet 1
Enet 2
HUB 0
Configuration #2 – Single drawer and two HMCs
Note: HUB is optional.
HMC #1
Enet
HUB 1HMC #2
Enet
Drawer 1
FSP Card
Enet 1
Enet 2
Configuration #3 – Multi-drawer with one HMC
HUB 0HMC #1
Enet
Drawer 2
FSP Card
Enet 1
Enet 2
Drawer 1
FSP Card
Enet 1
Enet 2
Configuration #4 – Multi-drawer with two HMCs
Drawer 2
FSP Card
Enet 1
Enet 2
HUB 1HMC #2
Enet
HUB 0HMC #1
Enet

If you want to migrate an LPAR from a POWER6 processor-based server onto a POWER7
processor-based server using PowerVM Live Partition Mobility, consider this: If the source
server is managed by one HMC and the destination server is managed by a different HMC,
ensure that the HMC managing the POWER6 processor-based server is at V7R7.3.5 or later
and the HMC managing the POWER7 processor-based server is at V7R7.4.0 or later.
2.13 IBM Systems Director Management Console
The newly released IBM Systems Director Management Console (SDMC) is intended to be
used in the same manner as the HMC. It provides the same functionality, including hardware,
service, and virtualization management, for Power Systems server and Power Systems
blades. Because SDMC uses IBM Systems Director Express Edition, it also provides all
Systems Director Express capabilities, such as monitoring of operating systems and creating
event action plans.
No configuration changes are required when a client moves from HMC management to
SDMC management.
Much of the SDMC function is equivalent to the HMC. This includes:
Server (host) management.
Virtualization management.
Redundancy and high availability: The SDMC offers console redundancy similar to the HMC.
The scalability and performance of the SDMC matches that of a current HMC. This includes
both the number of systems (hosts) and the number of partitions (virtual servers) that can be
managed. Currently, 48 small-tier entry servers or 32 large-tier servers can be managed by
the SDMC with up to 1,024 partitions (virtual servers) configured across those managed
systems (hosts).
The SDMC can be obtained as a hardware appliance in the same manner as an HMC.
Hardware appliances support managing all Power Systems servers. The SDMC can
optionally be obtained in a virtual appliance format, capable of running on VMware (ESX/i 4,
Tips: Note the following tips:
When upgrading the code of a dual HMC configuration, a good practice is to disconnect
one HMC to avoid having both HMCs connected to the same server but running different
levels of code. If no profiles or partition changes take place during the upgrade, both
HMCs can stay connected. If the HMCs are at different levels and a profile change is
made from the HMC at level V7R7.4.0, for example, the format of the data stored in the
server could be changed, causing the HMC at a previous level (for example, 3.50) to
possibly go into a recovery state because it does not understand the new data format.
Compatibility rules exist between the various software that is executing within a
POWER7 processor-based server environment:
– HMC
– VIO
– System firmware
– Partition operating systems
To check which combinations are supported and to identify required upgrades, you can
use the Fix Level Recommendation Tool web page:
http://www14.software.ibm.com/webapp/set2/flrt/home

or later), and KVM (Red Hat Enterprise Linux (RHEL) 5.5). The virtual appliance is only
supported or managing small-tier Power servers and Power Systems blades.
Table 2-38 and Table 2-38 detail whether the SDMC software appliance, hardware appliance,
or both are supported for each model.
Table 2-38 Type of SDMC appliance support for POWER7-based server
The IBM SDMC Hardware Appliance requires an IBM 7042-CR6 Rack-mounted Hardware
Management Console and IBM SDMC indicator (#0963).
Note: At the time of writing, the SDMC is not supported for the Power 770 (9117-MMC)
and Power 780 (9179-MHC) models.
IBM intends to enhance the IBM Systems Director Management Console (SDMC) to
support the Power 770 (9117-MMC) and Power 780 (9179-MHC). IBM also intends for the
current HMC 7042-CR6 to be upgradable to an IBM SDMC that supports the Power 770
(9117-MMC) and Power 780 (9179-MHC).
POWER7 models Type of SDMC appliance supported
7891-73X (IBM BladeCenter® PS703) Hardware or software appliance
7891-74X (IBM BladeCenter PS704) Hardware or software appliance
8202-E4B (IBM Power 720 Express) Hardware or software appliance
8406-70Y (IBM BladeCenter PS700) Hardware or software appliance
8406-71Y (IBM BladeCenter PS701 and PS702) Hardware or software appliance
8231-E2B (IBM Power 710 and IBM Power 730 Express) Hardware or software appliance
8236-E8C (IBM Power 755) Hardware or software appliance
9117-MMB (IBM Power 770) Hardware appliance only
9179-MHB (IBM Power 780) Hardware appliance only
9119-FHB (IBM Power 795) Hardware appliance only
Note: When ordering #0963, the features #0031(No Modem), #1946 (additional 4 GB
memory), and #1998 (additional 500 GB SATA HDD) are being configured automatically.
Feature #0963 replaces the HMC software with IBM Systems Director Management
Console Hardware Appliance V6.7.3 (5765-MCH).
Neither an external modem (#0032) nor an internal modem (#0033) can be selected with
IBM SDMC indicator (#0963).
To run HMC LMC (#0962), you cannot order the additional storage (#1998). However, you
can order the additional memory (#1946).

The IBM SDMC Virtual Appliance requires an IBM Systems Director Management Console
V6.7.3 (5765-MCV).
SDMC on POWER6 processor-based servers and blades requires eFirmware level 3.5.7.
SDMC on Power Systems POWER7 processor-based servers and blades requires
eFirmware level 7.3.0.
For more detailed information about the SDMC, see IBM Systems Director Management
Console: Introduction and Overview, SG24-7860.
2.14 Operating system support
The IBM POWER7 processor-based systems support three families of operating systems:
AIX
IBM i
Linux
In addition, the Virtual I/O Server can be installed in special partitions that provide support to
the other operating systems for using features such as virtualized I/O devices, PowerVM Live
Partition Mobility, or PowerVM Active Memory Sharing.
2.14.1 Virtual I/O Server
The minimum required level of Virtual I/O server for both the Power 770 and Power 780 is
VIOS 2.2.1.0.
IBM regularly updates the Virtual I/O Server code. To find information about the latest
updates, visit the Fix Central website:
2.14.2 IBM AIX operating system
The following sections discuss the various levels of AIX operating system support.
Note: If you want to use the software appliance, you have to provide the hardware and
virtualization environment.
At a minimum, the following resources must be available to the virtual machine:
2.53 GHz Intel Xeon E5630, Quad Core processor
500 GB storage
8 GB memory
The following hypervisors are supported:
VMware (ESXi 4.0 or later)
KVM (RHEL 5.5)
Note: For details about the software available on IBM Power Systems, visit the Power
Systems Software™ website:
http://www.ibm.com/systems/power/software/index.html

IBM periodically releases maintenance packages (service packs or technology levels) for the
AIX operating system. Information about these packages, downloading, and obtaining the
CD-ROM is on the Fix Central website:
The Fix Central website also provides information about how to obtain the fixes shipping
on CD-ROM.
The Service Update Management Assistant, which can help you to automate the task of
checking and downloading operating system downloads, is part of the base operating
system. For more information about the suma command, go to following website:
http://www14.software.ibm.com/webapp/set2/sas/f/genunix/suma.html
IBM AIX Version 5.3
The minimum level of AIX Version 5.3 to support the Power 770 and Power 780 is AIX 5.3
with the 5300-12 Technology Level and Service Pack 5 or later.
A partition using AIX Version 5.3 will be executing in POWER6 or POWER6+ compatibility
mode. This means that although the POWER7 processor has the ability to run four hardware
threads per core simultaneously, using AIX 5.3 limits the number of hardware threads per
core to two.
IBM AIX Version 6.1
The minimum level of AIX Version 6.1 to support the Power 770 and Power 780 is:
AIX 6.1 with the 6100-07 Technology Level or later
AIX 6.1 with the 6100-06 Technology Level and Service Pack 6 or later
AIX 6.1 with the 6100-05 Technology Level and Service Pack 7 or later
A partition using AIX 6.1 with TL6 can run in POWER6, POWER6+, or POWER7 mode. It is
best to run the partition in POWER7 mode to allow exploitation of new hardware capabilities
such as SMT4 and Active Memory Expansion (AME).
IBM AIX Version 7.1
The minimum level of AIX Version 7.1 to support the Power 770 and Power 780 is:
AIX 7.1 with the 7100-01 Technology Level or later
AIX 7.1 with the 7100-00 Technology Level and Service Pack 4 or 1 later
A partition using AIX 7.1 can run in POWER6, POWER6+, or POWER7 mode. It is best to run
the partition in POWER7 mode to allow exploitation of new hardware capabilities such as
SMT4 and AME.
2.14.3 IBM i operating system
The IBM i operating system is supported on the Power 720 and Power 740 with these
minimum required levels:
IBM i Version 6.1 with i 6.1.1 machine code or later
IBM i Version 7.1 or later
IBM periodically releases maintenance packages (service packs or technology levels) for the
IBM i operating system. Information about these packages, downloading, and obtaining the
CD-ROM is on the Fix Central website:

2.14.4 Linux operating system
Linux is an open source operating system that runs on numerous platforms from embedded
systems to mainframe computers. It provides a UNIX-like implementation across many
computer architectures.
The supported versions of Linux on POWER7 processor-based servers are:
SUSE Linux Enterprise Server 11 Service Pack 1, or later, with current maintenance
updates available from SUSE to enable all planned functionality
Red Hat Enterprise Linux AP 5 Update 7 for POWER, or later
Red Hat Enterprise Linux 6.1 for POWER, or later
If you want to configure Linux partitions in virtualized Power Systems, you have to be aware
of these conditions:
Not all devices and features that are supported by the AIX operating system are supported
in logical partitions running the Linux operating system.
Linux operating system licenses are ordered separately from the hardware. You can
acquire Linux operating system licenses from IBM to be included with the POWER7
processor-based servers, or from other Linux distributors.
For information about the features and external devices supported by Linux, go to:
http://www.ibm.com/systems/p/os/linux/index.html
For information about SUSE Linux Enterprise Server 10, go to:
http://www.novell.com/products/server
For information about Red Hat Enterprise Linux Advanced Server, go to:
http://www.redhat.com/rhel/features
2.14.5 Java supported versions
There are unique considerations when running Java 1.4.2 on POWER7 servers. For best
exploitation of the outstanding performance capabilities and most recent improvements of
POWER7 technology, IBM recommends upgrading Java-based applications to Java 7,
Java 6, or Java 5 whenever possible. For more information, visit:
http://www.ibm.com/developerworks/java/jdk/aix/service.html
2.14.6 Boosting performance and productivity with IBM compilers
IBM XL C, XL C/C++, and XL Fortran compilers for AIX and for Linux exploit the latest
POWER7 processor architecture. Release after release, these compilers continue to help
improve application performance and capability, exploiting architectural enhancements made
available through the advancement of the POWER technology.
IBM compilers are designed to optimize and tune your applications for execution on IBM
POWER platforms, to help you unleash the full power of your IT investment, to create and
maintain critical business and scientific applications, to maximize application performance,
and to improve developer productivity.
The performance gain from years of compiler optimization experience is seen in the
continuous release-to-release compiler improvements that support the POWER4 processors,

through to the POWER4+, POWER5, POWER5+, and POWER6 processors, and now
including the new POWER7 processors. With the support of the latest POWER7 processor
chip, IBM advances a more than 20-year investment in the XL compilers for POWER series
and PowerPC® series architectures.
XL C, XL C/C++, and XL Fortran features introduced to exploit the latest POWER7 processor
include vector unit and vector scalar extension (VSX) instruction set to efficiently manipulate
vector operations in your application, vector functions within the Mathematical Acceleration
Subsystem (MASS) libraries for improved application performance, built-in functions or
intrinsics and directives for direct control of POWER instructions at the application level, and
architecture and tune compiler options to optimize and tune your applications.
COBOL for AIX enables you to selectively target code generation of your programs to
either exploit POWER7 systems architecture or to be balanced among all supported
POWER systems. The performance of COBOL for AIX applications is improved by means
of an enhanced back-end optimizer. The back-end optimizer, a component common also
to the IBM XL compilers lets your applications leverage the latest industry-leading
optimization technology.
The performance of PL/I for AIX applications has been improved through both front-end
changes and back-end optimizer enhancements. The back-end optimizer, a component
common also to the IBM XL compilers, lets your applications leverage the latest
industry-leading optimization technology. For PL/I it will produce code that is intended to
perform well across all hardware levels, including POWER7 of AIX.
IBM Rational® Development Studio for IBM i 7.1 provides programming languages for
creating modern business applications. This includes the ILE RPG, ILE COBOL, C, and C++
compilers as well as the heritage RPG and COBOL compilers. The latest release includes
performance improvements and XML processing enhancements for ILE RPG and ILE
COBOL, improved COBOL portability with a new COMP-5 data type, and easier Unicode
migration with relaxed USC2 rules in ILE RPG. Rational has also released a product called
Rational Open Access: RPG Edition. This product opens the ILE RPG file I/O processing,
enabling partners, tool providers, and users to write custom I/O handlers that can access
other devices like databases, services, and web user interfaces.
IBM Rational Developer for Power Systems Software provides a rich set of integrated
development tools that support the XL C/C++ for AIX compiler, the XL C for AIX compiler, and
the COBOL for AIX compiler. Rational Developer for Power Systems Software offers
capabilities of file management, searching, editing, analysis, build, and debug, all integrated
into an Eclipse workbench. XL C/C++, XL C, and COBOL for AIX developers can boost
productivity by moving from older, text-based, command-line development tools to a rich set
of integrated development tools.
The IBM Rational Power Appliance solution provides a workload-optimized system and
integrated development environment for AIX development on IBM Power Systems. IBM
Rational Power Appliance includes a Power Express server preinstalled with a
comprehensive set of Rational development software along with the AIX operating system.
The Rational development software includes support for Collaborative Application Lifecycle
Management (C/ALM) through Rational Team Concert™, a set of software development tools
from Rational Developer for Power Systems Software, and a choice between the XL C/C++
for AIX or COBOL for AIX compilers.

2.15 Energy management
The Power 770 and 780 servers are designed with features to help clients become more
energy efficient. The IBM Systems Director Active Energy Manager exploits EnergyScale
technology, enabling advanced energy management features to dramatically and dynamically
conserve power and further improve energy efficiency. Intelligent Energy optimization
capabilities enable the POWER7 processor to operate at a higher frequency for increased
performance and performance per watt or dramatically reduce frequency to save energy.
2.15.1 IBM EnergyScale technology
IBM EnergyScale technology provides functions to help the user understand and dynamically
optimize the processor performance versus processor energy consumption, and system
workload, to control IBM Power Systems power and cooling usage.
On POWER7 processor-based systems, the thermal power management device (TPMD)
card is responsible for collecting the data from all system components, changing operational
parameters in components, and interacting with the IBM Systems Director Active Energy
Manager (an IBM Systems Directors plug-in) for energy management and control.
IBM EnergyScale makes use of power and thermal information collected from the system in
order to implement policies that can lead to better performance or better energy utilization.
IBM EnergyScale features include:
Power trending
EnergyScale provides continuous collection of real-time server energy consumption. This
enables administrators to predict power consumption across their infrastructure and to
react to business and processing needs. For example, administrators can use such
information to predict datacenter energy consumption at various times of the day, week,
or month.
Thermal reporting
IBM Director Active Energy Manager can display measured ambient temperature and
calculated exhaust heat index temperature. This information can help identify data center
hot spots that need attention.
Power saver mode
Power saver mode lowers the processor frequency and voltage on a fixed amount,
reducing the energy consumption of the system while still delivering predictable
performance. This percentage is predetermined to be within a safe operating limit and
is not user configurable. The server is designed for a fixed frequency drop of up to
30% down from nominal frequency (the actual value depends on the server type and
configuration). Power saver mode is not supported during boot or re-boot, although it is
a persistent condition that will be sustained after the boot when the system starts
executing instructions.
Dynamic power saver mode
Dynamic power saver mode varies processor frequency and voltage based on the
utilization of the POWER7 processors. Processor frequency and utilization are inversely
proportional for most workloads, implying that as the frequency of a processor increases,
its utilization decreases, given a constant workload. Dynamic power saver mode takes
advantage of this relationship to detect opportunities to save power, based on measured
real-time system utilization.

When a system is idle, the system firmware will lower the frequency and voltage to power
energy saver mode values. When fully utilized, the maximum frequency will vary
depending on whether the user favors power savings or system performance. If an
administrator prefers energy savings and a system is fully utilized, the system is designed
to reduce the maximum frequency to 95% of nominal values. If performance is favored
over energy consumption, the maximum frequency can be increased to up to 109% of
nominal frequency for extra performance.
Dynamic power saver mode is mutually exclusive with power saver mode. Only one of
these modes can be enabled at a given time.
Power capping
Power capping enforces a user-specified limit on power usage. Power capping is not a
power-saving mechanism. It enforces power caps by throttling the processors in the
system, degrading performance significantly. The idea of a power cap is to set a limit that
must never be reached but that frees up extra power never used in the data center. The
margined power is this amount of extra power that is allocated to a server during its
installation in a datacenter. It is based on the server environmental specifications that
usually are never reached because server specifications are always based on maximum
configurations and worst-case scenarios. The user must set and enable an energy cap
from the IBM Director Active Energy Manager user interface.
Soft power capping
There are two power ranges into which the power cap can be set, power capping, as
described previously, and soft power capping. Soft power capping extends the allowed
energy capping range further, beyond a region that can be guaranteed in all configurations
and conditions. If the energy management goal is to meet a particular consumption limit,
then soft power capping is the mechanism to use.
Processor core nap mode
The IBM POWER7 processor uses a low-power mode called nap that stops processor
execution when there is no work to do on that processor core. The latency of exiting nap
mode is very small, typically not generating any impact on applications running. Because
of that, the POWER Hypervisor™ can use nap mode as a general-purpose idle state.
When the operating system detects that a processor thread is idle, it yields control of a
hardware thread to the POWER Hypervisor. The POWER Hypervisor immediately puts
the thread into nap mode. Nap mode allows the hardware to turn the clock off on most of
the circuits inside the processor core. Reducing active energy consumption by turning off
the clocks allows the temperature to fall, which further reduces leakage (static) power of
the circuits causing a cumulative effect. Nap mode saves from 10 - 15% of power
consumption in the processor core.
Processor core sleep mode
To be able to save even more energy, the POWER7 processor has an even lower power
mode called sleep. Before a core and its associated L2 and L3 caches enter sleep mode,
caches are flushed and transition lookaside buffers (TLB) are invalidated, and the
hardware clock is turned off in the core and in the caches. Voltage is reduced to minimize
leakage current. Processor cores inactive in the system (such as CoD processor cores)
are kept in Sleep mode. Sleep mode saves about 35% power consumption in the
processor core and associated L2 and L3 caches.
Fan control and altitude input
System firmware will dynamically adjust fan speed based on energy consumption,
altitude, ambient temperature, and energy savings modes. Power Systems are designed
to operate in worst-case environments, in hot ambient temperatures, at high altitudes, and
with high power components. In a typical case, one or more of these constraints are not
valid. When no power savings setting is enabled, fan speed is based on ambient

temperature and assumes a high-altitude environment. When a power savings setting is
enforced (either Power Energy Saver Mode or Dynamic Power Saver Mode), fan speed
will vary based on power consumption, ambient temperature, and altitude available.
System altitude can be set in IBM Director Active Energy Manager. If no altitude is set, the
system will assume a default value of 350 meters above sea level.
Processor folding
Processor folding is a consolidation technique that dynamically adjusts, over the short
term, the number of processors available for dispatch to match the number of processors
demanded by the workload. As the workload increases, the number of processors made
available increases. As the workload decreases, the number of processors made
available decreases. Processor folding increases energy savings during periods of low to
moderate workload because unavailable processors remain in low-power idle states (nap
or sleep) longer.
EnergyScale for I/O
IBM POWER7 processor-based systems automatically power off hot pluggable PCI
adapter slots that are empty or not being used. System firmware automatically scans all
pluggable PCI slots at regular intervals, looking for those that meet the criteria for being
not in use and powering them off. This support is available for all POWER7
processor-based servers and the expansion units that they support.
Server power down
If overall data center processor utilization is low, workloads can be consolidated on fewer
numbers of servers so that some servers can be turned off completely. It makes sense to
do this when there will be long periods of low utilization, such as weekends. AEM provides
information, such as the power that will be saved and the time it will take to bring a server
back online, that can be used to help make the decision to consolidate and power off. As
with many of the features available in IBM Systems Director and Active Energy Manager,
this function is scriptable and can be automated.
Partition power management
Available with Active Energy Manager 4.3.1 or later, and POWER7 systems with the 730
firmware release or later, is the capability to set a power savings mode for partitions or the
system processor pool. As in the system-level power savings modes, the per-partition
power savings modes can be used to achieve a balance between the power consumption
and the performance of a partition. Only partitions that have dedicated processing units
can have a unique power savings setting. Partitions that run in shared processing mode
will have a common power savings setting, which is that of the system processor pool.
This is because processing unit fractions cannot be power- managed.
As in the case of system-level power savings, two Dynamic Power Saver options are
offered:
– Favor partition performance
– Favor partition power savings
The user must configure this setting from Active Energy Manager. When dynamic power
saver is enabled in either mode, system firmware continuously monitors the performance
and utilization of each of the computer's POWER7 processor cores that belong to the
partition. Based on this utilization and performance data, the firmware will dynamically
adjust the processor frequency and voltage, reacting within milliseconds to adjust
workload performance and also deliver power savings when the partition is under-utilized.
In addition to the two dynamic power saver options, the customer can select to have no
power savings on a given partition. This option will leave the processor cores assigned to
the partition running at their nominal frequencies and voltages.

