Ibm power systems e870 and e880 technical overview and introduction

ibm.com/redbooks Redpaper
Front cover
IBM Power Systems E870 and E880
Technical Overview and Introduction
Alexandre Bicas Caldeira
YoungHoon Cho
James Cruickshank
Bartłomiej Grabowski
Volker Haug
Andrew Laidlaw
Seulgi Yoppy Sung
New modular architecture for
increased reliability
Enterprise POWER8
processor-based servers
Exceptional memory and
I/O bandwidth

International Technical Support Organization
December 2014
REDP-5137-00

© Copyright International Business Machines Corporation 2014. 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 2014)
This edition applies to the IBM Power E870 (9119-MME) and Power E880 (9119-MHE) Power Systems
servers.
This document was created or updated on May 3, 2016.
Note: Before using this information and the product it supports, read the information in “Notices” on
page vii.

© Copyright IBM Corp. 2014. All rights reserved. iii
Contents
Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Authors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Now you can become a published author, too! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Comments welcome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Stay connected to IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Chapter 1. General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Systems overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1 Power E870 server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.2 Power E880 server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.3 System control unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.4 System nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.5 I/O drawers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2 Operating environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 Physical package . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4 System features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4.1 Power E870 system features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4.2 Power E880 system features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.4.3 Minimum features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.4.4 Power supply features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.4.5 Processor card features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.4.6 Summary of processor features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.4.7 Memory features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.4.8 System node PCIe slots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.5 Disk and media features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.6 I/O drawers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.6.1 PCIe Gen3 I/O expansion drawer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.6.2 I/O drawers and usable PCI slot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.7 EXP24S SFF Gen2-bay Drawer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.8 Comparison between models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.9 Build to order. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.10 IBM editions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.11 Model upgrades . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.11.1 Power E870 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.11.2 Power E880 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.11.3 Upgrade considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
1.12 Management consoles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
1.13 System racks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.13.1 IBM 7014 model T00 rack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.13.2 IBM 7014 model T42 rack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1.13.3 IBM 7953 model 94Y rack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1.13.4 Feature code #0551 rack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.13.5 Feature code #0553 rack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.13.6 The AC power distribution unit and rack content . . . . . . . . . . . . . . . . . . . . . . . . 32
Chapter 2. Architecture and technical overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

iv IBM Power Systems E870 and E880 Technical Overview and Introduction
2.1 Logical diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.2 The IBM POWER8 processor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.2.1 POWER8 processor overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.2.2 POWER8 processor core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.2.3 Simultaneous multithreading. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.2.4 Memory access. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.2.5 On-chip L3 cache innovation and Intelligent Cache . . . . . . . . . . . . . . . . . . . . . . . 49
2.2.6 Level 4 cache and memory buffer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.2.7 Hardware transactional memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2.2.8 Coherent Accelerator Processor Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2.2.9 Power management and system performance . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.2.10 Comparison of the POWER7, POWER7+, and POWER8 processors . . . . . . . . 53
2.3 Memory subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.3.1 Custom DIMM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.3.2 Memory placement rules. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.3.3 Memory activation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
2.3.4 Memory throughput. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.3.5 Active Memory Mirroring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
2.3.6 Memory Error Correction and Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.3.7 Special Uncorrectable Error handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.4 Capacity on Demand. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.4.1 Capacity Upgrade on Demand (CUoD). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.4.2 Power enterprise pools and mobile capacity on demand (Mobile CoD) . . . . . . . . 69
2.4.3 Elastic Capacity on Demand (Elastic CoD). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.4.4 Utility Capacity on Demand (Utility CoD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
2.4.5 Trial Capacity on Demand (Trial CoD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
2.4.6 Software licensing and CoD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2.5 System bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2.5.1 PCI Express Gen3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
2.5.2 Service Processor Bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
2.6 Internal I/O subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
2.6.1 Blind-swap cassettes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.6.2 System ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.7 PCI adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
2.7.1 PCI Express (PCIe). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
2.7.2 LAN adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
2.7.3 Graphics accelerator adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
2.7.4 SAS adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
2.7.5 Fibre Channel adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
2.7.6 Fibre Channel over Ethernet. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
2.7.7 USB adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
2.7.8 InfiniBand host channel adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
2.7.9 Cryptographic Coprocessor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
2.8 Internal storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2.8.1 DVD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2.9 External I/O subsystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
2.9.1 PCIe Gen3 I/O expansion drawer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
2.9.2 PCIe Gen3 I/O expansion drawer optical cabling . . . . . . . . . . . . . . . . . . . . . . . . . 87
2.9.3 PCIe Gen3 I/O expansion drawer SPCN cabling . . . . . . . . . . . . . . . . . . . . . . . . . 92
2.10 External disk subsystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
2.10.1 EXP24S SFF Gen2-bay Drawer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
2.10.2 EXP24S common usage scenarios. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
2.10.3 IBM System Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