A new power savings mode, called inherit host setting, is available and is only applicable
to partitions. When configured to use this setting, a partition will adopt the power savings
mode of its hosting server. By default, all partitions with dedicated processing units, and
the system processor pool, are set to the inherit host setting.
On POWER7 processor-based systems, several EnergyScales are imbedded in
the hardware and do not require an operating system or external management
component. More advanced functionality requires Active Energy Manager (AEM) and
IBM Systems Director.
Table 2-39 provides a list of all features supported, showing all cases in which AEM is not
required. Table 2-39 also details the features that can be activated by traditional user
interfaces (for example, ASMI and HMC).
Table 2-39 AEM support
The Power 770 and Power 780 systems implement all the EnergyScale capabilities listed
in 2.15.1, “IBM EnergyScale technology” on page 114.
2.15.2 Thermal power management device card
The thermal power management device (TPMD) card is a separate micro controller installed
on some POWER6 processor-based systems, and on all POWER7 processor-based
systems. It runs real-time firmware whose sole purpose is to manage system energy.
The TPMD card monitors the processor modules, memory, environmental temperature, and
fan speed. Based on this information, it can act upon the system to maintain optimal power
and energy conditions (for example, increase the fan speed to react to a temperature
change). It also interacts with the IBM Systems Director Active Energy Manager to report
power and thermal information and to receive input from AEM on policies to be set. The
TPMD is part of the EnergyScale infrastructure.
Feature Active Energy Manager (AEM) required ASMI HMC
Power Trending Y N N
Thermal Reporting Y N N
Static Power Saver N Y Y
Dynamic Power Saver Y N N
Power Capping Y N N
Energy-optimized Fans N - -
Processor Core Nap N - -
Processor Core Sleep N - -
Processor Folding N - -
EnergyScale for I/O N - -
Server Power Down Y - -
Partition Power
Management
Y - -

Chapter 3. Virtualization
As you look for ways to maximize the return on your IT infrastructure investments,
consolidating workloads becomes an attractive proposition.
IBM Power Systems combined with PowerVM technology are designed to help you
consolidate and simplify your IT environment with the following key capabilities:
Improve server utilization and sharing I/O resources to reduce total cost of ownership and
make better use of IT assets.
Improve business responsiveness and operational speed by dynamically re-allocating
resources to applications as needed, to better match changing business needs or handle
unexpected changes in demand.
Simplify IT infrastructure management by making workloads independent of hardware
resources, thereby enabling you to make business-driven policies to deliver resources
based on time, cost, and service-level requirements.
This chapter discusses the virtualization technologies and features on IBM Power Systems:
POWER Hypervisor
POWER Modes
Partitioning
Active Memory Expansion
PowerVM
System Planning Tool
3

3.1 POWER Hypervisor
Combined with features designed into the POWER7 processors, the POWER Hypervisor
delivers functions that enable other system technologies, including logical partitioning
technology, virtualized processors, IEEE VLAN compatible virtual switch, virtual SCSI
adapters, virtual Fibre Channel adapters, and virtual consoles. The POWER Hypervisor is a
basic component of the system’s firmware and offers the following functions:
Provides an abstraction between the physical hardware resources and the logical
partitions that use them
Enforces partition integrity by providing a security layer between logical partitions
Controls the dispatch of virtual processors to physical processors (See “Processing mode”
on page 131.)
Saves and restores all processor state information during a logical processor
context switch
Controls hardware I/O interrupt management facilities for logical partitions
Provides virtual LAN channels between logical partitions that help to reduce the need for
physical Ethernet adapters for inter-partition communication
Monitors the Service Processor and performs a reset or reload if it detects the loss of the
Service Processor, notifying the operating system if the problem is not corrected
The POWER Hypervisor is always active, regardless of the system configuration and also
when not connected to the managed console. It requires memory to support the resource
assignment to the logical partitions on the server. The amount of memory required by the
POWER Hypervisor firmware varies according to several factors. Factors influencing the
POWER Hypervisor memory requirements include these:
Number of logical partitions
Number of physical and virtual I/O devices used by the logical partitions
Maximum memory values specified in the logical partition profiles
The minimum amount of physical memory required to create a partition will be the size of the
system’s Logical Memory Block (LMB). The default LMB size varies according to the amount
of memory configured in the CEC (Table 3-1).
Table 3-1 Configured CEC memory-to-default Logical Memory Block size
In most cases, however, the actual minimum requirements and recommendations of the
supported operating systems are above 256 MB. Physical memory is assigned to partitions in
increments of LMB.
The POWER Hypervisor provides the following types of virtual I/O adapters:
Virtual SCSI
Virtual Ethernet
Virtual Fibre Channel
Virtual (TTY) console
Configurable CEC memory Default Logical Memory Block
Greater than 8 GB, up to 16 GB 64 MB
Greater than 16 GB, up to 32 GB 128 MB
Greater than 32 GB 256 MB

Chapter 3. Virtualization 121
Virtual SCSI
The POWER Hypervisor provides a virtual SCSI mechanism for virtualization of storage
devices. The storage virtualization is accomplished using two, paired adapters:
A virtual SCSI server adapter
A virtual SCSI client adapter
A Virtual I/O Server partition or a IBM i partition can define virtual SCSI server adapters.
Other partitions are client partitions. The Virtual I/O Server partition is a special logical
partition, as described in 3.4.4, “Virtual I/O Server” on page 137. The Virtual I/O Server
software is included on all PowerVM Editions and when using the PowerVM Standard Edition
and PowerVM Enterprise Edition, dual Virtual I/O Servers can be deployed to provide
maximum availability for client partitions when performing Virtual I/O Server maintenance.
Virtual Ethernet
The POWER Hypervisor provides a virtual Ethernet switch function that allows partitions on
the same server to use a fast and secure communication without any need for physical
interconnection. The virtual Ethernet allows a transmission speed in the range of 1 - 3 Gbps,
depending on the maximum transmission unit (MTU) size and CPU entitlement. Virtual
Ethernet support began with IBM AIX Version 5.3, or an appropriate level of Linux supporting
virtual Ethernet devices (see 3.4.9, “Operating system support for PowerVM” on page 148).
The virtual Ethernet is part of the base system configuration.
Virtual Ethernet has the following major features:
The virtual Ethernet adapters can be used for both IPv4 and IPv6 communication and can
transmit packets with a size up to 65,408 bytes. Therefore, the maximum MTU for the
corresponding interface can be up to 65,394 (65,390 if VLAN tagging is used).
The POWER Hypervisor presents itself to partitions as a virtual 802.1Q-compliant switch.
The maximum number of VLANs is 4096. Virtual Ethernet adapters can be configured as
either untagged or tagged (following the IEEE 802.1Q VLAN standard).
A partition can support 256 virtual Ethernet adapters. Besides a default port VLAN ID,
the number of additional VLAN ID values that can be assigned per virtual Ethernet
adapter is 20, which implies that each virtual Ethernet adapter can be used to access 21
virtual networks.
Each partition operating system detects the virtual local area network (VLAN) switch
as an Ethernet adapter without the physical link properties and asynchronous data
transmit operations.
Any virtual Ethernet can also have connectivity outside of the server if a layer-2 bridge to a
physical Ethernet adapter is set in one Virtual I/O Server partition (see 3.4.4, “Virtual I/O
Server” on page 137, for more details about shared Ethernet), also known as Shared
Ethernet Adapter.
Note: Virtual Ethernet is based on the IEEE 802.1Q VLAN standard. No physical I/O
adapter is required when creating a VLAN connection between partitions, and no access to
an outside network is required.

Virtual Fibre Channel
A virtual Fibre Channel adapter is a virtual adapter that provides client logical partitions with a
Fibre Channel connection to a storage area network through the Virtual I/O Server logical
partition. The Virtual I/O Server logical partition provides the connection between the virtual
Fibre Channel adapters on the Virtual I/O Server logical partition and the physical Fibre
Channel adapters on the managed system. Figure 3-1 depicts the connections between the
client partition virtual Fibre Channel adapters and the external storage. For additional
information, see 3.4.8, “N_Port ID virtualization” on page 147.
Figure 3-1 Connectivity between virtual Fibre Channels adapters and external SAN devices
Virtual (TTY) console
Each partition must have access to a system console. Tasks such as operating system
installation, network setup, and various problem analysis activities require a dedicated system
console. The POWER Hypervisor provides the virtual console by using a virtual TTY or serial
adapter and a set of Hypervisor calls to operate on them. Virtual TTY does not require the
purchase of any additional features or software, such as the PowerVM Edition features.
Depending on the system configuration, the operating system console can be provided by the
Hardware Management Console virtual TTY, IVM virtual TTY, or from a terminal emulator that
is connected to a system port.
3.2 POWER processor modes
Although, strictly speaking, not a virtualization feature, the POWER modes are described
here because they affect various virtualization features.
Client logical
partition 1
Client virtual
fibre channel
adapter
Virtual I/O Server 1
Client logical
partition 2
Client virtual
fibre channel
adapter
Client logical
partition 3
Client virtual
fibre channel
adapter
Hypervisor
Physical fibre
channel adapter
Server virtual fibre
channel adapter
channel adapter
channel adapter
Storage Area
Network
Physical
storage 1
Physical
storage 2
Physical
storage 3

On Power System servers, partitions can be configured to run in several modes, including:
POWER6 compatibility mode
This execution mode is compatible with Version 2.05 of the Power Instruction Set
Architecture (ISA). For more information, visit the following address:
http://www.power.org/resources/reading/PowerISA_V2.05.pdf
POWER6+ compatibility mode
This mode is similar to POWER6, with eight additional Storage Protection Keys.
POWER7 mode
This is the native mode for POWER7 processors, implementing the v2.06 of the Power
Instruction Set Architecture. For more information, visit the following address:
http://www.power.org/resources/downloads/PowerISA_V2.06_PUBLIC.pdf
The selection of the mode is made on a per-partition basis, from the managed console, by
editing the partition profile (Figure 3-2).
Figure 3-2 Configuring partition profile compatibility mode from the managed console

Table 3-2 lists the differences between these modes.
Table 3-2 Differences between POWER6 and POWER7 mode
3.3 Active Memory Expansion
Active Memory Expansion enablement is an optional feature of POWER7 processor-based
servers that must be specified when creating the configuration in the e-Config tool, as follows:
IBM Power 770 #4791
IBM Power 780 #4791
This feature enables memory expansion on the system. Using compression/decompression
of memory content can effectively expand the maximum memory capacity, providing
additional server workload capacity and performance.
Active Memory Expansion is an innovative POWER7 technology that allows the effective
maximum memory capacity to be much larger than the true physical memory maximum.
Compression/decompression of memory content can allow memory expansion up to 100%,
which in turn enables a partition to perform significantly more work or support more users
with the same physical amount of memory. Similarly, it can allow a server to run more
partitions and do more work for the same physical amount of memory.
Active Memory Expansion is available for partitions running AIX 6.1, Technology Level 4 with
SP2, or later.
Active Memory Expansion uses CPU resource of a partition to compress/decompress the
memory contents of this same partition. The trade-off of memory capacity for processor
cycles can be an excellent choice, but the degree of expansion varies based on how
compressible the memory content is, and it also depends on having adequate spare CPU
capacity available for this compression/decompression. Tests in IBM laboratories, using
POWER6 and POWER6+
mode
POWER7 mode Customer value
2-thread SMT 4-thread SMT Throughput performance,
processor core utilization
Vector Multimedia Extension/
AltiVec (VMX)
Vector Scalar Extension (VSX) High-performance computing
Affinity OFF by default 3-tier memory, Micropartition
Affinity
Improved system performance
for system images spanning
sockets and nodes
Barrier Synchronization
Fixed 128-byte array,
Kernel Extension Access
Enhanced Barrier
Synchronization
Variable Sized Array, User
Shared Memory Access
High-performance computing
parallel programming
synchronization facility
64-core and 128-thread
scaling
scaling
scaling
scaling
Performance and scalability for
large scale-up single system
image workloads (such as
OLTP, ERP scale-up, and
WPAR consolidation)
EnergyScale CPU Idle EnergyScale CPU Idle and
Folding with NAP and SLEEP
Improved energy efficiency

sample work loads, showed excellent results for many workloads in terms of memory
expansion per additional CPU utilized. Other test workloads had more modest results.
Clients have much control over Active Memory Expansion usage. Each individual AIX
partition can turn on or turn off Active Memory Expansion. Control parameters set the amount
of expansion desired in each partition to help control the amount of CPU used by the Active
Memory Expansion function. An initial program load (IPL) is required for the specific partition
that is turning memory expansion on or off. After turned on, monitoring capabilities are
available in standard AIX performance tools, such as lparstat, vmstat, topas, and svmon.
Figure 3-3 represents the percentage of CPU that is used to compress memory for two
partitions with separate profiles. The green curve corresponds to a partition that has spare
processing power capacity. The blue curve corresponds to a partition constrained in
processing power.
Figure 3-3 CPU usage versus memory expansion effectiveness
Both cases show that there is a knee-of-curve relationship for CPU resource required for
memory expansion:
Busy processor cores do not have resources to spare for expansion.
The more memory expansion done, the more CPU resource required.
The knee varies depending on how compressible that the memory contents are. This
example demonstrates the need for a case-by-case study of whether memory expansion can
provide a positive return on investment.
% CPU
utilization
for
expansion
Amount of memory expansion
1 = Plenty of spare
CPU resource
available
2 = Constrained
CPU resource –
already running at
significant utilization
1
2
Very cost effective

To help you perform this study, a planning tool is included with AIX 6.1 Technology Level 4,
allowing you to sample actual workloads and estimate how expandable the partition's
memory is and how much CPU resource is needed. Any model Power System can run the
planning tool. Figure 3-4 shows an example of the output returned by this planning tool. The
tool outputs various real memory and CPU resource combinations to achieve the desired
effective memory. It also recommends one particular combination. In this example, the tool
recommends that you allocate 58% of a processor to benefit from 45% extra memory
capacity.
Figure 3-4 Output from Active Memory Expansion planning tool
Active Memory Expansion Modeled Statistics:
-----------------------
Modeled Expanded Memory Size : 8.00 GB
Expansion True Memory Modeled Memory CPU Usage
Factor Modeled Size Gain Estimate
--------- -------------- ----------------- -----------
1.21 6.75 GB 1.25 GB [ 19%] 0.00
1.31 6.25 GB 1.75 GB [ 28%] 0.20
1.41 5.75 GB 2.25 GB [ 39%] 0.35
1.51 5.50 GB 2.50 GB[ 45%] 0.58
1.61 5.00 GB 3.00 GB [ 60%] 1.46
Active Memory Expansion Recommendation:
---------------------
The recommended AME configuration for this workload is to configure
the LPAR with a memory size of 5.50 GB and to configure a memory
expansion factor of 1.51. This will result in a memory expansion of
45% from the LPAR's current memory size. With this configuration,
the estimated CPU usage due to Active Memory Expansion is
approximately 0.58 physical processors, and the estimated overall
peak CPU resource required for the LPAR is 3.72 physical processors.

After you select the value of the memory expansion factor that you want to achieve, you can
use this value to configure the partition from the managed console (Figure 3-5).
Figure 3-5 Using the planning tool result to configure the partition
On the HMC menu describing the partition, check the Active Memory Expansion box and
enter true and maximum memory, and the memory expansion factor. To turn off expansion,
clear the check box. In both cases, a reboot of the partition is needed to activate the change.
In addition, a one-time, 60-day trial of Active Memory Expansion is available to provide more
exact memory expansion and CPU measurements. The trial can be requested using the
Capacity on Demand web page:
http://www.ibm.com/systems/power/hardware/cod/
Active Memory Expansion can be ordered with the initial order of the server or as an MES
order. A software key is provided when the enablement feature is ordered that is applied to
the server. Rebooting is not required to enable the physical server. The key is specific to an
individual server and is permanent. It cannot be moved to a separate server. This feature is
ordered per server, independently of the number of partitions using memory expansion.
Active Memory Expansion Modeled Statistics:
-----------------------
Modeled Expanded Memory Size : 8.00 GB
Expansion True Memory Modeled Memory CPU Usage
Factor Modeled Size Gain Estimate
--------- -------------- ----------------- -----------
1.21 6.75 GB 1.25 GB [ 19%] 0.00
1.31 6.25 GB 1.75 GB [ 28%] 0.20
1.41 5.75 GB 2.25 GB [ 39%] 0.35
1.51 5.50 GB 2.50 GB[ 45%] 0.58
1.61 5.00 GB 3.00 GB [ 60%] 1.46
Active Memory Expansion Recommendation:
---------------------
The recommended AME configuration for this workload is to
configure the LPAR with a memory size of 5.50 GB and to configure
a memory expansion factor of 1.51. This will result in a memory
expansion of 45% from the LPAR's current memory size. With this
configuration, the estimated CPU usage due to Active Memory
Expansion is approximately 0.58 physical processors, and the
estimated overall peak CPU resource required for the LPAR is 3.72
physical processors.
5.5 true
8.0 max
Sample output

hypervisorFrom the HMC, you can view whether the Active Memory Expansion feature has
been activated (Figure 3-6).
Figure 3-6 Server capabilities listed from the HMC
For detailed information regarding Active Memory Expansion, you can download the
document Active Memory Expansion: Overview and Usage Guide from this location:
http://www-01.ibm.com/common/ssi/cgi-bin/ssialias?infotype=SA&subtype=WH&appname=S
TGE_PO_PO_USEN&htmlfid=POW03037USEN
3.4 PowerVM
The PowerVM platform is the family of technologies, capabilities, and offerings that deliver
industry-leading virtualization on the IBM Power Systems. It is the new umbrella branding
term for Power Systems Virtualization (Logical Partitioning, Micro-Partitioning, POWER
Hypervisor, Virtual I/O Server, Live Partition Mobility, Workload Partitions, and more). As with
Advanced Power Virtualization in the past, PowerVM is a combination of hardware
enablement and value-added software. Section 3.4.1, “PowerVM editions” on page 129,
discusses the licensed features of each of the three separate editions of PowerVM.
Note: If you want to move an LPAR using Active Memory Expansion to a different system
using Live Partition Mobility, the target system must support AME (the target system must
have AME activated with the software key). If the target system does not have AME
activated, the mobility operation fails during the pre-mobility check phase, and an
appropriate error message displays to the user.

3.4.1 PowerVM editions
This section provides information about the virtualization capabilities of the PowerVM. The
three editions of PowerVM are suited for various purposes, as follows:
PowerVM Express Edition
PowerVM Express Edition is designed for customers looking for an introduction to
more advanced virtualization features at a highly affordable price, generally in
single-server projects.
PowerVM Standard Edition
This edition provides advanced virtualization functions and is intended for production
deployments and server consolidation.
PowerVM Enterprise Edition
This edition is suitable for large server deployments such as multi-server deployments and
cloud infrastructure. It includes unique features like Active Memory Sharing and Live
Partition Mobility.
Table 3-3 lists the version of PowerVM that are available on Power 770 and Power 780.
Table 3-3 Availability of PowerVM per POWER7 processor technology-based server model
For more information about the features included on each version of PowerVM, see IBM
PowerVM Virtualization Introduction and Configuration, SG24-7940-04.
3.4.2 Logical partitions (LPARs)
LPARs and virtualization increase utilization of system resources and add a new level of
configuration possibilities. This section provides details and configuration specifications
about this topic.
Dynamic logical partitioning
Logical partitioning was introduced with the POWER4 processor-based product line and the
AIX Version 5.1 operating system. This technology offered the capability to divide a pSeries®
system into separate logical systems, allowing each LPAR to run an operating environment
on dedicated attached devices, such as processors, memory, and I/O components.
Later, dynamic logical partitioning increased the flexibility, allowing selected system
resources, such as processors, memory, and I/O components, to be added and deleted from
logical partitions while they are executing. AIX Version 5.2, with all the necessary
enhancements to enable dynamic LPAR, was introduced in 2002. The ability to reconfigure
dynamic LPARs encourages system administrators to dynamically redefine all available
system resources to reach the optimum capacity for each defined dynamic LPAR.
PowerVM editions Express Standard Enterprise
IBM Power 770 N/A #7942 #7995
IBM Power 780 N/A #7942 #7995
Note: At the time of writing, the IBM Power 770 (9117-MMC) and Power 780 (9179-MHC)
can only be managed by the Hardware Management Console.

Micro-Partitioning
Micro-Partitioning technology allows you to allocate fractions of processors to a logical
partition. This technology was introduced with POWER5 processor-based systems. A logical
partition using fractions of processors is also known as a Shared Processor Partition or
micro-partition. Micro-partitions run over a set of processors called a Shared Processor Pool,
and virtual processors are used to let the operating system manage the fractions of
processing power assigned to the logical partition. From an operating system perspective, a
virtual processor cannot be distinguished from a physical processor, unless the operating
system has been enhanced to be made aware of the difference. Physical processors are
abstracted into virtual processors that are available to partitions. The meaning of the term
physical processor in this section is a processor core. For example, a 2-core server has two
physical processors.
When defining a shared processor partition, several options have to be defined:
The minimum, desired, and maximum processing units
Processing units are defined as processing power, or the fraction of time that the partition
is dispatched on physical processors. Processing units define the capacity entitlement of
the partition.
The Shared Processor Pool
Pick one from the list with the names of each configured Shared Processor Pool. This list
also displays the pool ID of each configured Shared Processor Pool in parentheses. If the
name of the desired Shared Processor Pool is not available here, you must first configure
the desired Shared Processor Pool using the Shared Processor Pool Management
window. Shared processor partitions use the default Shared Processor Pool called
DefaultPool by default. See 3.4.3, “Multiple Shared Processor Pools” on page 132, for
details about Multiple Shared Processor Pools.
Whether the partition will be able to access extra processing power to “fill up” its virtual
processors above its capacity entitlement (selecting either to cap or uncap your partition)
If there is spare processing power available in the Shared Processor Pool or
other partitions are not using their entitlement, an uncapped partition can use
additional processing units if its entitlement is not enough to satisfy its application
processing demand.
The weight (preference) in the case of an uncapped partition
The minimum, desired, and maximum number of virtual processors
The POWER Hypervisor calculates partition processing power based on minimum, desired,
and maximum values, processing mode, and is also based on requirements of other active
partitions. The actual entitlement is never smaller than the processing unit’s desired value,
but can exceed that value in the case of an uncapped partition and up to the number of virtual
processors allocated.
A partition can be defined with a processor capacity as small as 0.10 processing units. This
represents 0.10 of a physical processor. Each physical processor can be shared by up to 10
shared processor partitions, and the partition’s entitlement can be incremented fractionally by
as little as 0.01 of the processor. The shared processor partitions are dispatched and
time-sliced on the physical processors under control of the POWER Hypervisor. The shared
processor partitions are created and managed by the HMC.
The IBM Power 770 supports up to 64 cores, and has the following maximums:
Up to 64 dedicated partitions
Up to 640 micro-partitions (10 micro-partitions per physical active core)

The Power 780 allows up to 96 cores in a single system, supporting the following maximums:
Up to 96 dedicated partitions
Up to 960 micro-partitions (10 micro-partitions per physical active core)
An important point is that the maximums stated are supported by the hardware, but the
practical limits depend on application workload demands.
Additional information about virtual processors includes:
A virtual processor can be running (dispatched) either on a physical processor or as
standby waiting for a physical processor to became available.
Virtual processors do not introduce any additional abstraction level. They are only a
dispatch entity. When running on a physical processor, virtual processors run at the same
speed as the physical processor.
Each partition’s profile defines CPU entitlement that determines how much processing
power any given partition should receive. The total sum of CPU entitlement of all partitions
cannot exceed the number of available physical processors in a Shared Processor Pool.
The number of virtual processors can be changed dynamically through a dynamic
LPAR operation.
Processing mode
When you create a logical partition you can assign entire processors for dedicated use, or you
can assign partial processing units from a Shared Processor Pool. This setting defines the
processing mode of the logical partition. Figure 3-7 shows a diagram of the concepts
discussed in this section.
Figure 3-7 Logical partitioning concepts
Set of micro-partitions
KEY:
vp Virtual processor
lp Logical processor
PrU Processing Units
Shared-Processor Pool 0
Shared-Processor Pool 1
lp lp lp lp lp lp lp lp lp lp lp lp
AIXV6.1
1.5PrU
AIXV5.3
0.5PrU
AIXV6.1
1.5PrU
Linux
0.5PrU
vpvp vp vpvp vp
lp lp lp lp lp lp
AIXV5.3
AIXV6.1
Dedicated processors Dedicated processors
POWER Hypervisor
8-core SMP System

Dedicated mode
In dedicated mode, physical processors are assigned as a whole to partitions. The
simultaneous multithreading feature in the POWER7 processor core allows the core to
execute instructions from two or four independent software threads simultaneously. To
support this feature we use the concept of logical processors. The operating system (AIX,
IBM i, or Linux) sees one physical processor as two or four logical processors if the
simultaneous multithreading feature is on. It can be turned off and on dynamically while the
operating system is executing (for AIX, use the smtctl command). If simultaneous
multithreading is off, each physical processor is presented as one logical processor, and thus
only one thread.
Shared dedicated mode
On POWER7 processor technology-based servers, you can configure dedicated partitions to
become processor donors for idle processors that they own, allowing for the donation of
spare CPU cycles from dedicated processor partitions to a Shared Processor Pool. The
dedicated partition maintains absolute priority for dedicated CPU cycles. Enabling this feature
can help to increase system utilization without compromising the computing power for critical
workloads in a dedicated processor.
Shared mode
In shared mode, logical partitions use virtual processors to access fractions of physical
processors. Shared partitions can define any number of virtual processors (the maximum
number is 10 times the number of processing units assigned to the partition). From the
POWER Hypervisor point of view, virtual processors represent dispatching objects. The
POWER Hypervisor dispatches virtual processors to physical processors according to the
partition’s processing units entitlement. One processing unit represents one physical
processor’s processing capacity. At the end of the POWER Hypervisor’s dispatch cycle
(10 ms), all partitions receive total CPU time equal to their processing unit’s entitlement. The
logical processors are defined on top of virtual processors. So, even with a virtual processor,
the concept of a logical processor exists and the number of logical processors depends
whether the simultaneous multithreading is turned on or off.
3.4.3 Multiple Shared Processor Pools
Multiple Shared Processor Pools (MSPPs) is a capability supported on POWER7 processor
and POWER6 processor-based servers. This capability allows a system administrator to
create a set of micro-partitions with the purpose of controlling the processor capacity that can
be consumed from the physical Shared Processor Pool.

To implement MSPPs, there is a set of underlying techniques and technologies. Figure 3-8
shows an overview of the architecture of Multiple Shared Processor Pools.
Figure 3-8 Overview of the architecture of Multiple Shared Processor Pools
Micro-partitions are created and then identified as members of either the default Shared
Processor Pool0 or a user-defined Shared Processor Pooln. The virtual processors that exist
within the set of micro-partitions are monitored by the POWER Hypervisor, and processor
capacity is managed according to user-defined attributes.
If the Power Systems server is under heavy load, each micro-partition within a Shared
Processor Pool is guaranteed its processor entitlement plus any capacity that it might be
allocated from the reserved pool capacity if the micro-partition is uncapped.
If certain micro-partitions in a Shared Processor Pool do not use their capacity entitlement,
the unused capacity is ceded and other uncapped micro-partitions within the same Shared
Processor Pool are allocated the additional capacity according to their uncapped weighting.
In this way, the entitled pool capacity of a Shared Processor Pool is distributed to the set of
micro-partitions within that Shared Processor Pool.
All Power Systems servers that support the Multiple Shared Processor Pools capability will
have a minimum of one (the default) Shared Processor Pool and up to a maximum of 64
Shared Processor Pools.
POWER Hypervisor
p1p0
Physical Shared-Processor Pool
p2 p3 p4 p5 p6 p7
Shared Processor Pool0
AIX V5.3
EC 1.6
AIX V6.1
EC 0.8
Linux
EC 0.5
vp0
vp1
vp2 vp3
vp4
AIX V6.1
EC 1.6
AIX V6.1
EC 0.8
Linux
EC 0.5
vp5
vp6
vp7
vp8 vp9
vp10
Shared Processor Pool1
Unused capacity in SPP0 is
redistributed to uncapped
micro-partitions within SPP0
Unused capacity in SPP1 is
redistributed to uncapped
micro-partitions within SPP1
KEY:
EC Entitled Capacity
p Physical processor
vp Virtual processor
SPPn Shared-Processor Pooln

Default Shared Processor Pool (SPP0)
On any Power Systems server supporting Multiple Shared Processor Pools, a default Shared
Processor Pool is always automatically defined. The default Shared Processor Pool has a
pool identifier of zero (SPP-ID = 0) and can also be referred to as SPP0. The default Shared
Processor Pool has the same attributes as a user-defined Shared Processor Pool except that
these attributes are not directly under the control of the system administrator. They have fixed
values (Table 3-4).
Table 3-4 Attribute values for the default Shared Processor Pool (SPP0)
Creating Multiple Shared Processor Pools
The default Shared Processor Pool (SPP0) is automatically activated by the system and is
always present.
All other Shared Processor Pools exist, but by default are inactive. By changing the maximum
pool capacity of a Shared Processor Pool to a value greater than zero, it becomes active and
can accept micro-partitions (either transferred from SPP0 or newly created).
Levels of processor capacity resolution
The two levels of processor capacity resolution implemented by the POWER Hypervisor and
Multiple Shared Processor Pools are:
Level0
The first level, Level0, is the resolution of capacity within the same Shared Processor
Pool. Unused processor cycles from within a Shared Processor Pool are harvested and
then redistributed to any eligible micro-partition within the same Shared Processor Pool.
Level1
This is the second level of processor capacity resolution. When all Level0 capacity has
been resolved within the Multiple Shared Processor Pools, the POWER Hypervisor
harvests unused processor cycles and redistributes them to eligible micro-partitions
regardless of the Multiple Shared Processor Pools structure.
SPP0 attribute Value
Shared Processor Pool ID 0.
Maximum pool capacity The value is equal to the capacity in the physical Shared
Processor Pool.
Reserved pool capacity 0.
Entitled pool capacity Sum (total) of the entitled capacities of the micro-partitions in the
default Shared Processor Pool.

Figure 3-9 shows the levels of unused capacity redistribution implemented by the POWER
Hypervisor.
Figure 3-9 The levels of unused capacity redistribution
Capacity allocation above the entitled pool capacity (Level1)
The POWER Hypervisor initially manages the entitled pool capacity at the Shared Processor
Pool level. This is where unused processor capacity within a Shared Processor Pool is
harvested and then redistributed to uncapped micro-partitions within the same Shared
Processor Pool. This level of processor capacity management is sometimes referred to as
Level0 capacity resolution.
At a higher level, the POWER Hypervisor harvests unused processor capacity from the
Multiple Shared Processor Pools that do not consume all of their entitled pool capacity. If a
particular Shared Processor Pool is heavily loaded and several of the uncapped
micro-partitions within it require additional processor capacity (above the entitled pool
capacity), then the POWER Hypervisor redistributes some of the extra capacity to the
uncapped micro-partitions. This level of processor capacity management is sometimes
referred to as Level1 capacity resolution.
To redistribute unused processor capacity to uncapped micro-partitions in Multiple Shared
Processor Pools above the entitled pool capacity, the POWER Hypervisor uses a higher level
of redistribution, Level1.
POWER Hypervisor
SPPnSPP2SPP1SPP0
Micro-partitionn
SPP2 capacity
resolution
SPPn capacity
resolution
SPP1 capacity
resolution
SPP0 capacity
resolution
Physical Shared Processor Pool
p0 p1 p2 p3 p4 p5
Level1 capacity
resolution
Level1 capacity resolution
POWER Hypervisor harvests unused
processor capacity from Shared-Processor
Pools and redistributes it across all
uncapped micro-partitions regardless of the
Shared-Processor Pool structure
Level0 capacity resolution
Resolution of the Entitled Pool Capacity
within the same Shared-Processor Pool
Level0
capacity
resolution
Micro-partition0
Micro-partition1
Micro-partition2
Micro-partition3
Micro-partitionn

Where there is unused processor capacity in under-utilized Shared Processor Pools,
the micro-partitions within the Shared Processor Pools cede the capacity to the
POWER Hypervisor.
In busy Shared Processor Pools, where the micro-partitions have used all of the entitled pool
capacity, the POWER Hypervisor allocates additional cycles to micro-partitions, in which all
of the following statements are true:
The maximum pool capacity of the Shared Processor Pool hosting the micro-partition has
not been met.
The micro-partition is uncapped.
The micro-partition has enough virtual-processors to take advantage of the
additional capacity.
Under these circumstances, the POWER Hypervisor allocates additional processor capacity
to micro-partitions on the basis of their uncapped weights independent of the Shared
Processor Pool hosting the micro-partitions. This can be referred to as Level1 capacity
resolution. Consequently, when allocating additional processor capacity in excess of the
entitled pool capacity of the Shared Processor Pools, the POWER Hypervisor takes the
uncapped weights of all micro-partitions in the system into account, regardless of the Multiple
Shared Processor Pool structure.
Dynamic adjustment of maximum pool capacity
The maximum pool capacity of a Shared Processor Pool, other than the default Shared
Processor Pool0, can be adjusted dynamically from the managed console, using either the
graphical interface or the command-line interface (CLI).
Dynamic adjustment of reserved pool capacity
The reserved pool capacity of a Shared Processor Pool, other than the default Shared
Processor Pool0, can be adjusted dynamically from the managed console, using either the
graphical interface or the CLI interface.
Dynamic movement between Shared Processor Pools
A micro-partition can be moved dynamically from one Shared Processor Pool to another
using the managed console using either the graphical interface or the CLI interface. Because
the entitled pool capacity is partly made up of the sum of the entitled capacities of the
micro-partitions, removing a micro-partition from a Shared Processor Pool reduces the
entitled pool capacity for that Shared Processor Pool. Similarly, the entitled pool capacity of
the Shared Processor Pool that the micro-partition joins will increase.
Deleting a Shared Processor Pool
Shared Processor Pools cannot be deleted from the system. However, they are deactivated
by setting the maximum pool capacity and the reserved pool capacity to zero. The Shared
Processor Pool will still exist but will not be active. Use the managed console interface to
deactivate a Shared Processor Pool. A Shared Processor Pool cannot be deactivated unless
all micro-partitions hosted by the Shared Processor Pool have been removed.
Important: Level1 capacity resolution: When allocating additional processor capacity in
excess of the entitled pool capacity of the Shared Processor Pool, the POWER Hypervisor
takes the uncapped weights of all micro-partitions in the system into account, regardless
of the Multiple Shared Processor Pool structure.

Live Partition Mobility and Multiple Shared Processor Pools
A micro-partition can leave a Shared Processor Pool because of PowerVM Live Partition
Mobility. Similarly, a micro-partition can join a Shared Processor Pool in the same way. When
performing PowerVM Live Partition Mobility, you are given the opportunity to designate a
destination Shared Processor Pool on the target server to receive and host the migrating
micro-partition.
Because several simultaneous micro-partition migrations are supported by PowerVM Live
Partition Mobility, it is conceivable to migrate the entire Shared Processor Pool from one
server to another.
3.4.4 Virtual I/O Server
The Virtual I/O Server is part of all PowerVM Editions. It is a special-purpose partition that
allows the sharing of physical resources between logical partitions to allow more efficient
utilization (for example, consolidation). In this case, the Virtual I/O Server owns the physical
resources (SCSI, Fibre Channel, network adapters, and optical devices) and allows client
partitions to share access to them, thus minimizing the number of physical adapters in the
system. The Virtual I/O Server eliminates the requirement that every partition owns a
dedicated network adapter, disk adapter, and disk drive. The Virtual I/O Server supports
OpenSSH for secure remote logins. It also provides a firewall for limiting access by ports,
network services, and IP addresses. Figure 3-10 shows an overview of a Virtual I/O
Server configuration.
Figure 3-10 Architectural view of the Virtual I/O Server
Because the Virtual I/O Server is an operating system-based appliance server, redundancy
for physical devices attached to the Virtual I/O Server can be provided by using capabilities
such as Multipath I/O and IEEE 802.3ad Link Aggregation.
Installation of the Virtual I/O Server partition is performed from a special system backup DVD
that is provided to clients who order any PowerVM edition. This dedicated software is only for
the Virtual I/O Server (and IVM in case it is used) and is only supported in special Virtual I/O
Server partitions. Three major virtual devices are supported by the Virtual I/O Server:
Shared Ethernet Adapter
Virtual SCSI
Virtual Fibre Channel adapter
Virtual I/O Server Hypervisor
Shared Ethernet
Adapter
Physical Ethernet
Adapter
Physical Disk
Adapter
Virtual I/O Client 1
Virtual Ethernet
Adapter
Virtual SCSI
Adapter
Virtual I/O Client 2
Virtual Ethernet
Adapter
Virtual SCSI
Adapter
Virtual Ethernet
Adapter
Virtual SCSI
Adapter
Physical
Disk
Physical
Disk
External Network

The Virtual Fibre Channel adapter is used with the NPIV feature, described in 3.4.8, “N_Port
ID virtualization” on page 147.
Shared Ethernet Adapter
A Shared Ethernet Adapter (SEA) can be used to connect a physical Ethernet network to a
virtual Ethernet network. The Shared Ethernet Adapter provides this access by connecting
the internal hypervisor VLANs with the VLANs on the external switches. Because the Shared
Ethernet Adapter processes packets at layer 2, the original MAC address and VLAN tags of
the packet are visible to other systems on the physical network. IEEE 802.1 VLAN tagging
is supported.
The Shared Ethernet Adapter also provides the ability for several client partitions to share
one physical adapter. With an SEA, you can connect internal and external VLANs using a
physical adapter. The Shared Ethernet Adapter service can only be hosted in the Virtual I/O
Server, not in a general-purpose AIX or Linux partition, and acts as a layer-2 network bridge
to securely transport network traffic between virtual Ethernet networks (internal) and one or
more (EtherChannel) physical network adapters (external). These virtual Ethernet network
adapters are defined by the POWER Hypervisor on the Virtual I/O Server.
Figure 3-11 shows a configuration example of an SEA with one physical and two virtual
Ethernet adapters. An SEA can include up to 16 virtual Ethernet adapters on the Virtual I/O
Server that share the same physical access.
Figure 3-11 Architectural view of a Shared Ethernet Adapter
Tip: A Linux partition can provide bridging function also, by using the brctl command.
VIOS Client 1
Ethernet
switch
VLAN=2 PVID=1
ent3
(sea)
en3
(if.)
en0
(if.)
Client 2
en0
(if.)
ent0
(virt.)
Client 3
en0
(if.)
ent0
(virt.)
ent1
(virt.)
ent2
(virt.)
ent0
(virt.)
VLAN=2
PVID=2
PVID=99
VID=2
PVID=1
PVID=1
PVID=1
VLAN=1
Hypervisor
External
Network
ent0
(phy.)