Contents v
2.11 Hardware Management Console . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
2.11.1 HMC code level. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
2.11.2 HMC RAID 1 support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
2.11.3 HMC connectivity to the POWER8 processor-based systems . . . . . . . . . . . . . 107
2.11.4 High availability HMC configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
2.12 Operating system support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
2.12.1 Virtual I/O Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
2.12.2 IBM AIX operating system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
2.12.3 IBM i operating system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
2.12.4 Linux operating systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
2.12.5 Java versions that are supported . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
2.12.6 Boosting performance and productivity with IBM compilers . . . . . . . . . . . . . . . 112
2.13 Energy management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
2.13.1 IBM EnergyScale technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
2.13.2 On Chip Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
2.13.3 Energy consumption estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Chapter 3. Virtualization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
3.1 IBM POWER Hypervisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
3.1.1 Virtual SCSI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
3.1.2 Virtual Ethernet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
3.1.3 Virtual Fibre Channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
3.1.4 Virtual (TTY) console . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
3.2 POWER processor modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
3.3 Active Memory Expansion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
3.4 Single root I/O virtualization (SR-IOV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.4.1 Direct access I/O and performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
3.4.2 Adapter sharing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
3.4.3 Adapter resource provisioning (QoS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
3.4.4 Flexible deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
3.4.5 Reduced costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.5 PowerVM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.5.1 PowerVM edition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.5.2 Logical partitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.5.3 Multiple shared processor pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
3.5.4 Virtual I/O Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
3.5.5 Live Partition Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
3.5.6 Active Memory Sharing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
3.5.7 Active Memory Deduplication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
3.5.8 Operating system support for PowerVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
3.5.9 Linux support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
3.5.10 PowerVM simplification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
3.5.11 Single view of all Virtual I/O servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
3.5.12 Virtual Storage Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
3.5.13 Virtual Network Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
3.6 System Planning Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
3.7 IBM Power Virtualization Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
3.8 IBM Power Virtualization Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
3.9 VIOS 2.2.3.51 features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
3.10 Dynamic Partition Remote Restart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Chapter 4. Reliability, availability, and serviceability. . . . . . . . . . . . . . . . . . . . . . . . . . 159
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