A single SEA setup can have up to 16 Virtual Ethernet trunk adapters and each virtual
Ethernet trunk adapter can support up to 20 VLAN networks. Therefore, a possibility is for a
single physical Ethernet to be shared between 320 internal VLAN networks. The number of
shared Ethernet adapters that can be set up in a Virtual I/O Server partition is limited only by
the resource availability, because there are no configuration limits.
Unicast, broadcast, and multicast are supported, so protocols that rely on broadcast or
multicast, such as Address Resolution Protocol (ARP), Dynamic Host Configuration
Protocol (DHCP), Boot Protocol (BOOTP), and Neighbor Discovery Protocol (NDP), can
work on an SEA.
For a more detailed discussion about virtual networking, see:
http://www.ibm.com/servers/aix/whitepapers/aix_vn.pdf
Virtual SCSI
Virtual SCSI is used to refer to a virtualized implementation of the SCSI protocol. Virtual SCSI
is based on a client/server relationship. The Virtual I/O Server logical partition owns the
physical resources and acts as a server or, in SCSI terms, a target device. The client logical
partitions access the virtual SCSI backing storage devices provided by the Virtual I/O Server
as clients.
The virtual I/O adapters (virtual SCSI server adapter and a virtual SCSI client adapter) are
configured using a managed console or through the Integrated Virtualization Manager on
smaller systems. The virtual SCSI server (target) adapter is responsible for executing any
SCSI commands that it receives. It is owned by the Virtual I/O Server partition. The virtual
SCSI client adapter allows a client partition to access physical SCSI and SAN attached
devices and LUNs that are assigned to the client partition. The provisioning of virtual disk
resources is provided by the Virtual I/O Server.
Physical disks presented to the Virtual/O Server can be exported and assigned to a client
partition in a number of ways:
The entire disk is presented to the client partition.
The disk is divided into several logical volumes, which can be presented to a single client
or multiple clients.
As of Virtual I/O Server 1.5, files can be created on these disks, and file-backed storage
devices can be created.
The logical volumes or files can be assigned to separate partitions. Therefore, virtual SCSI
enables sharing of adapters and disk devices.
Note: A Shared Ethernet Adapter does not need to have an IP address configured to be
able to perform the Ethernet bridging functionality. Configuring IP on the Virtual I/O Server
is convenient because the Virtual I/O Server can then be reached by TCP/IP, for example,
to perform dynamic LPAR operations or to enable remote login. This task can be done
either by configuring an IP address directly on the SEA device or on an additional virtual
Ethernet adapter in the Virtual I/O Server. This leaves the SEA without the IP address,
allowing for maintenance on the SEA without losing IP connectivity in case SEA failover
is configured.

Figure 3-12 shows an example where one physical disk is divided into two logical volumes by
the Virtual I/O Server. Each client partition is assigned one logical volume, which is then
accessed through a virtual I/O adapter (VSCSI Client Adapter). Inside the partition, the disk is
seen as a normal hdisk.
Figure 3-12 Architectural view of virtual SCSI
At the time of writing, virtual SCSI supports Fibre Channel, parallel SCSI, iSCSI, SAS, SCSI
RAID devices, and optical devices, including DVD-RAM and DVD-ROM. Other protocols such
as SSA and tape devices are not supported.
For more information about the specific storage devices supported for Virtual I/O Server, see:
http://www14.software.ibm.com/webapp/set2/sas/f/vios/documentation/datasheet.html
Virtual I/O Server functions
The Virtual I/O Server has a number of features, including monitoring solutions:
Support for Live Partition Mobility starting on POWER6 processor-based systems with the
PowerVM Enterprise Edition. For more information about Live Partition Mobility, see 3.4.5,
“PowerVM Live Partition Mobility” on page 141.
Support for virtual SCSI devices backed by a file, which are then accessed as standard
SCSI-compliant LUNs.
Support for virtual Fibre Channel devices that are used with the NPIV feature.
Virtual I/O Server Expansion Pack with additional security functions such as Kerberos
(Network Authentication Service for users and client and server applications), Simple
Network Management Protocol (SNMP) v3, and Lightweight Directory Access Protocol
(LDAP) client functionality.
System Planning Tool (SPT) and Workload Estimator, which are designed to ease the
deployment of a virtualized infrastructure. For more information about the System
Planning Tool, see 3.5, “System Planning Tool” on page 150.
Client Partition 1 Client Partition 2I/O Server Partition
POWER Hypervisor
LVM
Physical
Adapter
Hdisk HdiskLogical
Volume 1
Logical
Volume 2
VSCSI
Server
Adapter
VSCSI
Client
Adapter
VSCSI
Client
Adapter
Physical Disk
(SCSI, FC)
VSCSI
Server
Adapter

Includes IBM Systems Director agent and a number of pre-installed Tivoli agents, such as:
– Tivoli Identity Manager (TIM), to allow easy integration into an existing Tivoli Systems
Management infrastructure
– Tivoli Application Dependency Discovery Manager (ADDM), which creates and
automatically maintains application infrastructure maps including dependencies,
change-histories, and deep configuration values
vSCSI eRAS.
Additional CLI statistics in svmon, vmstat, fcstat, and topas.
Monitoring solutions to help manage and monitor the Virtual I/O Server and shared
resources. New commands and views provide additional metrics for memory, paging,
processes, Fibre Channel HBA statistics, and virtualization.
For more information about the Virtual I/O Server and its implementation, see IBM PowerVM
Virtualization Introduction and Configuration, SG24-7940.
3.4.5 PowerVM Live Partition Mobility
PowerVM Live Partition Mobility allows you to move a running logical partition, including its
operating system and running applications, from one system to another without any shutdown
or without disrupting the operation of that logical partition. Inactive partition mobility allows
you to move a powered-off logical partition from one system to another.
Partition mobility provides systems management flexibility and improves system availability,
as follows:
Avoid planned outages for hardware or firmware maintenance by moving logical partitions
to another server and then performing the maintenance. Live Partition Mobility can help
lead to zero downtime maintenance because you can use it to work around scheduled
maintenance activities.
Avoid downtime for a server upgrade by moving logical partitions to another server and
then performing the upgrade. This approach allows your users to continue their work
without disruption.
Avoid unplanned downtime. With preventive failure management, if a server indicates a
potential failure, you can move its logical partitions to another server before the failure
occurs. Partition mobility can help avoid unplanned downtime.
Take advantage of server optimization:
– Consolidation: You can consolidate workloads running on several small, under-used
servers onto a single large server.
– Deconsolidation: You can move workloads from server to server to optimize resource
use and workload performance within your computing environment. With active
partition mobility, you can manage workloads with minimal downtime.
Mobile partition’s operating system requirements
The operating system running in the mobile partition has to be AIX or Linux. The Virtual I/O
Server partition itself cannot be migrated. All versions of AIX and Linux supported on the IBM
POWER7 processor-based servers also support partition mobility.
Source and destination system requirements
The source partition must be one that has only virtual devices. If there are any physical
devices in its allocation, they must be removed before the validation or migration is initiated.

An N_Port ID virtualization (NPIV) device is considered virtual and is compatible with
partition migration.
The hypervisor must support the Partition Mobility functionality (also called migration process)
available on POWER 6 and POWER 7 processor-based hypervisors. Firmware must be at
firmware level eFW3.2 or later. All POWER7 processor-based hypervisors support Live
Partition Mobility. Source and destination systems can have separate firmware levels, but
they must be compatible with each other.
A possibility is to migrate partitions back and forth between POWER6 and POWER7
processor-based servers. Partition Mobility leverages the POWER6 Compatibility Modes
that are provided by POWER7 processor-based servers. On the POWER7
processor-based server, the migrated partition is then executing in POWER6 or POWER6+
Compatibility Mode.
If you want to move an active logical partition from a POWER6 processor-based server to a
POWER7 processor-based server so that the logical partition can take advantage of the
additional capabilities available with the POWER7 processor, perform these steps:
1. Set the partition-preferred processor compatibility mode to the default mode. When you
activate the logical partition on the POWER6 processor-based server, it runs in the
POWER6 mode.
2. Move the logical partition to the POWER7 processor-based server. Both the current
and preferred modes remain unchanged for the logical partition until you restart the
logical partition.
3. Restart the logical partition on the POWER7 processor-based server. The hypervisor
evaluates the configuration. Because the preferred mode is set to default and the logical
partition now runs on a POWER7 processor-based server, the highest mode available is
the POWER7 mode. The hypervisor determines that the most fully featured mode that is
supported by the operating environment installed in the logical partition is the POWER7
mode and changes the current mode of the logical partition to the POWER7 mode.
Now the current processor compatibility mode of the logical partition is the POWER7 mode,
and the logical partition runs on the POWER7 processor-based server.
The Virtual I/O Server on the source system provides the access to the client resources and
must be identified as a mover service partition (MSP). The Virtual Asynchronous Services
Interface (VASI) device allows the mover service partition to communicate with the
hypervisor. It is created and managed automatically by the managed console and will be
configured on both the source and destination Virtual I/O Servers, which are designated as
the mover service partitions for the mobile partition, to participate in active mobility. Other
requirements include a similar time-of-day on each server, systems must not be running on
battery power, and shared storage (external hdisk with reserve_policy=no_reserve). In
addition, all logical partitions must be on the same open network with RMC established to the
managed console.
The managed console is used to configure, validate, and orchestrate. You use the managed
console to configure the Virtual I/O Server as an MSP and to configure the VASI device. An
managed console wizard validates your configuration and identifies issues that can cause the
Tip: The “Migration combinations of processor compatibility modes for active Partition
Mobility” web page offers presentations of the supported migrations:
http://publib.boulder.ibm.com/infocenter/powersys/v3r1m5/topic/p7hc3/iphc3pcmco
mbosact.htm

migration to fail. During the migration, the managed console controls all phases of
the process.
Improved Live Partition Mobility benefits
The possibility to move partitions between POWER6 and POWER7 processor-based servers
greatly facilitates the deployment of POWER7 processor-based servers, as follows:
Installation of the new server can be performed while the application is executing on a
POWER6 server. After the POWER7 processor-based server is ready, the application can
be migrated to its new hosting server without application down time.
When adding POWER7 processor-based servers to a POWER6 environment, you get the
additional flexibility to perform workload balancing across the entire set of POWER6 and
POWER7 processor-based servers.
When performing server maintenance, you get the additional flexibility to use POWER6
Servers for hosting applications usually hosted on POWER7 processor-based servers,
and vice versa, allowing you to perform this maintenance with no application planned
down time.
For more information about Live Partition Mobility and how to implement it, see IBM
PowerVM Live Partition Mobility, SG24-7460.
3.4.6 Active Memory Sharing
Active Memory Sharing is an IBM PowerVM advanced memory virtualization technology that
provides system memory virtualization capabilities to IBM Power Systems, allowing multiple
partitions to share a common pool of physical memory.
Active Memory Sharing is only available with the Enterprise version of PowerVM.
The physical memory of an IBM Power System can be assigned to multiple partitions either in
dedicated or shared mode. The system administrator has the capability to assign some
physical memory to a partition and some physical memory to a pool that is shared by other
partitions. A single partition can have either dedicated or shared memory:
With a pure dedicated memory model, the system administrator’s task is to optimize
available memory distribution among partitions. When a partition suffers degradation
because of memory constraints and other partitions have unused memory, the
administrator can manually issue a dynamic memory reconfiguration.
With a shared memory model, the system automatically decides the optimal distribution of
the physical memory to partitions and adjusts the memory assignment based on partition
load. The administrator reserves physical memory for the shared memory pool, assigns
partitions to the pool, and provides access limits to the pool.
Active Memory Sharing can be exploited to increase memory utilization on the system either
by decreasing the global memory requirement or by allowing the creation of additional
partitions on an existing system. Active Memory Sharing can be used in parallel with Active
Memory Expansion on a system running a mixed workload of several operating system. For
example, AIX partitions can take advantage of Active Memory Expansion. Other operating
systems take advantage of Active Memory Sharing.
For additional information regarding Active Memory Sharing, see IBM PowerVM Virtualization
Active Memory Sharing, REDP-4470.

3.4.7 Active Memory Deduplication
In a virtualized environment, the systems might have a considerable amount of
duplicated information stored on RAM after each partition has its own operating system,
and some of them might even share the same kind of applications. On heavily loaded
systems this might lead to a shortage of the available memory resources, forcing paging by
the AMS partition operating systems, the AMD pool, or both, which might decrease overall
system performance.
Figure 3-13 shows the standard behavior of a system without Active Memory Deduplication
(AMD) enabled on its AMS shared memory pool. Identical pages within the same or different
LPARs each require their own unique physical memory page, consuming space with
repeated information.
Figure 3-13 AMS shared memory pool without AMD enabled
Active Memory Deduplication allows the hypervisor to dynamically map identical partition
memory pages to a single physical memory page within a shared memory pool. This enables
a better utilization of the AMS shared memory pool, increasing the system’s overall
performance by avoiding paging. Deduplication can cause the hardware to incur fewer cache
misses, which will also lead to improved performance.
D UD D D U U U U U U
U U U U U U U U U U U
U U U
D U U
U U U U U
D D U U U
U U
U U
D U
U U
Without
Active Memory
Deduplication
Mappings
AMS shared memory pool
LPAR1
Logical Memory
LPAR2
Logical Memory
LPAR3
Logical Memory
D
U
Duplicate pages
Unique pages
KEY:

Figure 3-14 shows the behavior of a system with Active Memory Deduplication enabled on its
AMS shared memory pool. Duplicated pages from different LPARs are stored just once,
providing the AMS pool with more free memory.
Figure 3-14 Identical memory pages mapped to a single physical memory page with Active Memory
Duplication enabled
Active Memory Deduplication (AMD) depends on the Active Memory Sharing (AMS) feature
to be available, and consumes CPU cycles donated by the AMS pool's VIOS partitions to
identify deduplicated pages. The operating systems running on the AMS partitions can hint to
the PowerVM Hypervisor that some pages (such as frequently referenced read-only code
pages) are particularly good for deduplication.
To perform deduplication, the hypervisor cannot compare every memory page in the AMS
pool with every other page. Instead, it computes a small signature for each page that it visits
and stores the signatures in an internal table. Each time that a page is inspected, its signature
is looked up against the known signatures in the table. If a match is found, the memory pages
are compared to be sure that the pages are really duplicates. When a duplicate is found, the
hypervisor remaps the partition memory to the existing memory page and returns the
duplicate page to the AMS pool.
D U U U U U U U
U U U U U U U U U U U
U U U
D U U
U U U U U
D D U U U
U U
U U
D U
U U
With
Active Memory
Deduplication
Mappings
AMS shared memory pool
LPAR1
Logical Memory
LPAR2
Logical Memory
LPAR3
Logical Memory
D
U
Duplicate pages
Unique pages
KEY:
Free

Figure 3-15 shows two pages being written in the AMS memory pool and having their
signatures matched on the deduplication table.
Figure 3-15 Memory pages having their signatures matched by Active Memory Deduplication
From the LPAR point of view, the AMD feature is completely transparent. If an LPAR attempts
to modify a deduplicated page, the hypervisor grabs a free page from the AMS pool, copies
the duplicate page contents into the new page, and maps the LPAR's reference to the new
page so that the LPAR can modify its own unique page.
System administrators can dynamically configure the size of the deduplication table, ranging
from 1/8192 up to 1/256 of the configured maximum AMS memory pool size. Having this table
too small might lead to missed deduplication opportunities. Conversely, having a table that is
too large might waste a small amount of overhead space.
The management of the Active Memory Deduplication feature is done via managed console,
allowing administrators to take the following steps:
Enable and disable Active Memory Deduplication at an AMS Pool level.
Display deduplication metrics.
Display and modify the deduplication table size.
AMS
Memory
Pool
Page A
Dedup
Table
Sign A
Signature
Function
AMS
Memory
Pool
Page A
Dedup
Table
Sign A
Signature
Function
Page B
Signature
Function
Signature of Page A being written
on the Deduplication Table
Signature of Page B matching
Sign A on the Deduplication Table

Figure 3-16 shows the Active Memory Deduplication being enabled to a shared memory pool.
Figure 3-16 Enabling the Active Memory Deduplication for a shared memory pool
The Active Memory Deduplication feature requires the following minimum components:
PowerVM Enterprise edition
System firmware level 740
AIX Version 6: AIX 6.1 TL7 or later
AIX Version 7: AIX 7.1 TL1 SP1 or later
IBM i: 7.14 or 7.2 or later
SLES 11 SP2 or later
RHEL 6.2 or later
3.4.8 N_Port ID virtualization
N_Port ID virtualization (NPIV) is a technology that allows multiple logical partitions to access
independent physical storage through the same physical Fibre Channel adapter. This adapter
is attached to a Virtual I/O Server partition that acts only as a pass-through, managing the
data transfer through the POWER Hypervisor.
Each partition using NPIV is identified by a pair of unique worldwide port names, enabling you
to connect each partition to independent physical storage on a SAN. Unlike virtual SCSI, only
the client partitions see the disk.
For additional information and requirements for NPIV, see these resources:
PowerVM Migration from Physical to Virtual Storage, SG24-7825
IBM PowerVM Virtualization Managing and Monitoring, SG24-7590
NPIV is supported in PowerVM Standard and Enterprise Editions on the IBM Power 770 and
Power 780 servers.

3.4.9 Operating system support for PowerVM
Table 3-5 summarizes the PowerVM features supported by the operating systems compatible
with the POWER7 processor-based servers.
Table 3-5 PowerVM features supported by AIX, IBM i, and Linux
Feature AIX
V5.3
AIX
V6.1
AIX
V7.1
IBM i
6.1.1
IBM i
7.1
RHEL
V5.7
RHEL
V6.1
SLES
V10
SP4
SLES
V11
SP1
Virtual SCSI Yes Yes Yes Yes Yes Yes Yes Yes Yes
Virtual Ethernet Yes Yes Yes Yes Yes Yes Yes Yes Yes
Shared Ethernet
Adapter
Yes Yes Yes Yes Yes Yes Yes Yes Yes
Virtual Fibre
Channel
Virtual Tape Yes Yes Yes Yes Yes Yes Yes Yes Yes
Logical Partitioning Yes Yes Yes Yes Yes Yes Yes Yes Yes
DLPAR I/O adapter
add/remove
DLPAR I/O
processor
add/remove
DLPAR I/O
memory add
DLPAR I/O
memory remove
Yes Yes Yes Yes Yes Yes Yes No Yes
Micro-Partitioning Yes Yes Yes Yes Yes Yes Yes Yes Yes
Shared Dedicated
Capacity
Multiple Shared
Processor Pools
Virtual I/O Server Yes Yes Yes Yes Yes Yes Yes Yes Yes
Suspend/Resume No Yes Yes No No No No No No
Shared Storage
Pools
Yes Yes Yes Yes Yesa No No No No
Thin Provisioning Yes Yes Yes Yesb Yesb No No No No
Active Memory
Sharing and Active
Memory
Deduplication
No Yes Yes Yes Yes No Yes No Yes
Live Partition
Mobility
Yes Yes Yes No No Yes Yes Yes Yes

3.4.10 POWER7 Linux programming support
IBM Linux Technology Center (LTC) contributes to the development of Linux by providing
support for IBM hardware in Linux distributions. In particular, the LTC makes tools and code
available to the Linux communities to take advantage of the POWER7 technology and
develop POWER7 optimized software.
Table 3-6 lists the support of specific programming features for various versions of Linux.
Table 3-6 Linux support for POWER7 features
Simultaneous
Multi-Threading
(SMT)
Yesc
Yesd
Yes Yese
Yes Yesc
Yesc
Yesc
Yes
Active Memory
Expansion
No Yesf
Yes No No No No No No
a. Requires IBM i 7.1 TR1.
b. Will become a fully provisioned device when used by IBM i.
c. Only supports two threads.
d. AIX 6.1 up to TL4 SP2 only supports two threads, and supports four threads as of TL4 SP3.
e. IBM i 6.1.1 and up support SMT4.
f. On AIX 6.1 with TL4 SP2 and later.
Feature AIX
V5.3
AIX
V6.1
AIX
V7.1
IBM i
6.1.1
IBM i
7.1
RHEL
V5.7
RHEL
V6.1
SLES
V10
SP4
SLES
V11
SP1
Features Linux releases Comments
SLES 10 SP4 SLES 11 RHEL 5.7 RHEL 6.1
POWER6
compatibility mode
Yes Yes Yes Yes -
POWER7 mode No Yes No Yes -
Strong Access
Ordering
No Yes No Yes Can improve Lx86
performance
Scale to 256 cores/
1024 threads
No Yes No Yes Base OS support
available
4-way SMT No Yes No Yes -
VSX support No Yes No Yes Full exploitation
requires Advance
Toolchain.
Distro toolchain
mcpu/mtune=p7
No Yes No Yes SLES11/GA toolchain
has minimal P7
enablementnecessary
to support kernel build
Advance Toolchain
support
Yes,
execution
restricted to
Power6
instructions
Yes Yes,
execution
restricted to
Power6
instructions
Yes Alternative IBM GNU
Toolchain
64k base page size No Yes Yes Yes -

For information regarding Advance Toolchain, see the following website:
http://www.ibm.com/developerworks/wikis/display/hpccentral/How+to+use+Advance+Tool
chain+for+Linux+on+POWER
Also see the University of Illinois Linux on Power Open Source Repository:
http://ppclinux.ncsa.illinois.edu
ftp://linuxpatch.ncsa.uiuc.edu/toolchain/at/at05/suse/SLES_11/release_notes.at05-
2.1-0.html
ftp://linuxpatch.ncsa.uiuc.edu/toolchain/at/at05/redhat/RHEL5/release_notes.at05-
2.1-0.html
3.5 System Planning Tool
The IBM System Planning Tool (SPT) helps you design systems to be partitioned with logical
partitions. You can also plan for and design non-partitioned systems by using the SPT. The
resulting output of your design is called a system plan, which is stored in a .sysplan file. This
file can contain plans for a single system or multiple systems. The .sysplan file can be used
for the following reasons:
To create reports
As input to the IBM configuration tool (e-Config)
To create and deploy partitions on your system (or systems) automatically
System plans that are generated by the SPT can be deployed on the system by the Hardware
Management Console (HMC), Systems Director Management Console (SDMC), or
Integrated Virtualization Manager (IVM).
You can create an entirely new system configuration, or you can create a system
configuration based on any of these items:
Performance data from an existing system that the new system is to replace
Performance estimates that anticipates future workloads that you must support
Sample systems that you can customize to fit your needs
Integration between the SPT and both the Workload Estimator (WLE) and IBM Performance
Management (PM) allows you to create a system that is based on performance and capacity
data from an existing system or that is based on new workloads that you specify.
Tickless idle No Yes No Yes Improved energy
utilization and
virtualization of
partially to fully idle
partitions
Features Linux releases Comments
SLES 10 SP4 SLES 11 RHEL 5.7 RHEL 6.1
Note: Ask your IBM representative or Business Partner to use the Customer Specified
Placement manufacturing option if you want to automatically deploy your partitioning
environment on a new machine. SPT looks for the resource’s allocation to be the same as
that specified in your .sysplan file.

You can use the SPT before you order a system to determine what you must order to support
your workload. You can also use the SPT to determine how you can partition a system that
you already have.
Using the System Planning Tool is an effective way of documenting and backing up key
system settings and partition definitions. It allows the user to create records of systems and
export them to their personal workstation or backup system of choice. These same backups
can then be imported back onto the same managed console when needed. This can be
useful when cloning systems enabling the user to import the system plan to any managed
console multiple times.
The SPT and its supporting documentation can be found on the IBM System Planning
Tool site:

Chapter 4. Continuous availability and
manageability
This chapter provides information about IBM reliability, availability, and serviceability (RAS)
design and features. This set of technologies implemented on IBM Power Systems servers
provides the possibility to improve your architecture’s total cost of ownership (TCO) by
reducing unplanned down time.
RAS can be described as follows:
Reliability: Indicates how infrequently a defect or fault in a server manifests itself
Availability: Indicates how infrequently the functionality of a system or application is
impacted by a fault or defect
Serviceability: Indicates how well faults and their impacts are communicated to users and
services, and how efficiently and nondisruptively the faults are repaired
Each successive generation of IBM servers is designed to be more reliable than the previous
server family. POWER7 processor-based servers have new features to support new levels of
virtualization, help ease administrative burden, and increase system utilization.
Reliability starts with components, devices, and subsystems designed to be fault-tolerant.
POWER7 uses lower voltage technology, improving reliability with stacked latches to reduce
soft error (SER) susceptibility. During the design and development process, subsystems go
through rigorous verification and integration testing processes. During system manufacturing,
systems go through a thorough testing process to help ensure high product quality levels.
The processor and memory subsystem contain a number of features designed to avoid or
correct environmentally induced, single-bit, intermittent failures, as well as handle solid faults
in components, including selective redundancy to tolerate certain faults without requiring an
outage or parts replacement.
4

IBM is the only vendor that designs, manufactures, and integrates its most critical server
components, including:
POWER processors
Caches
Memory buffers
Hub-controllers
Clock cards
Service processors
Design and manufacturing verification and integration, as well as field support information, is
used as feedback for continued improvement on the final products.
This chapter also includes a manageability section describing the means to successfully
manage your systems.
Several software-based availability features exist that are based on the benefits available
when using AIX and IBM i as the operating system. Support of these features when using
Linux can vary.

Chapter 4. Continuous availability and manageability 155
4.1 Reliability
Highly reliable systems are built with highly reliable components. On IBM POWER
processor-based systems, this basic principle is expanded upon with a clear design for
reliability architecture and methodology. A concentrated, systematic, architecture-based
approach is designed to improve overall system reliability with each successive generation of
system offerings.
4.1.1 Designed for reliability
Systems designed with fewer components and interconnects have fewer opportunities to fail.
Simple design choices, such as integrating processor cores on a single POWER chip, can
dramatically reduce the opportunity for system failures. In this case, an 8-core server can
include one-fourth as many processor chips (and chip socket interfaces) as with a double
CPU-per-processor design. Not only does this case reduce the total number of system
components, it reduces the total amount of heat generated in the design, resulting in an
additional reduction in required power and cooling components. POWER7 processor-based
servers also integrate L3 cache into the processor chip for a higher integration of parts.
Parts selection also plays a critical role in overall system reliability. IBM uses three grades of
components, grade 3 defined as industry standard (off-the-shelf). As shown in Figure 4-1,
using stringent design criteria and an extensive testing program, the IBM manufacturing team
can produce grade 1 components that are expected to be 10 times more reliable than
industry standard. Engineers select grade 1 parts for the most critical system components.
Newly introduced organic packaging technologies, rated grade 5, achieve the same reliability
as grade 1 parts.
Figure 4-1 Component failure rates
Component failure rates
0
0.2
0.4
0.6
0.8
1
Grade 3 Grade 1 Grade 5

4.1.2 Placement of components
Packaging is designed to deliver both high performance and high reliability. For example,
the reliability of electronic components is directly related to their thermal environment, that
is, large decreases in component reliability are directly correlated with relatively small
increases in temperature. POWER processor-based systems are carefully packaged to
ensure adequate cooling. Critical system components such as the POWER7 processor chips
are positioned on printed circuit cards so that they receive fresh air during operation. In
addition, POWER processor-based systems are built with redundant, variable-speed fans that
can automatically increase output to compensate for increased heat in the central
electronic complex.
4.1.3 Redundant components and concurrent repair
High-opportunity components, or those that most affect system availability, are protected with
redundancy and the ability to be repaired concurrently.
The use of redundant parts allows the system to remain operational. Among the parts are:
POWER7 cores, which include redundant bits in L1-I, L1-D, and L2 caches, and in L2 and
L3 directories
Power 770 and Power 780 main memory DIMMs, which contain an extra DRAM chip for
improved redundancy
Power 770 and 780 redundant system clock and service processor for configurations with
two or more central electronics complex (CEC) drawers
Redundant and hot-swap cooling
Redundant and hot-swap power supplies
Redundant 12X loops to I/O subsystem
For maximum availability, be sure to connect power cords from the same system to two
separate Power Distribution Units (PDUs) in the rack and to connect each PDU to
independent power sources. Deskside form factor power cords must be plugged into two
independent power sources to achieve maximum availability.
4.2 Availability
The IBM hardware and microcode capability to continuously monitor execution of hardware
functions is generally described as the process of first-failure data capture (FFDC). This
process includes the strategy of predictive failure analysis, which refers to the ability to track
intermittent correctable errors and to vary components off-line before they reach the point of
hard failure, causing a system outage, and without the need to re-create the problem.
Note: Check your configuration for optional redundant components before ordering
your system.

The POWER7 family of systems continues to introduce significant enhancements that are
designed to increase system availability and ultimately a high availability objective with
hardware components that are able to perform the following functions:
Self-diagnose and self-correct during run time.
Automatically reconfigure to mitigate potential problems from suspect hardware.
Self-heal or automatically substitute good components for failing components.
Throughout this chapter, we describe IBM POWER technology’s capabilities that are focused
on keeping a system environment up and running. For a specific set of functions that are
focused on detecting errors before they become serious enough to stop computing work, see
4.3.1, “Detecting” on page 169.
4.2.1 Partition availability priority
Also available is the ability to assign availability priorities to partitions. If an alternate
processor recovery event requires spare processor resources and there are no other means
of obtaining the spare resources, the system determines which partition has the lowest
priority and attempts to claim the needed resource. On a properly configured POWER
processor-based server, this approach allows that capacity to first be obtained from a
low-priority partition instead of a high-priority partition.
This capability is relevant to the total system availability because it gives the system an
additional stage before an unplanned outage. In the event that insufficient resources exist to
maintain full system availability, these servers attempt to maintain partition availability by
user-defined priority.
Partition availability priority is assigned to partitions using a weight value or integer rating, the
lowest priority partition rated at 0 (zero) and the highest priority partition valued at 255. The
default value is set at 127 for standard partitions and 192 for Virtual I/O Server (VIOS)
partitions. You can vary the priority of individual partitions.
Partition availability priorities can be set for both dedicated and shared processor partitions.
The POWER Hypervisor uses the relative partition weight value among active partitions to
favor higher priority partitions for processor sharing, adding and removing processor capacity,
and favoring higher priority partitions for normal operation.
Note that the partition specifications for minimum, desired, and maximum capacity are also
taken into account for capacity-on-demand options and if total system-wide processor
capacity becomes disabled because of deconfigured failed processor cores. For example, if
total system-wide processor capacity is sufficient to run all partitions, at least with the
minimum capacity, the partitions are allowed to start or continue running. If processor
capacity is insufficient to run a partition at its minimum value, then starting that partition
results in an error condition that must be resolved.
4.2.2 General detection and deallocation of failing components
Runtime correctable or recoverable errors are monitored to determine if there is a pattern of
errors. If these components reach a predefined error limit, the service processor initiates an
action to deconfigure the faulty hardware, helping to avoid a potential system outage and to
enhance system availability.
Note: POWER7 processor-based servers are independent of the operating system for
error detection and fault isolation within the central electronics complex.

Persistent deallocation
To enhance system availability, a component that is identified for deallocation or
deconfiguration on a POWER processor-based system is flagged for persistent deallocation.
Component removal can occur either dynamically (while the system is running) or at boot
time (IPL), depending both on the type of fault and when the fault is detected.
In addition, runtime unrecoverable hardware faults can be deconfigured from the system after
the first occurrence. The system can be rebooted immediately after failure and resume
operation on the remaining stable hardware. This way prevents the same faulty hardware
from affecting system operation again. The repair action is deferred to a more convenient,
less critical time.
Persistent deallocation functions include:
Processor
L2/L3 cache lines (cache lines are dynamically deleted)
Memory
Deconfigure or bypass failing I/O adapters
Processor instruction retry
As in POWER6, the POWER7 processor has the ability to retry processor instruction and
alternate processor recovery for a number of core related faults. This ability significantly
reduces exposure to both permanent and intermittent errors in the processor core.
Intermittent errors, often because of cosmic rays or other sources of radiation, are generally
not repeatable.
With this function, when an error is encountered in the core, in caches and certain logic
functions, the POWER7 processor first automatically retries the instruction. If the source of
the error was truly transient, the instruction succeeds and the system continues as before.
On IBM systems prior to POWER6, this error caused a checkstop.
Alternate processor retry
Hard failures are more difficult, being permanent errors that are replicated each time that the
instruction is repeated. Retrying the instruction does not help in this situation because the
instruction will continue to fail.
As in POWER6, POWER7 processors have the ability to extract the failing instruction from
the faulty core and retry it elsewhere in the system for a number of faults, after which the
failing core is dynamically deconfigured and scheduled for replacement.
Dynamic processor deallocation
Dynamic processor deallocation enables automatic deconfiguration of processor cores when
patterns of recoverable core-related faults are detected. Dynamic processor deallocation
prevents a recoverable error from escalating to an unrecoverable system error, which might
otherwise result in an unscheduled server outage. Dynamic processor deallocation relies on
the service processor’s ability to use FFDC-generated recoverable error information to notify
the POWER Hypervisor when a processor core reaches its predefined error limit. Then the
POWER Hypervisor dynamically deconfigures the failing core and is called out for
replacement. The entire process is transparent to the partition owning the failing instruction.
If there are available inactivated processor cores or CoD processor cores, the system
effectively puts a CoD processor into operation after an activated processor is determined to
no longer be operational. In this way, the server remains with its total processor power.

If there are no CoD processor cores available system-wide, total processor capacity is
lowered below the licensed number of cores.
Single processor checkstop
As in POWER6, POWER7 provides single-processor check-stopping for certain processor
logic, command, or control errors that cannot be handled by the availability enhancements in
the preceding section.
This way significantly reduces the probability of any one processor affecting total system
availability by containing most processor checkstops to the partition that was using the
processor at the time that the full checkstop goes into effect.
Even with all these availability enhancements to prevent processor errors from affecting
system-wide availability, errors might result on a system-wide outage.
4.2.3 Memory protection
A memory protection architecture that provides good error resilience for a relatively small L1
cache might be very inadequate for protecting the much larger system main store. Therefore,
a variety of protection methods is used in POWER processor-based systems to avoid
uncorrectable errors in memory.
Memory protection plans must take into account many factors, including:
Size
Desired performance
Memory array manufacturing characteristics
POWER7 processor-based systems have a number of protection schemes designed to
prevent, protect, or limit the effect of errors in main memory. These capabilities include:
64-byte ECC code
This innovative ECC algorithm from IBM research allows a full 8-bit device kill to be
corrected dynamically. This ECC code mechanism works on DIMM pairs on a rank basis.
(Depending on the size, a DIMM might have one, two, or four ranks.) With this ECC code,
an entirely bad DRAM chip can be marked as bad (chip mark). After marking the DRAM
as bad, the code corrects all the errors in the bad DRAM. It can additionally mark a 2-bit
symbol as bad and correct the 2-bit symbol, providing a double-error detect or single-error
correct ECC, or a better level of protection in addition to the detection or correction of a
chipkill event.
This improvement in the ECC word algorithm replaces the redundant bit steering used on
POWER6 systems.
The Power 770 and 780, and future POWER7 high-end machines, have a spare DRAM
chip per rank on each DIMM that can be spared out. Effectively, this protection means that
on a rank basis, a DIMM pair can detect and correct two and sometimes three chipkill
events and still provide better protection than ECC, explained in the previous paragraph.
Hardware scrubbing
Hardware scrubbing is a method used to deal with intermittent errors. IBM POWER
processor-based systems periodically address all memory locations. Any memory
locations with a correctable error are rewritten with the correct data.

CRC
The bus that is transferring data between the processor and the memory uses CRC error
detection with a failed operation-retry mechanism and the ability to dynamically retune bus
parameters when a fault occurs. In addition, the memory bus has spare capacity to
substitute a spare data bit-line, for that which is determined to be faulty.
Chipkill
Chipkill is an enhancement that enables a system to sustain the failure of an entire DRAM
chip. Chipkill spreads the bit lines from a DRAM over multiple ECC words so that a
catastrophic DRAM failure does not affect more of what is protected by the ECC code
implementation. The system can continue indefinitely in this state with no performance
degradation until the failed DIMM can be replaced. Figure 4-2 shows an example of how
chipkill technology spreads bit lines across multiple ECC words.
Figure 4-2 Chipkill in action with a spare memory DRAM chip on a Power 770 and Power 780
POWER7 memory subsystem
The POWER7 chip contains two memory controllers with four channels per memory
controller. Each channel connects to a single DIMM, but because the channels work in pairs,
a processor chip can address four DIMM pairs, two pairs per memory controller.
The bus transferring data between the processor and the memory uses CRC error detection
with a failed operation-retry mechanism and the ability to dynamically retune bus parameters
when a fault occurs. In addition, the memory bus has spare capacity to substitute a spare
data bit-line, for that which is determined to be faulty.
ECC word ECC word ECC word ECC word
Chipkill
DRAM sparing
Spare
memory
chip
Scattered memory chip bits across separate ECC words for Chipkill

Figure 4-3 shows a POWER7 chip, with its memory interface, consisting of two controllers
and four DIMMs per controller. Advanced memory buffer chips are exclusive to IBM and help
to increase performance, acting as read/write buffers. On the Power 770 and Power 780, the
advanced memory buffer chips are integrated into the DIMM that they support.
Figure 4-3 POWER7 memory subsystem
Memory page deallocation
Although coincident cell errors in separate memory chips are a statistic rarity, IBM POWER
processor-based systems can contain these errors by using a memory page deallocation
scheme for partitions that are running IBM AIX and IBM i operating systems, as well as for
memory pages owned by the POWER Hypervisor. If a memory address experiences an
uncorrectable or repeated correctable single cell error, the service processor sends the
memory page address to the POWER Hypervisor to be marked for deallocation.
Pages used by the POWER Hypervisor are deallocated as soon as the page is released.
In other cases, the POWER Hypervisor notifies the owning partition that the page should be
deallocated. Where possible, the operating system moves any data currently contained in
that memory area to another memory area and removes the page (or pages) that are
associated with this error from its memory map, no longer addressing these pages. The
operating system performs memory page deallocation without any user intervention and is
transparent to users and applications.
The POWER Hypervisor maintains a list of pages that are marked for deallocation during the
current platform Initial Program Load (IPL). During a partition IPL, the partition receives a list
of all the bad pages in its address space. In addition, if memory is dynamically added to a
partition (through a dynamic LPAR operation), the POWER Hypervisor warns the operating
system when memory pages are included that need to be deallocated.
Ctrl
DIMM
Ctrl
DIMM
Ctrl
DIMM
Ctrl
DIMM
GX
POWER7
Core
256 KB L2
POWER7
Core
256 KB L2
POWER7
Core
256 KB L2
POWER7
Core
256 KB L2
POWER7
Core
256 KB L2
POWER7
Core
256 KB L2
POWER7
Core
256 KB L2
POWER7
Core
256 KB L2
32 MB L3 Cache
Memory Controller Memory Controller
SMPFabric
Port Port Port Port
Ctrl
DIMM
Ctrl
DIMM
Ctrl
DIMM
Ctrl
DIMM
Buffer Chip Buffer Chip Buffer Chip Buffer Chip

Finally, if an uncorrectable error in memory is discovered, the logical memory block
associated with the address with the uncorrectable error is marked for deallocation by the
POWER Hypervisor. This deallocation takes effect on a partition reboot if the logical memory
block is assigned to an active partition at the time of the fault.
In addition, the system deallocates the entire memory group that is associated with the error
on all subsequent system reboots until the memory is repaired. This way is intended to guard
against future uncorrectable errors while waiting for parts replacement.
Memory persistent deallocation
Defective memory that is discovered at boot time is automatically switched off. If the service
processor detects a memory fault at boot time, it marks the affected memory as bad so that it
is not to be used on subsequent reboots.
If the service processor identifies faulty memory in a server that includes CoD memory, the
POWER Hypervisor attempts to replace the faulty memory with available CoD memory.
Faulty resources are marked as deallocated, and working resources are included in the active
memory space. Because these activities reduce the amount of CoD memory available for
future use, schedule repair of the faulty memory as soon as convenient.
Upon reboot, if not enough memory is available to meet minimum partition requirements, the
POWER Hypervisor reduces the capacity of one or more partitions.
Depending on the configuration of the system, the HMC Service Focal Point™, the OS
Service Focal Point, or the service processor receives a notification of the failed component
and triggers a service call.
4.2.4 Active Memory Mirroring for Hypervisor
Active Memory Mirroring (AMM) for Hypervisor is a hardware and firmware function of
Power 770 and Power 780 systems that provides the ability of the POWER7 chip to create
two copies of data in memory. Having two copies eliminates a system-wide outage due to an
uncorrectable failure of a single DIMM in the main memory used by the hypervisor (also
called System firmware). This capability is standard and enabled by default on the Power 780
server. On the Power 770 it is an optional chargeable feature.
What memory is mirrored
These are the areas of memory that are mirrored:
Hypervisor data that is mirrored
– Hardware Page Tables (HPTs) that are managed by the hypervisor on behalf of
partitions to track the state of the memory pages assigned to the partition
– Translation control entries (TCEs) that are managed by the hypervisor on behalf of
partitions to communicate with partition I/O buffers for I/O devices
– Hypervisor code (instructions that make up the hypervisor kernel)
– Memory used by hypervisor to maintain partition configuration, I/O states, Virtual I/O
information, partition state, and so on
Note: Memory page deallocation handles single cell failures, but because of the sheer size
of data in a data bit line, it might be inadequate for dealing with more catastrophic failures.