vi IBM Power Systems E870 and E880 Technical Overview and Introduction
4.1.1 RAS enhancements of POWER8 processor-based servers. . . . . . . . . . . . . . . . 160
4.2 Reliability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
4.2.1 Designed for reliability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
4.2.2 Placement of components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
4.3 Processor/Memory availability details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
4.3.1 Correctable error introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
4.3.2 Uncorrectable error introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
4.3.3 Processor Core/Cache correctable error handling . . . . . . . . . . . . . . . . . . . . . . . 165
4.3.4 Processor Instruction Retry and other try again techniques . . . . . . . . . . . . . . . . 165
4.3.5 Alternative processor recovery and Partition Availability Priority . . . . . . . . . . . . 166
4.3.6 Core Contained Checkstops and other PowerVM error recovery. . . . . . . . . . . . 166
4.3.7 Cache uncorrectable error handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
4.3.8 Other processor chip functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
4.3.9 Other fault error handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
4.3.10 Memory protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
4.3.11 I/O subsystem availability and Enhanced Error Handling . . . . . . . . . . . . . . . . . 169
4.4 Enterprise systems availability details. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
4.5 Availability impacts of a solution architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
4.5.1 Clustering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
4.5.2 Virtual I/O redundancy configurations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
4.5.3 PowerVM Live Partition Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
4.6 Serviceability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
4.6.1 Detecting errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
4.6.2 Error checkers, fault isolation registers, and First-Failure Data Capture . . . . . . 174
4.6.3 Service processor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
4.6.4 Diagnosing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
4.6.5 Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
4.6.6 Notifying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
4.6.7 Locating and servicing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
4.7 Manageability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
4.7.1 Service user interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
4.7.2 IBM Power Systems Firmware maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
4.7.3 Concurrent firmware update improvements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
4.7.4 Electronic Services and Electronic Service Agent . . . . . . . . . . . . . . . . . . . . . . . 191
4.8 Selected POWER8 RAS capabilities by operating system . . . . . . . . . . . . . . . . . . . . . 195
Related publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Online resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Help from IBM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200

© Copyright IBM Corp. 2014. 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 grant 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 measurements 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 Systems E870 and E880 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®
DB2®
DS8000®
Easy Tier®
Electronic Service Agent™
EnergyScale™
eServer™
FlashSystem™
Focal Point™
IBM®
IBM FlashSystem™
Micro-Partitioning®
POWER®
Power Architecture®
POWER Hypervisor™
Power Systems™
Power Systems Software™
POWER6®
POWER6+™
POWER7®
POWER7+™
POWER8™
PowerHA®
PowerPC®
PowerVM®
PowerVP™
Rational®
Real-time Compression™
Redbooks®
Redpaper™
Redbooks (logo) ®
RS/6000®
Storwize®
System p®
System Storage®
System z®
SystemMirror®
Tivoli®
XIV®
The following terms are trademarks of other companies:
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.
Linux is a trademark of Linus Torvalds in the United States, other countries, or both.
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.
UNIX is a registered trademark of The Open Group in the United States and other countries.
Other company, product, or service names may be trademarks or service marks of others.

© Copyright IBM Corp. 2014. All rights reserved. ix
Preface
This IBM® Redpaper™ publication is a comprehensive guide covering the IBM Power System
E870 (9119-MME) and IBM Power System E880 (9119-MHE) servers that support IBM AIX®,
IBM i, and Linux operating systems. The objective of this paper is to introduce the major
innovative Power E870 and Power E880 offerings and their relevant functions:
The new IBM POWER8™ processor, available at frequencies of 4.024 GHz, 4.190 GHz,
and 4.350 GHz
Up to 16 TB of memory in the E870 and 32 TB in the E880
Significantly strengthened cores and larger caches
Two integrated memory controllers with improved latency and bandwidth
Integrated I/O subsystem and hot-pluggable PCIe Gen3 I/O slots
I/O drawer expansion options offers greater flexibility
Improved reliability, serviceability, and availability (RAS) functions
IBM EnergyScale™ technology that provides features such as power trending,
power-saving, capping of power, and thermal measurement
This publication is for professionals who want to acquire a better understanding of IBM Power
Systems™ products. The intended audience includes the following roles:
Clients
Sales and marketing professionals
Technical support professionals
IBM Business Partners
Independent software vendors
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 E870 and
Power E880 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.
Authors
This paper was produced by a team of specialists from around the world working at the
International Technical Support Organization, Austin Center.
Alexandre Bicas Caldeira is a Certified IT Specialist and is a member of the Power Systems
Advanced Technical Sales Support team for IBM Brazil. He holds a degree in Computer
Science from the Universidade Estadual Paulista (UNESP) and an MBA in Marketing. His
major areas of focus are competition, sales, and technical sales support. Alexandre has more
than 14 years of experience working on IBM Systems & Technology Group Solutions and has
worked also as an IBM Business Partner on Power Systems hardware, AIX, and IBM
PowerVM® virtualization products.