Hypervisor data that is not mirrored
– Advanced Memory Sharing (AMS) pool
– Memory used to hold contents of platform dump while waiting for offload to
management console
Partition data that is not mirrored
– Desired memory configured for individual partitions is not mirrored.
To enable mirroring, the requirement is to have eight equally sized functional memory DIMMs
behind at least one POWER7 chip in each CEC enclosure. The DIMMs will be managed by
the same memory controller. The sizes of DIMMs might be different from one Power 7 chip
to another.
A write operation in the memory begins on the first DIMM of a mirrored DIMM pair. When
this write is complete, the POWER7 chip writes the same data to a second DIMM of the
DIMM pair.
The read operations alternate between both DIMMs.
Figure 4-4 shows the hardware implementation of Memory Mirroring for Hypervisor.
Figure 4-4 Hardware implementation of Memory Mirroring for Hypervisor
The impact on performance is very low. Whereas writes operations are slightly slower
because two writes are actually done, reads are faster because two sources for the data are
used. Measured commercial workloads show no gain or loss in performance due to mirroring.
HPC workload performing huge amounts of string manipulation might see a slight
performance effect.
DIMMs
Memory controller 0
Power7 processor
chip
0
1
2
3
4
5
6
7
Mirrored Data
Data
Memory controller 1

The Active Memory Mirroring can be disabled or enabled on the management console using
the Advanced tab of the server properties (Figure 4-5).
Figure 4-5 Enabling or disabling active memory sharing
The system must be entirely powered off and then powered on to change from mirroring
mode to non-mirrored mode.
This same frame also gives informations about the mirroring status:
Desired mirroring mode: Takes the values “Off” or “System firmware only”
System firmware mirroring status
– Fully mirrored: The mirroring is completely functional.
– Partially functional: Due to uncorrectable memory failures, some of the hypervisor
elements or objects are not mirrored. The system remains partially mirrored until DIMM
is replaced and the system is rebooted.
– Not mirrored: At the last power on of the system, the desired state was “mirroring off.’
Mirrorable memory: Total amount of physical memory that can be mirrored, which is
based on the DIMMs that are plugged
Mirrored memory in use
Available mirrored memory

Mirroring optimization
Hypervisor mirroring requires specific memory locations. Those locations might be assign to
other purposes (for LPAR memory, for example) due to memory’s management based on the
logical memory block. To “reclaim” those memory locations, an optimization tool is available
on the Advanced tab of the system properties (Figure 4-6).
Figure 4-6 Optimization Tool
You can define the amount of memory available for mirroring by either selecting a custom
value or making available as much mirrorable memory as possible. After selecting OK, this
action copies the active partition’s contents from one LMB to another to free pairs of mirrored
memory. The copy operation will have a slight impact on performance while in progress.
The operation can be stopped by selecting Cancel. A time limit can also be specified.
DIMM guard at system boot
During system boot the FSP will guard a failing DIMM. Because there will not be eight
functional DIMMs behind a memory controller, hypervisor mirroring is not possible on this
chip. Then at boot time:
If there are other chips in the book with mirrorable memory, the system will boot
fully mirrored.
If this was the only mirrorable memory in this book, hypervisor enters a partially mirrored
state. Not all of the hypervisor objects are mirrored, and therefore are unprotected.
Hypervisor will continue to mirror as much as possible to continue to provide protection. If
a second uncorrectable error occurs in the same CEC while in partial mirror state, this will
likely result in system failure. The system remains partially mirrored until the DIMM is
replaced and the CEC is rebooted.

Advanced memory mirroring features
On the Power 770 server, the Advanced Memory Mirroring for Hypervisor function is an
optional chargable feature. It must be selected in econfig.
On this server, the advanced memory mirroring is activated by entering an activation code
(also called Virtualization Technology Code, or VET) in the management console. If the
customer enables mirroring from the management console without entering the activation
code, the system boots only to standby and will wait for the customer to enter the VET code
(New SRC A700474A displays). If mirroring was enabled by mistake, you must disable it and
power cycle the CEC, as mirroring state requires a CEC reboot to change. Hypervisor
mirroring is disabled by default on the Power 770 server.
On the Power 780 server, this feature is standard. There is no individual feature code in
econfig. The mirroring is enabled by default on the server.
4.2.5 Cache protection
POWER7 processor-based systems are designed with cache protection mechanisms,
including cache-line delete in both L2 and L3 arrays, Processor Instruction Retry and
Alternate Processor Recovery protection on L1-I and L1-D, and redundant Repair bits in L1-I,
L1-D, and L2 caches, and in L2 and L3 directories.
L1 instruction and data array protection
The POWER7 processor’s instruction and data caches are protected against intermittent
errors by using Processor Instruction Retry and against permanent errors by Alternate
Processor Recovery, both mentioned previously. L1 cache is divided into sets. POWER7
processor can deallocate all but one set before doing a Processor Instruction Retry.
In addition, faults in the Segment Lookaside Buffer (SLB) array are recoverable by the
POWER Hypervisor. The SLB is used in the core to perform address translation calculations.
L2 and L3 array protection
The L2 and L3 caches in the POWER7 processor are protected with double-bit detect
single-bit correct error detection code (ECC). Single-bit errors are corrected before being
forwarded to the processor and are subsequently written back to L2 and L3.
In addition, the caches maintain a cache-line delete capability. A threshold of correctable
errors detected on a cache line can result in the data in the cache line being purged and the
cache line removed from further operation without requiring a reboot. An ECC uncorrectable
error detected in the cache can also trigger a purge and deleting of the cache line. This
results in no loss of operation because an unmodified copy of the data can be held on system
memory to reload the cache line from main memory. Modified data is handled through
Special Uncorrectable Error handling.
L2-deleted and L3-deleted cache lines are marked for persistent deconfiguration on
subsequent system reboots until the processor card can be replaced.
4.2.6 Special uncorrectable error handling
Although rare, an uncorrectable data error can occur in memory or a cache. IBM POWER7
processor-based systems attempt to limit, to the least possible disruption, the impact of an
uncorrectable error using a well-defined strategy that first considers the data source.

Sometimes an uncorrectable error is temporary in nature and occurs in data that can be
recovered from another repository. For example:
Data in the instruction L1 cache is never modified within the cache itself. Therefore, an
uncorrectable error discovered in the cache is treated like an ordinary cache-miss, and
correct data is loaded from the L2 cache.
The L2 and L3 cache of the POWER7 processor-based systems can hold an unmodified
copy of data in a portion of main memory. In this case, an uncorrectable error simply
triggers a reload of a cache line from main memory.
In cases where the data cannot be recovered from another source, a technique called Special
Uncorrectable Error (SUE) handling is used to prevent an uncorrectable error in memory or
cache from immediately causing the system to terminate. Instead, the system tags the data
and determines whether it can ever be used again.
If the error is irrelevant, it does not force a checkstop.
If the data is used, termination can be limited to the program, kernel, or hypervisor owning
the data, or a freezing of the I/O adapters that are controlled by an I/O hub controller if
data is to be transferred to an I/O device.
When an uncorrectable error is detected, the system modifies the associated ECC word,
thereby signaling to the rest of the system that the standard ECC is no longer valid. The
service processor is then notified and takes appropriate actions. When running AIX V5.2 (or
later) or Linux, and a process attempts to use the data, the operating system is informed of
the error and might terminate, or only terminate a specific process associated with the corrupt
data, depending on the operating system and firmware level and whether the data was
associated with a kernel or non-kernel process.
Only when the corrupt data is being used by the POWER Hypervisor can the entire system be
rebooted, thereby preserving overall system integrity. If Active Memory Mirroring is enabled,
the entire system is protected and continues to run.
Depending on the system configuration and the source of the data, errors encountered during
I/O operations might not result in a machine check. Instead, the incorrect data is handled by
the PCI host bridge (PHB) chip. When the PHB chip detects a problem, it rejects the data,
preventing data from being written to the I/O device. The PHB then enters a freeze mode,
halting normal operations. Depending on the model and type of I/O being used, the freeze
can include the entire PHB chip, or simply a single bridge, resulting in the loss of all I/O
operations that use the frozen hardware until a power-on reset of the PHB. The impact to
partitions depends on how the I/O is configured for redundancy. In a server that is configured
for fail-over availability, redundant adapters spanning multiple PHB chips can enable the
system to recover transparently, without partition loss.
4.2.7 PCI enhanced error handling
IBM estimates that PCI adapters can account for a significant portion of the hardware-based
errors on a large server. Although servers that rely on boot-time diagnostics can identify
failing components to be replaced by hot-swap and reconfiguration, runtime errors pose a
more significant problem.
PCI adapters are generally complex designs involving extensive on-board instruction
processing, often on embedded microcontrollers. They tend to use industry standard grade
components with an emphasis on product cost that is relative to high reliability. In certain
cases, they might be more likely to encounter internal microcode errors or many of the
hardware errors described for the rest of the server.

The traditional means of handling these problems is through adapter internal-error reporting
and recovery techniques, in combination with operating system device-driver management
and diagnostics. In certain cases, an error in the adapter can cause transmission of bad data
on the PCI bus itself, resulting in a hardware-detected parity error and causing a global
machine check interrupt, eventually requiring a system reboot to continue.
PCI enhanced error handling-enabled adapters respond to a special data packet that is
generated from the affected PCI slot hardware by calling system firmware, which examines
the affected bus, allows the device driver to reset it, and continues without a system reboot.
For Linux, enhanced error handling (EEH) support extends to the majority of frequently used
devices, although various third-party PCI devices might not provide native EEH support.
To detect and correct PCIe bus errors, POWER7 processor-based systems use CRC
detection and instruction retry correction. For PCI-X, it uses ECC.
Figure 4-7 shows the location and mechanisms used throughout the I/O subsystem for
PCI-enhanced error handling.
Figure 4-7 PCI-enhanced error handling
4.2.8 POWER7 I/O chip freeze behavior
The POWER7 I/O chip implements a “freeze behavior” for uncorrectable errors borne on the
GX+ bus and for internal POWER7 I/O chip errors detected by the POWER7 I/O chip. With
this freeze behavior, the chip refuses I/O requests to the attached I/O, but does not check
stop the system. This allows systems with redundant I/O to continue operating without an
outage instead of system checkstops seen in earlier chips, such as the POWER5 I/O chip
used on POWER6 processor-based systems.
PCIe
Adapter
PCI-X
Adapter
Parity error
Parity error
I/O drawer concurrent add
CRC with
retry or ECC
PCI Bridge Enhanced
Error Handling
PCI-X to PCI-X
POWER7
12X Channel
Hub
PCI-X
Bridge
PCI-X
Bridge
POWER7
12X Channel
Hub
12X Channel –
PCIe Bridge
GX+ / GX++ bus
adapter
12x channel failover
support
PCI Bus Enhanced Error
Handling

4.3 Serviceability
IBM Power Systems design considers both IBM and client needs. The IBM Serviceability
Team has enhanced the base service capabilities and continues to implement a strategy that
incorporates best-of-breed service characteristics from diverse IBM systems offerings.
Serviceability includes system installation, system upgrades and downgrades (MES), and
system maintenance and repair.
The goal of the IBM Serviceability Team is to design and provide the most efficient system
service environment that includes:
Easy access to service components, design for Customer Set Up (CSU), Customer
Installed Features (CIF), and Customer Replaceable Units (CRU)
On demand service education
Error detection and fault isolation (ED/FI)
First-failure data capture (FFDC)
An automated guided repair strategy that uses common service interfaces for a converged
service approach across multiple IBM server platforms
By delivering on these goals, IBM Power Systems servers enable faster and more accurate
repair and reduce the possibility of human error.
Client control of the service environment extends to firmware maintenance on all of the
POWER processor-based systems. This strategy contributes to higher systems availability
with reduced maintenance costs.
This section provides an overview of the progressive steps of error detection, analysis,
reporting, notification, and repairing that are found in all POWER processor-based systems.
4.3.1 Detecting
The first and most crucial component of a solid serviceability strategy is the ability to
accurately and effectively detect errors when they occur. Although not all errors are a
guaranteed threat to system availability, those that go undetected can cause problems
because the system does not have the opportunity to evaluate and act if necessary. POWER
processor-based systems employ System z® server-inspired error detection mechanisms
that extend from processor cores and memory to power supplies and hard drives.
Service processor
The service processor is a microprocessor that is powered separately from the main
instruction processing complex. The service processor provides the capabilities for:
POWER Hypervisor (system firmware) and Hardware Management Console
connection surveillance
Several remote power control options
Reset and boot features
Environmental monitoring
The service processor monitors the server’s built-in temperature sensors, sending
instructions to the system fans to increase rotational speed when the ambient temperature
is above the normal operating range. Using an architected operating system interface, the
service processor notifies the operating system of potential environmentally related

problems so that the system administrator can take appropriate corrective actions before
a critical failure threshold is reached.
The service processor can also post a warning and initiate an orderly system shutdown in
the following circumstances:
– The operating temperature exceeds the critical level (for example, failure of air
conditioning or air circulation around the system).
– The system fan speed is out of operational specification (for example, because of
multiple fan failures).
– The server input voltages are out of operational specification.
The service processor can immediately shut down a system in the
following circumstances:
– Temperature exceeds the critical level or remains above the warning level for too long.
– Internal component temperatures reach critical levels.
– Non-redundant fan failures occur.
Placing calls
On systems without a Hardware Management Console, the service processor can place
calls to report surveillance failures with the POWER Hypervisor, critical environmental
faults, and critical processing faults even when the main processing unit is inoperable.
Mutual surveillance
The service processor monitors the operation of the POWER Hypervisor firmware during
the boot process and watches for loss of control during system operation. It also allows
the POWER Hypervisor to monitor service processor activity. The service processor can
take appropriate action, including calling for service, when it detects that the POWER
Hypervisor firmware has lost control. Likewise, the POWER Hypervisor can request a
service processor repair action if necessary.
Availability
The auto-restart (reboot) option, when enabled, can reboot the system automatically
following an unrecoverable firmware error, firmware hang, hardware failure, or
environmentally induced (AC power) failure.

Figure 4-8 ASMI Auto Power Restart setting panel
Fault monitoring
Built-in self-test (BIST) checks processor, cache, memory, and associated hardware that
is required for proper booting of the operating system, when the system is powered on at
the initial installation or after a hardware configuration change (for example, an upgrade).
If a non-critical error is detected or if the error occurs in a resource that can be removed
from the system configuration, the booting process is designed to proceed to completion.
The errors are logged in the system nonvolatile random access memory (NVRAM). When
the operating system completes booting, the information is passed from the NVRAM to the
system error log where it is analyzed by error log analysis (ELA) routines. Appropriate
actions are taken to report the boot-time error for subsequent service, if required.
Note: The auto-restart (reboot) option has to be enabled from the Advanced System
Manager Interface or from the Control (Operator) Panel. Figure 4-8 shows this option
using the ASMI.

Concurrent access to the service processors menus of the Advanced System
Management Interface (ASMI)
This access allows nondisruptive abilities to change system default parameters,
interrogate service processor progress and error logs andset and reset server indicators
(Guiding Light for midrange and high-end servers, Light Path for low-end servers),
accessing all service processor functions without having to power down the system to the
standby state. This allows the administrator or service representative to dynamically
access the menus from any eeb browser-enabled console that is attached to the Ethernet
service network, concurrently with normal system operation.
Managing the interfaces for connecting uninterruptible power source systems to the
POWER processor-based systems, performing Timed Power-On (TPO) sequences, and
interfacing with the power and cooling subsystem
Error checkers
IBM POWER processor-based systems contain specialized hardware detection circuitry that
is used to detect erroneous hardware operations. Error checking hardware ranges from parity
error detection coupled with processor instruction retry and bus retry, to ECC correction on
caches and system buses. All IBM hardware error checkers have distinct attributes:
Continuous monitoring of system operations to detect potential calculation errors.
Attempts to isolate physical faults based on run time detection of each unique failure.
Ability to initiate a wide variety of recovery mechanisms designed to correct the problem.
The POWER processor-based systems include extensive hardware and firmware
recovery logic.
Fault isolation registers
Error checker signals are captured and stored in hardware fault isolation registers (FIRs). The
associated logic circuitry is used to limit the domain of an error to the first checker that
encounters the error. In this way, runtime error diagnostics can be deterministic so that for
every check station, the unique error domain for that checker is defined and documented.
Ultimately, the error domain becomes the field-replaceable unit (FRU) call, and manual
interpretation of the data is not normally required.
First-failure data capture
FFDC is an error isolation technique, which ensures that when a fault is detected in a
system through error checkers or other types of detection methods, the root cause of the fault
will be captured without the need to re-create the problem or run an extended tracing or
diagnostics program.
For the vast majority of faults, a good FFDC design means that the root cause is detected
automatically without intervention by a service representative. Pertinent error data related to
the fault is captured and saved for analysis. In hardware, FFDC data is collected from the
fault isolation registers and from the associated logic. In firmware, this data consists of return
codes, function calls, and so forth.
FFDC check stations are carefully positioned within the server logic and data paths to
ensure that potential errors can be quickly identified and accurately tracked to a
field-replaceable unit (FRU).
This proactive diagnostic strategy is a significant improvement over the classic, less accurate
reboot and diagnose service approaches.

Figure 4-9 shows a schematic of a fault isolation register implementation.
Figure 4-9 Schematic of FIR implementation
Fault isolation
The service processor interprets error data that is captured by the FFDC checkers (saved in
the FIRs or other firmware-related data capture methods) to determine the root cause of the
error event.
Root cause analysis might indicate that the event is recoverable, meaning that a service
action point or need for repair has not been reached. Alternatively, it could indicate that a
service action point has been reached, where the event exceeded a pre-determined
threshold or was unrecoverable. Based on the isolation analysis, recoverable error
threshold counts can be incremented. No specific service action is necessary when the
event is recoverable.
When the event requires a service action, additional required information is collected to
service the fault. For unrecoverable errors or for recoverable events that meet or exceed their
service threshold, meaning that a service action point has been reached, a request for
service is initiated through an error logging component.
4.3.2 Diagnosing
Using the extensive network of advanced and complementary error detection logic that is built
directly into hardware, firmware, and operating systems, the IBM Power Systems servers can
perform considerable self-diagnosis.
Memory
CPU
L2 / L3
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
L1
Disk
Text
Text
Text
Text
Text
Text
Text
Text
Non-volatile
RAM
Service
Processor
Error checkers
Text Fault isolation register (FIR)
Unique fingerprint of each
captured error
Log error

Boot time
When an IBM Power Systems server powers up, the service processor initializes the system
hardware. Boot-time diagnostic testing uses a multi-tier approach for system validation,
starting with managed low-level diagnostics that are supplemented with system firmware
initialization and configuration of I/O hardware, followed by OS-initiated software test
routines. Boot-time diagnostic routines include:
Built-in self-tests (BISTs) for both logic components and arrays ensure the internal
integrity of components. Because the service processor assists in performing these tests,
the system is enabled to perform fault determination and isolation, whether or not the
system processors are operational. Boot-time BISTs can also find faults undetectable by
processor-based power-on self-test (POST) or diagnostics.
Wire-tests discover and precisely identify connection faults between components such as
processors, memory, or I/O hub chips.
Initialization of components such as ECC memory, typically by writing patterns of data and
allowing the server to store valid ECC data for each location, can help isolate errors.
To minimize boot time, the system determines which of the diagnostics are required to be
started to ensure correct operation, based on the way that the system was powered off, or on
the boot-time selection menu.
Run time
All Power Systems servers can monitor critical system components during run time, and they
can take corrective actions when recoverable faults occur. IBM hardware error-check
architecture provides the ability to report non-critical errors in an out-of-band communications
path to the service processor without affecting system performance.
A significant part of IBM runtime diagnostic capabilities originate with the service processor.
Extensive diagnostic and fault analysis routines have been developed and improved over
many generations of POWER processor-based servers, and enable quick and accurate
predefined responses to both actual and potential system problems.
The service processor correlates and processes runtime error information using logic
derived from IBM engineering expertise to count recoverable errors (called thresholding) and
predict when corrective actions must be automatically initiated by the system. These actions
can include:
Requests for a part to be replaced
Dynamic invocation of built-in redundancy for automatic replacement of a failing part
Dynamic deallocation of failing components so that system availability is maintained
Device drivers
In certain cases diagnostics are best performed by operating system-specific drivers, most
notably I/O devices that are owned directly by a logical partition. In these cases, the operating
system device driver often works in conjunction with I/O device microcode to isolate and
recover from problems. Potential problems are reported to an operating system device driver,
which logs the error. I/O devices can also include specific exercisers that can be invoked by
the diagnostic facilities for problem recreation if required by service procedures.
4.3.3 Reporting
In the unlikely event that a system hardware or environmentally induced failure is diagnosed,
IBM Power Systems servers report the error through a number of mechanisms. The analysis

result is stored in system NVRAM. Error log analysis (ELA) can be used to display the failure
cause and the physical location of the failing hardware.
With the integrated service processor, the system has the ability to automatically send out an
alert through a phone line to a pager, or call for service in the event of a critical system failure.
A hardware fault also illuminates the amber system fault LED located on the system unit to
alert the user of an internal hardware problem.
On POWER7 processor-based servers, hardware and software failures are recorded in the
system log. When a management console is attached, an ELA routine analyzes the error,
forwards the event to the Service Focal Point (SFP) application running on the management
console, and has the capability to notify the system administrator that it has isolated a likely
cause of the system problem. The service processor event log also records unrecoverable
checkstop conditions, forwards them to the Service Focal Point (SFP) application, and
notifies the system administrator. After the information is logged in the SFP application, if the
system is properly configured, a call-home service request is initiated and the pertinent failure
data with service parts information and part locations is sent to the IBM service
organization.This information will also contain the client contact information as defined in the
Electronic Service Agent (ESA) guided set-up wizard.
Error logging and analysis
When the root cause of an error has been identified by a fault isolation component, an error
log entry is created with basic data such as:
An error code uniquely describing the error event
The location of the failing component
The part number of the component to be replaced, including pertinent data such as
engineering and manufacturing levels
Return codes
Resource identifiers
FFDC data
Data containing information about the effect that the repair will have on the system is also
included. Error log routines in the operating system and FSP can then use this information
and decide whether the fault is a call home candidate. If the fault requires support
intervention, then a call will be placed with service and support and a notifcation sent to the
contact defined in the ESA guided set-up wizard
Remote support
The Remote Management and Control (RMC) subsystem is delivered as part of the base
operating system, including the operating system running on the Hardware Management
Console. RMC provides a secure transport mechanism across the LAN interface between the
operating system and the Hardware Management Console and is used by the operating
system diagnostic application for transmitting error information. It performs a number of other
functions also, but these are not used for the service infrastructure.
Service Focal Point
A critical requirement in a logically partitioned environment is to ensure that errors are not lost
before being reported for service, and that an error should only be reported once, regardless
of how many logical partitions experience the potential effect of the error. The Manage
Serviceable Events task on the management console is responsible for aggregating duplicate
error reports, and ensures that all errors are recorded for review and management.

When a local or globally reported service request is made to the operating system, the
operating system diagnostic subsystem uses the Remote Management and Control
Subsystem (RMC) to relay error information to the Hardware Management Console. For
global events (platform unrecoverable errors, for example) the service processor will also
forward error notification of these events to the Hardware Management Console, providing a
redundant error-reporting path in case of errors in the RMC network.
The first occurrence of each failure type is recorded in the Manage Serviceable Events task
on the management console. This task then filters and maintains a history of duplicate reports
from other logical partitions on the service processor. It then looks at all active service event
requests, analyzes the failure to ascertain the root cause and, if enabled, initiates a call home
for service. This methodology ensures that all platform errors will be reported through at least
one functional path, ultimately resulting in a single notification for a single problem.
Extended error data
Extended error data (EED) is additional data that is collected either automatically at the time
of a failure or manually at a later time. The data collected is dependent on the invocation
method but includes information like firmware levels, operating system levels, additional fault
isolation register values, recoverable error threshold register values, system status, and any
other pertinent data.
The data is formatted and prepared for transmission back to IBM to assist the service support
organization with preparing a service action plan for the service representative or for
additional analysis.
System dump handling
In certain circumstances, an error might require a dump to be automatically or manually
created. In this event, it is off-loaded to the management console. Specific management
console information is included as part of the information that can optionally be sent to IBM
support for analysis. If additional information relating to the dump is required, or if it becomes
necessary to view the dump remotely, the management console dump record notifies the IBM
support center regarding on which management console the dump is located.
4.3.4 Notifying
After a Power Systems server has detected, diagnosed, and reported an error to an
appropriate aggregation point, it then takes steps to notify the client, and if necessary the IBM
support organization. Depending on the assessed severity of the error and support
agreement, this could range from a simple notification to having field service personnel
automatically dispatched to the client site with the correct replacement part.
Client Notify
When an event is important enough to report, but does not indicate the need for a repair
action or the need to call home to IBM service and support, it is classified as Client Notify.
Clients are notified because these events might be of interest to an administrator. The event
might be a symptom of an expected systemic change, such as a network reconfiguration or
failover testing of redundant power or cooling systems. Examples of these events include:
Network events such as the loss of contact over a local area network (LAN)
Environmental events such as ambient temperature warnings
Events that need further examination by the client (although these events do not
necessarily require a part replacement or repair action)

Client Notify events are serviceable events, by definition, because they indicate that
something has happened that requires client awareness in the event that the client
wants to take further action. These events can always be reported back to IBM at the
client’s discretion.
Call home
A correctly configured POWER processor-based system can initiate an automatic or manual
call from a client location to the IBM service and support organization with error data, server
status, or other service-related information. The call-home feature invokes the service
organization in order for the appropriate service action to begin, automatically opening a
problem report and, in certain cases, also dispatching field support. This automated reporting
provides faster and potentially more accurate transmittal of error information. Although
configuring call-home is optional, clients are strongly encouraged to configure this feature to
obtain the full value of IBM service enhancements.
Vital product data (VPD) and inventory management
Power Systems store vital product data (VPD) internally, which keeps a record of how much
memory is installed, how many processors are installed, the manufacturing level of the parts,
and so on. These records provide valuable information that can be used by remote support
and service representatives, enabling them to provide assistance in keeping the firmware and
software on the server up-to-date.
IBM problem management database
At the IBM support center, historical problem data is entered into the IBM Service and
Support Problem Management database. All of the information that is related to the error,
along with any service actions taken by the service representative, is recorded for problem
management by the support and development organizations. The problem is then tracked
and monitored until the system fault is repaired.
4.3.5 Locating and servicing
The final component of a comprehensive design for serviceability is the ability to effectively
locate and replace parts requiring service. POWER processor-based systems use a
combination of visual cues and guided maintenance procedures to ensure that the identified
part is replaced correctly, every time.
Packaging for service
The following service enhancements are included in the physical packaging of the systems to
facilitate service:
Color coding (touch points)
– Terra-cotta-colored touch points indicate that a component (FRU or CRU) can be
concurrently maintained.
– Blue-colored touch points delineate components that are not concurrently maintained
(those that require the system to be turned off for removal or repair).
Tool-less design: Selected IBM systems support tool-less or simple tool designs. These
designs require no tools or simple tools, such as flathead screw drivers to service the
hardware components.
Positive retention: Positive retention mechanisms help to ensure proper connections
between hardware components, such as from cables to connectors, and between two
cards that attach to each other. Without positive retention, hardware components run the
risk of becoming loose during shipping or installation, preventing a good electrical

connection. Positive retention mechanisms such as latches, levers, thumb-screws, pop
Nylatches (U-clips), and cables are included to help prevent loose connections and aid in
installing (seating) parts correctly. These positive retention items do not require tools.
Light Path
The Light Path LED feature is for low-end systems, including Power Systems up to models
750 and 755, that can be repaired by clients. In the Light Path LED implementation, when a
fault condition is detected on the POWER7 processor-based system, an amber FRU fault
LED is illuminated, which is then rolled up to the system fault LED. The Light Path system
pinpoints the exact part by turning on the amber FRU fault LED that is associated with the
part to be replaced.
The system can clearly identify components for replacement by using specific
component-level LEDs, and can also guide the servicer directly to the component by
signaling (staying on solid) the system fault LED, enclosure fault LED, and the component
FRU fault LED.
After the repair, the LEDs shut off automatically if the problem is fixed.
Guiding Light
Midrange and high-end systems, including models 770 and 780 and later, are usually
repaired by IBM Support personnel.
The enclosure and system identify LEDs turn on solid and can be used to follow the path from
the system to the enclosure and down to the specific FRU.
Guiding Light uses a series of flashing LEDs, allowing a service provider to quickly and easily
identify the location of system components. Guiding Light can also handle multiple error
conditions simultaneously, which might be necessary in some very complex high-end
configurations.
In these situations, Guiding Light waits for the servicer’s indication of what failure to attend
first and then illuminates the LEDs to the failing component.
Data centers can be complex places, and Guiding Light is designed to do more than identify
visible components. When a component might be hidden from view, Guiding Light can flash a
sequence of LEDs that extends to the frame exterior, clearly guiding the service
representative to the correct rack, system, enclosure, drawer, and component.

Service labels
Service providers use these labels to assist them in performing maintenance actions. Service
labels are found in various formats and positions, and are intended to transmit readily
available information to the servicer during the repair process.
Several of these service labels and their purposes are described in the following list:
Location diagrams are strategically located on the system hardware, relating information
regarding the placement of hardware components. Location diagrams can include location
codes, drawings of physical locations, concurrent maintenance status, or other data that is
pertinent to a repair. Location diagrams are especially useful when multiple components
are installed, such as DIMMs, sockets, processor cards, fans, adapter cards, LEDs, and
power supplies.
Remove or replace procedure labels contain procedures often found on a cover of the
system or in other spots that are accessible to the servicer. These labels provide
systematic procedures, including diagrams, detailing how to remove and replace certain
serviceable hardware components.
Numbered arrows are used to indicate the order of operation and serviceability direction of
components. Various serviceable parts such as latches, levers, and touch points must be
pulled or pushed in a certain direction and order so that the mechanical mechanisms can
engage or disengage. Arrows generally improve the ease of serviceability.
The operator panel
The operator panel on a POWER processor-based system is a four-row by 16-element LCD
display that is used to present boot progress codes, indicating advancement through the
system power-on and initialization processes. The operator panel is also used to display error
and location codes when an error occurs that prevents the system from booting. It includes
several buttons, enabling a service support representative (SSR) or client to change various
boot-time options and for other limited service functions.
Concurrent maintenance
The IBM POWER7 processor-based systems are designed with the understanding that
certain components have higher intrinsic failure rates than others. The movement of fans,
power supplies, and physical storage devices naturally make them more susceptible to
wearing down or burning out. Other devices such as I/O adapters can begin to wear from
repeated plugging and unplugging. For these reasons, these devices have been specifically
designed to be concurrently maintainable when properly configured.
In other cases, a client might be in the process of moving or redesigning a data center or
planning a major upgrade. At times like these, flexibility is crucial. The IBM POWER7
processor-based systems are designed for redundant or concurrently maintainable power,
fans, physical storage, and I/O towers.
The most recent members of the IBM Power Systems family, based on the POWER7
processor, continue to support concurrent maintenance of power, cooling, PCI adapters,
media devices, I/O drawers, GX adapter, and the operator panel. In addition, they support
concurrent firmware fix pack updates when possible. The determination of whether a
firmware fix pack release can be updated concurrently is identified in the readme file that is
released with the firmware.

Hot-node add, hot-node repair, and memory upgrade
With the proper configuration and required protective measures, the Power 770 and
Power 780 servers are designed for node add, node repair, or memory upgrade without
powering down the system.
The Power 770 and Power 780 servers support the addition of another CEC enclosure (node)
to a system (hot-node add) or adding more memory (memory upgrade) to an existing node.
The additional Power 770 and Power 780 enclosure or memory can be ordered as a system
upgrade (MES order) and added to the original system. The additional resources of the newly
added CEC enclosure (node) or memory can then be assigned to existing OS partitions or
new partitions as required. Hot-node add and memory upgrade enable the upgrading of a
server by integrating a second, third, or fourth CEC enclosure or additional memory into the
server, with reduced impact to the system operation.
In the unlikely event that CEC hardware (for example, processor or memory) experienced a
failure, the hardware can be repaired by freeing the processors and memory in the node and
its attached I/O resources (node evacuation) dependant on the partition configuration.
To guard against any potential impact to system operation during hot-node add, memory
upgrade, or node repair, clients must comply with these protective measures:
For memory upgrade and node repair, ensure that the system has sufficient inactive
or spare processors and memory. Critical I/O resources must be configured with
redundant paths.
Schedule upgrades or repairs during non-peak operational hours.
Move business applications to another server by using the PowerVM Live Partition
Mobility feature or quiesce them. The use of LPM means that all critical applications
must be halted or moved to another system before the operation begins. Non-critical
applications can remain running. The partitions can be left running at the operating
system command prompt.
Back up critical application and system state information.
Checkpoint the databases.
Blind-swap cassette
Blind-swap PCIe adapters represent significant service and ease-of-use enhancements in I/O
subsystem design while maintaining high PCIe adapter density.
Blind-swap allows PCIe adapters to be concurrently replaced or installed without having to
put the I/O drawer or system into a service position. Since first delivered, minor carrier design
adjustments have improved an already well-thought-out service design.
For PCIe adapters on the POWER7 processor-based servers, blind-swap cassettes include
the PCIe slot, to avoid the top to bottom movement for inserting the card on the slot that was
required on previous designs. The adapter is correctly connected by just sliding the cassette
in and actuacting a latch.
Firmware updates
System Firmware is delivered as a release level or a service pack. Release Levels support
the general availability (GA) of new function or features, and new machine types or models.
Upgrading to a higher release level is disruptive to customer operations. IBM intends to
introduce no more than two new release levels per year. These release levels will be
supported by service packs. Service packs are intended to contain only firmware fixes and
not to introduce new function. A service pack is an update to an existing release level.

If the system is managed by a management console, you will use the management console
for firmware updates. Using the management console allows you to take advantage of the
Concurrent Firmware Maintenance (CFM) option when concurrent service packs are
available. CFM is the IBM term used to describe the IBM Power Systems firmware updates
that can be partially or wholy concurrent or non-disruptive. With the introduction of CFM, IBM
is significantly increasing a client’s opportunity to stay on a given release level for longer
periods of time. Clients wanting maximum stability can defer until there is a compelling
reason to upgrade, such as:
A release level is approaching its end-of-service date (that is, it has been available for
about a year and hence will go out of service support soon).
Moving a system to a more standardized release level when there are multiple systems in
an environment with similar hardware.
A new release has new functionality that is needed in the environment.
A scheduled maintenance action will cause a platform reboot. This provides an
opportunity to also upgrade to a new firmware release.
The update and upgrade of system firmware is dependant on several factors, such as
whether the system is standalone or managed by a management console, the current
firmware installed, and what operating systems are running on the system. These scenarios
and the associated installation instructions are comprehensively outlined in the firmware
section of Fix Central:
http://www.ibm.com/support/fixcentral/
You might also want to review the best practice white papers, which can be found here:
http://www14.software.ibm.com/webapp/set2/sas/f/best/home.html
Repair and verify system
Repair and verify (R&V) is a system used to guide a service provider step-by-step through the
process of repairing a system and verifying that the problem has been repaired. The steps
are customized in the appropriate sequence for the particular repair for the specific system
being repaired. Repair scenarios covered by repair and verify include:
Replacing a defective field-replaceable unit (FRU) or a customer replacable unit (CRU)
Reattaching a loose or disconnected component
Correcting a configuration error
Removing or replacing an incompatible FRU
Updating firmware, device drivers, operating systems, middleware components, and IBM
applications after replacing a part
Repair and verify procedures can be used by both service representative providers who are
familiar with the task and those who are not. Education On Demand content is placed in the
procedure at the appropriate locations. Throughout the repair and verify procedure, repair
history is collected and provided to the Service and Support Problem Management Database
for storage with the serviceable event, to ensure that the guided maintenance procedures are
operating correctly.
If a server is managed by a management console, then many of the R&V procedures are
performed from the management console. If the FRU to be replaced is a PCI adapter or an
internal storage device, then the service action is always performed from the operating
system of the partition owning that resource.

Clients can subscribe through the subscription services to obtain the notifications about the
latest updates available for service-related documentation. The latest version of the
documentation is accessible through the internet.
4.4 Manageability
Several functions and tools help manageability and enable you to efficiently and effectively
manage your system.
4.4.1 Service user interfaces
The service interface allows support personnel or the client to communicate with the service
support applications in a server using a console, interface, or terminal. Delivering a clear,
concise view of available service applications, the service interface allows the support team
to manage system resources and service information in an efficient and effective way.
Applications available through the service interface are carefully configured and placed to
give service providers access to important service functions.
Various service interfaces are used, depending on the state of the system and its operating
environment. The primary service interfaces are:
Light Path and Guiding Light
For more information, see “Light Path” on page 178 and “Guiding Light” on page 178.
Service processor, Advanced System Management Interface (ASMI)
Operator panel
Operating system service menu
Service Focal Point on the Hardware Management Console
Service Focal Point Lite on Integrated Virtualization Manager
Service processor
The service processor is a controller that is running its own operating system. It is a
component of the service interface card.
The service processor operating system has specific programs and device drivers for the
service processor hardware. The host interface is a processor support interface that is
connected to the POWER processor. The service processor is always working, regardless of
the main system unit’s state. The system unit can be in these states:
Standby (power off)
Operating, ready to start partitions
Operating with running logical partitions
The service processor is used to monitor and manage the system hardware resources and
devices. The service processor checks the system for errors, ensuring that the connection to
the management console for manageability purposes and accepting Advanced System
Management Interface (ASMI) Secure Sockets Layer (SSL) network connections. The
service processor provides the ability to view and manage the machine-wide settings by
using the ASMI, and enables complete system and partition management from the
management console.

With two CEC enclosures and more, there are two redundant FSP, one in each of the first
CECs. While one is active, the second one is in standby mode. In case of a failure, there will
be a automatic takeover.
The service processor uses two Ethernet 10/100Mbps ports. Note the following information:
Both Ethernet ports are only visible to the service processor and can be used to attach the
server to an HMC or to access the ASMI. The ASMI options can be accessed through an
HTTP server that is integrated into the service processor operating environment.
Both Ethernet ports support only auto-negotiation. Customer selectable media speed and
duplex settings are not available.
Both Ethernet ports have a default IP address, as follows:
– Service processor Eth0 or HMC1 port is configured as 169.254.2.147.
When a redundant service processor is present, the default IP addresses are:
The functions available through service processor include:
Call Home
Advanced System Management Interface (ASMI)
Error Information (error code, PN, Location Codes) menu
View of guarded components
Limited repair procedures
Generate dump
LED Management menu
Remote view of ASMI menus
Firmware update through USB key
Advanced System Management Interface (ASMI)
ASMI is the interface to the service processor that enables you to manage the operation of
the server, such as auto-power restart, and to view information about the server, such as the
error log and vital product data. Various repair procedures require connection to the ASMI.
The ASMI is accessible through the management console. It is also accessible by using a
web browser on a system that is connected directly to the service processor (in this case,
either a standard Ethernet cable or a crossed cable) or through an Ethernet network. ASMI
can also be accessed from an ASCII terminal, but this is only available while the system is in
the platform powered-off mode.
Use the ASMI to change the service processor IP addresses or to apply certain security
policies and prevent access from undesired IP addresses or ranges.
Note: The service processor enables a system that does not boot to be analyzed. The
error log analysis can be performed from either the ASMI or the management console.

You might be able to use the service processor’s default settings. In that case, accessing the
ASMI is not necessary. To access ASMI, use one of the following methods:
Access the ASMI by using an management console.
If configured to do so, the management console connects directly to the ASMI for a
selected system from this task.
To connect to the Advanced System Management interface from an management
console, follow these steps:
a. Open Systems Management from the navigation pane.
b. From the work pane, select one or more managed systems to work with.
c. From the System Management tasks list, select Operations Advanced System
Management (ASM).
Access the ASMI by using a web browser.
At the time of writing, supported web browsers are Microsoft Internet Explorer
(Version 7.0), Mozilla Firefox (Version 2.0.0.11), and Opera (Version 9.24). Later versions
of these browsers might work but are not officially supported. The JavaScript language
and cookies must be enabled.
The web interface is available during all phases of system operation, including the initial
program load (IPL) and run time. However, several of the menu options in the web
interface are unavailable during IPL or run time to prevent usage or ownership conflicts if
the system resources are in use during that phase. The ASMI provides a Secure Sockets
Layer (SSL) web connection to the service processor. To establish an SSL connection,
open your browser using this adress:
https://<ip_address_of_service_processor>
Access the ASMI using an ASCII terminal.
The ASMI on an ASCII terminal supports a subset of the functions that are provided by the
web interface and is available only when the system is in the platform powered-off mode. The
ASMI on an ASCII console is not available during several phases of system operation, such
as the IPL and run time.
The operator panel
The service processor provides an interface to the operator panel, which is used to display
system status and diagnostic information.
Note: To make the connection through Internet Explorer, click Tools Internet Options.
Clear the Use TLS 1.0 check box, and click OK.

The operator panel can be accessed in two ways:
By using the normal operational front view.
By pulling it out to access the switches and viewing the LCD display. Figure 4-10 shows
that the operator panel on a Power 770 and Power 780 is pulled out.
Figure 4-10 Operator panel is pulled out from the chassis
Several of the operator panel features include:
A 2 x 16 character LCD display
Reset, enter, power On/Off, increment, and decrement buttons
Amber System Information/Attention, green Power LED
Blue Enclosure Identify LED on the Power 770 and Power 780
Altitude sensor
USB Port
Speaker/Beeper
The functions available through the operator panel include:
Error Information
Generate dump
View Machine Type, Model, and Serial Number
Limited set of repair functions
Operating system service menu
The system diagnostics consist of IBM i service tools, stand-alone diagnostics that are loaded
from the DVD drive, and online diagnostics (available in AIX).
Online diagnostics, when installed, are a part of the AIX or IBM i operating system on the disk
or server. They can be booted in single-user mode (service mode), run in maintenance mode,
or run concurrently (concurrent mode) with other applications. They have access to the AIX
Release Lever
(slide left to release operator panel and pull out from chassis)

error log and the AIX configuration data. IBM i has a service tools problem log, IBM i history
log (QHST), and IBM i problem log.
These are the modes:
Service mode
Requires a service mode boot of the system and enables the checking of system devices
and features. Service mode provides the most complete checkout of the system
resources. All system resources, except the SCSI adapter and the disk drives used for
paging, can be tested.
Concurrent mode
Enables the normal system functions to continue while selected resources are being
checked. Because the system is running in normal operation, certain devices might
require additional actions by the user or diagnostic application before testing can be done.
Maintenance mode
Enables the checking of most system resources. Maintenance mode provides the same
test coverage as service mode. The difference between the two modes is the way that
they are invoked. Maintenance mode requires that all activity on the operating system be
stopped. The shutdown -m command is used to stop all activity on the operating system
and put the operating system into maintenance mode.
The System Management Services (SMS) error log is accessible on the SMS menus.
This error log contains errors that are found by partition firmware when the system or
partition is booting.
The service processor’s error log can be accessed on the ASMI menus.
You can also access the system diagnostics from a Network Installation Management
(NIM) server.
The IBM i operating system and associated machine code provide Dedicated Service Tools
(DST) as part of the IBM i licensed machine code (Licensed Internal Code) and System
Service Tools (SST) as part of the IBM i operating system. DST can be run in dedicated
mode (no operating system loaded). DST tools and diagnostics are a superset of those
available under SST.
The IBM i End Subsystem (ENDSBS *ALL) command can shut down all IBM and customer
applications subsystems except the controlling subsystem QTCL. The Power Down System
(PWRDWNSYS) command can be set to power down the IBM i partition and restart the
partition in DST mode.
You can start SST during normal operations, which leaves all applications up and running
using the IBM i Start Service Tools (STRSST) command (when signed onto IBM i with the
appropriately secured user ID).
With DST and SST you can look at various logs, run various diagnostics, or take several
kinds of system dumps or other options.
Note: When you order a Power System, a DVD-ROM or DVD-RAM might be optional. An
alternate method for maintaining and servicing the system must be available if you do not
order the DVD-ROM or DVD-RAM.

Depending on the operating system, these are the service-level functions that you typically
see when using the operating system service menus:
Product activity log
Trace Licensed Internal Code
Work with communications trace
Display/Alter/Dump
Licensed Internal Code log
Main storage dump manager
Hardware service manager
Call Home/Customer Notification
Error information menu
LED management menu
Concurrent/Non-concurrent maintenance (within scope of the OS)
Managing firmware levels
– Server
– Adapter
Remote support (access varies by OS)
Service Focal Point on the Hardware Management Console
Service strategies become more complicated in a partitioned environment. The Manage
Serviceable Events task in the management console can help to streamline this process.
Each logical partition reports errors that it detects and forwards the event to the Service Focal
Point (SFP) application that is running on the management console, without determining
whether other logical partitions also detect and report the errors. For example, if one logical
partition reports an error for a shared resource, such as a managed system power supply,
other active logical partitions might report the same error.
By using the Manage Serviceable Events task in the management console, you can avoid
long lists of repetitive call-home information by recognizing that these are repeated errors and
consolidating them into one error.
In addition, you can use the Manage Serviceable Events task to initiate service functions on
systems and logical partitions, including the exchanging of parts, configuring connectivity,
and managing dumps.
4.4.2 IBM Power Systems firmware maintenance
The IBM Power Systems Client-Managed Microcode is a methodology that enables you to
manage and install microcode updates on Power Systems and associated I/O adapters.
The system firmware consists of service processor microcode, Open Firmware microcode,
SPCN microcode, and the POWER Hypervisor.
The firmware and microcode can be downloaded and installed from a management console,
a running partition, or USB port number 1 on the rear of a Power 770 and Power 780, if that
system is not managed by a management console.
Power Systems has a permanent firmware boot side, or A side, and a temporary firmware
boot side, or B side. New levels of firmware must be installed on the temporary side first in
order to test the update’s compatibility with existing applications. When the new level of
firmware has been approved, it can be copied to the permanent side.

For access to the initial web pages that address this capability, see the Support for
IBM Systems web page:
http://www.ibm.com/systems/support
For Power Systems, select the Power link (Figure 4-11).
Figure 4-11 Support for Power servers web page
Although the content under the Popular links section can change, click Firmware and HMC
updates to go to the resources for keeping your system’s firmware current.

If there is a management console to manage the server, the management console interface
can be use to view the levels of server firmware and power subsystem firmware that are
installed and are available to download and install.
Each IBM Power Systems server has the following levels of server firmware and power
subsystem firmware:
Installed level
This level of server firmware or power subsystem firmware has been installed and will be
installed into memory after the managed system is powered off and then powered on. It is
installed on the temporary side of system firmware.
Activated level
This level of server firmware or power subsystem firmware is active and running
in memory.
Accepted level
This level is the backup level of server or power subsystem firmware. You can return to
this level of server or power subsystem firmware if you decide to remove the installed
level. It is installed on the permanent side of system firmware.
IBM provides the Concurrent Firmware Maintenance (CFM) function on selected Power
Systems. This function supports applying nondisruptive system firmware service packs to the
system concurrently (without requiring a reboot operation to activate changes). For systems
that are not managed by an management console, the installation of system firmware is
always disruptive.
The concurrent levels of system firmware can, on occasion, contain fixes that are known as
deferred. These deferred fixes can be installed concurrently but are not activated until the
next IPL. For deferred fixes within a service pack, only the fixes in the service pack, which
cannot be concurrently activated, are deferred. Table 4-1 shows the file-naming convention
for system firmware.
Table 4-1 Firmware naming convention
The following example uses the convention:
01AM710_086_063 = Managed System Firmware for 9117-MMB Release 710 Fixpack 086
PPNNSSS_FFF_DDD
PP Package identifier 01 -
02 -
NN Platform and class AL Low End
AM Mid Range
AS IH Server
AH High End
AP Bulk Power for IH
AB Bulk Power
SSS Release indicator
FFF Current fix pack
DDD Last disruptive fix pack

An installation is disruptive if the following statements are true:
The release levels (SSS) of currently installed and new firmware differ.
The service pack level (FFF) and the last disruptive service pack level (DDD) are equal in
new firmware.
Otherwise, an installation is concurrent if the service pack level (FFF) of the new firmware is
higher than the service pack level currently installed on the system and the conditions for
disruptive installation are not met.
4.4.3 Electronic Services and Electronic Service Agent
IBM has transformed its delivery of hardware and software support services to help you
achieve higher system availability. Electronic Services is a web-enabled solution that offers
an exclusive, no-additional-charge enhancement to the service and support available for IBM
servers. These services provide the opportunity for greater system availability with faster
problem resolution and preemptive monitoring. The Electronic Services solution consists of
two separate, but complementary, elements:
Electronic Services news page
The Electronic Services news page is a single internet entry point that replaces the
multiple entry points that are traditionally used to access IBM internet services and
support. The news page enables you to gain easier access to IBM resources for
assistance in resolving technical problems.
Electronic Service Agent
The Electronic Service Agent is software that resides on your server. It monitors events
and transmits system inventory information to IBM on a periodic, client-defined timetable.
The Electronic Service Agent automatically reports hardware problems to IBM.
Early knowledge about potential problems enables IBM to deliver proactive service that can
result in higher system availability and performance. In addition, information that is collected
through the Service Agent is made available to IBM service support representatives when
they help answer your questions or diagnose problems. Installation and use of IBM Electronic
Service Agent for problem reporting enables IBM to provide better support and service for
your IBM server.
To learn how Electronic Services can work for you, visit:
https://www.ibm.com/support/electronic/portal
4.5 Operating system support for RAS features
Table 4-2 gives an overview of features for continuous availability that are supported by the
various operating systems running on the Power 770 and Power 780 systems.
Table 4-2 Operating system support for RAS features
RAS feature
AIX
5.3
AIX
6.1
AIX
7.1
IBM i RHEL
5.7
RHEL
6.1
SLES11
SP1
System deallocation of failing components
Dynamic Processor Deallocation X X X X X X X
Dynamic Processor Sparing X X X X X X X

Processor Instruction Retry X X X X X X X
Alternate Processor Recovery X X X X X X X
Partition Contained Checkstop X X X X X X X
Persistent processor deallocation X X X X X X X
GX++ bus persistent deallocation X X X X - - X
PCI bus extended error detection X X X X X X X
PCI bus extended error recovery X X X X Most Most Most
PCI-PCI bridge extended error handling X X X X - - -
Redundant RIO or 12x Channel link X X X X X X X
PCI card hot-swap X X X X X X X
Dynamic SP failover at run time X X X X X X X
Memory sparing with CoD at IPL time X X X X X X X
Clock failover runtime or IPL X X X X X X X
Memory availability
64-byte ECC code X X X X X X X
Hardware scrubbing X X X X X X X
CRC X X X X X X X
Chipkill X X X X X X X
L1 instruction and data array protection X X X X X X X
L2/L3 ECC and cache line delete X X X X X X X
Special uncorrectable error handling X X X X X X X
Active Memory Mirroring X X X X X X X
Fault detection and isolation
Platform FFDC diagnostics X X X X X X X
Run-time diagnostics X X X X Most Most Most
Storage Protection Keys - X X X - - -
Dynamic Trace X X X X - - X
Operating System FFDC - X X X - - -
Error log analysis X X X X X X X
Freeze mode of I/O Hub X X X X - - -
Service Processor support for:
Built-in-Self-Tests (BIST) for logic and arrays X X X X X X X
Wire tests X X X X X X X
RAS feature
AIX
5.3
AIX
6.1
AIX
7.1
IBM i RHEL
5.7
RHEL
6.1
SLES11
SP1

Component initialization X X X X X X X
Serviceability
Boot-time progress indicators X X X X Most Most Most
Electronic Service Agent Call Home from management
console
X X X X X X X
Firmware error codes X X X X X X X
Operating system error codes X X X X Most Most Most
Inventory collection X X X X X X X
Environmental and power warnings X X X X X X X
Hot-plug fans, power supplies X X X X X X X
Extended error data collection X X X X X X X
I/O drawer redundant connections X X X X X X X
I/O drawer hot add and concurrent repair X X X X X X X
Concurrent RIO/GX adapter add X X X X X X X
SP mutual surveillance with POWER Hypervisor X X X X X X X
Dynamic firmware update with management console X X X X X X X
Electronic Service Agent Call Home Application X X X X - - -
Guiding light LEDs X X X X X X X
System dump for memory, POWER Hypervisor, SP X X X X X X X
Infocenter / Systems Support Site service publications X X X X X X X
System Support Site education X X X X X X X
Operating system error reporting to management
console SFP
X X X X X X X
RMC secure error transmission subsystem X X X X X X X
Health check scheduled operations with management
console
X X X X X X X
Operator panel (real or virtual) X X X X X X X
Concurrent operator panel maintenance X X X X X X X
Redundant management consoles X X X X X X X
Automated server recovery/restart X X X X X X X
High availability clustering support X X X X X X X
Repair and Verify Guided Maintenance X X X X Most Most Most
Concurrent kernel update - X X X X X X
Concurrent Hot Add/Repair Maintenance X X X X X X X
RAS feature
AIX
5.3
AIX
6.1
AIX
7.1
IBM i RHEL
5.7
RHEL
6.1
SLES11
SP1

Related publications
The publications listed in this section are considered particularly suitable for a more detailed
discussion of the topics covered in this paper.
IBM Redbooks
The following IBM Redbooks publications provide additional information about the topic in this
document. Note that some publications referenced in this list might be available in softcopy
only.
IBM BladeCenter PS700, PS701, and PS702 Technical Overview and Introduction,
REDP-4655
IBM BladeCenter PS703 and PS704 Technical Overview and Introduction, REDP-4744
IBM Power 710 and 730 Technical Overview and Introduction, REDP-4796
IBM Power 795 Technical Overview and Introduction, REDP-4640
IBM PowerVM Virtualization Introduction and Configuration, SG24-7940
IBM PowerVM Virtualization Managing and Monitoring, SG24-7590
IBM PowerVM Live Partition Mobility, SG24-7460
IBM System p Advanced POWER Virtualization (PowerVM) Best Practices, REDP-4194
PowerVM Migration from Physical to Virtual Storage, SG24-7825
IBM System Storage DS8000: Copy Services in Open Environments, SG24-6788
IBM System Storage DS8700 Architecture and Implementation, SG24-8786
PowerVM and SAN Copy Services, REDP-4610
SAN Volume Controller V4.3.0 Advanced Copy Services, SG24-7574
You can search for, view, download or order these documents and other Redbooks,
Redpapers, Web Docs, draft and additional materials, at the following website:
ibm.com/redbooks
Other publications
These publications are also relevant as further information sources:
IBM Power Systems Facts and Features POWER7 Blades and Servers
http://www.ibm.com/systems/power/hardware/reports/factsfeatures.html
Specific storage devices supported for Virtual I/O Server
http://www14.software.ibm.com/webapp/set2/sas/f/vios/documentation/datasheet.html

IBM Power 710 server Data Sheet
http://public.dhe.ibm.com/common/ssi/ecm/en/pod03048usen/POD03048USEN.PDF
Active Memory Expansion: Overview and Usage Guide
http://public.dhe.ibm.com/common/ssi/ecm/en/pow03037usen/POW03037USEN.PDF
Migration combinations of processor compatibility modes for active Partition Mobility
http://publib.boulder.ibm.com/infocenter/powersys/v3r1m5/topic/p7hc3/iphc3pcmco
mbosact.htm
Advance Toolchain for Linux website
http://www.ibm.com/developerworks/wikis/display/hpccentral/How+to+use+Advance+
Toolchain+for+Linux+on+POWER
Online resources
These websites are also relevant as further information sources:
IBM Power Systems Hardware Information Center
http://publib.boulder.ibm.com/infocenter/systems/scope/hw/index.jsp
IBM System Planning Tool website
IBM Fix Central website
http://www.ibm.com/support/fixcentral/
Power Systems Capacity on Demand website
http://www.ibm.com/systems/power/hardware/cod/

Related publications 195
Support for IBM Systems website
http://www.ibm.com/support/entry/portal/Overview?brandind=Hardware~Systems~Power
IBM Power Systems website
http://www.ibm.com/systems/power/
IBM Storage website
http://www.ibm.com/systems/storage/
Help from IBM
IBM Support and downloads
ibm.com/support
IBM Global Services
ibm.com/services

®
REDP-4798-00
INTERNATIONAL
TECHNICAL
SUPPORT
ORGANIZATION
BUILDING TECHNICAL
INFORMATION BASED ON
PRACTICAL EXPERIENCE
IBM Redbooks are developed
by the IBM International
Technical Support
Organization. Experts from
IBM, Customers and Partners
from around the world create
timely technical information
based on realistic scenarios.
Specific recommendations
are provided to help you
implement IT solutions more
effectively in your
environment.
For more information:
ibm.com/redbooks
Redpaper™
IBM Power 770 and 780
Technical Overview
and Introduction
Features the
9117-MMC and
9179-MHC based on
the latest POWER7
processor technology
Describes MaxCore
and TurboCore for
redefining
performance
Discusses Active
Memory Mirroring for
Hypervisor
This IBM Redpaper publication is a comprehensive guide covering the
IBM Power 770 (9117-MMC) and Power 780 (9179-MHC) servers
supporting IBM AIX, IBM i, and Linux operating systems. The goal of
this paper is to introduce the major innovative Power 770 and
Power 780 offerings and their prominent functions, including:
The IBM POWER7 processor available at frequencies of 3.3 GHz,
3.44 GHz, 3.72 GHz, and 3.92 GHz, and 4.14 GHz
The specialized IBM POWER7 Level 3 cache that provides greater
bandwidth, capacity, and reliability
The 1 Gb or 10 Gb Integrated Multifunction Card that provides two
USB ports, one serial port, and four Ethernet connectors for a
processor enclosure and does not require a PCI slot
The new Active Memory Mirroring (AMM) for Hypervisor feature
that mirrors the main memory used by the firmware
IBM PowerVM virtualization, including PowerVM Live Partition
Mobility and PowerVM Active Memory Sharing
Active Memory Expansion that provides more usable memory than
what is physically installed on the system
IBM EnergyScale technology that provides features such as power
trending, power-saving, capping of power, and thermal
measurement
Enterprise-ready reliability, serviceability, and availability
Professionals who want to acquire a better understanding of IBM
Power Systems products should read this paper. This paper expands
the current set of IBM Power Systems documentation by providing a
desktop reference that offers a detailed technical description of the
Power 770 and Power 780 systems.
Back cover

IBM Power 770 and 780 Technical Overview and Introduction

More Related Content

What's hot (19)

Similar to IBM Power 770 and 780 Technical Overview and Introduction (19)

More from IBM India Smarter Computing (20)

Recently uploaded (20)

IBM Power 770 and 780 Technical Overview and Introduction