x IBM Power Systems E870 and E880 Technical Overview and Introduction
YoungHoon Cho is a Power Systems Top Gun with the post-sales Technical Support Team
for IBM in Korea. He has over 10 years of experience working on IBM RS/6000®, IBM
System p®, and Power Systems products. He provides second-line technical support to Field
Engineers working on Power Systems and system management.
James Cruickshank works in the Power System Client Technical Specialist team for IBM in
the UK. He holds an Honors degree in Mathematics from the University of Leeds. James has
over 13 years experience working with RS/6000, pSeries, System p and Power Systems
products. James supports customers in the financial services sector in the UK.
Bartłomiej Grabowski is a Principal System Support Specialist in DHL IT Services in the
Czech Republic. He holds a Bachelor’s degree in Computer Science from the Academy of
Computer Science and Management in Bielsko-Biala. His areas of expertise include IBM i,
IBM PowerHA® solutions that are based on hardware and software replication, Power
Systems hardware, and PowerVM. He is an IBM Certified Systems Expert and a coauthor of
several IBM Redbooks® publications.
Volker Haug is an Open Group Certified IT Specialist within IBM Systems & Technology
Group in Germany supporting Power Systems clients and Business Partners. He holds a
Diploma degree in Business Management from the University of Applied Studies in Stuttgart.
His career includes more than 28 years of experience with Power Systems, AIX, and
PowerVM virtualization. He has written several IBM Redbooks publications about Power
Systems and PowerVM. Volker is an IBM POWER8 Champion and a member of the German
Technical Expert Council, which is an affiliate of the IBM Academy of Technology.
Andrew Laidlaw is a Client Technical Specialist for IBM working in the UK. He supports
Service Provider clients within the UK and Ireland, focusing primarily on Power Systems
running AIX and Linux workloads. His expertise extends to open source software package
including the KVM hypervisor and various management tools. Andrew holds an Honors
degree in Mathematics from the University of Leeds, which includes credits from the
University of California in Berkeley.
Seulgi Yoppy Sung is a very passionate Engineer, supporting multi-platform systems as a
System Services Representative almost three year, include Power System hardware, AIX,
high-end and low-end storage DS8000 and V7000. She is very positive and enthusiastic
about Power Systems.
The project that produced this publication was managed by:
Scott Vetter
Executive Project Manager, PMP
Thanks to the following people for their contributions to this project:
Tamikia Barrow, Ella Buslovich, Mark Clark, Gareth Coates, Tim Damron, Dan DeLapp,
David Dilling, Nigel Griffiths, Stacy L Haugen, Daniel Henderson, Eddie Hoyal,
Tenley Jackson, Stephanie Jensen, Roxette Johnson, Caroyln K Jones, Bob Kovacs,
Stephen Lutz, Jai Lei Ma, Cesar D Maciel, Chris Mann, Dwayne Moore, Michael J Mueller,
Mark Olson, Michael Poli, Michael Quaranta, Marc Rauzier, Monica Sanchez, Bill Starke,
Dawn C Stokes, Jeff Stuecheli, Doug Szerdi, Kristin Thomas, Jacobo Vargas, Diane Wallace,
Steve Will, Gerlinde Wochele, Doug Yakesch
IBM
Louis Bellanger
Bull

Preface xi
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
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

xii IBM Power Systems E870 and E880 Technical Overview and Introduction

© Copyright IBM Corp. 2014. All rights reserved. 1
Chapter 1. General description
The IBM Power System E870 (9119-MME) and IBM Power System E880 (9119-MHE)
servers use the latest POWER8 processor technology that is designed to deliver
unprecedented performance, scalability, reliability, and manageability for demanding
commercial workloads.
The Power E870 is a highly-scalable rack-mount system optimized for running AIX, IBM i, and
Linux workloads. The Power E870 is a modular-built system and uses one or two system
nodes together with a system control unit. Each system node is five EIA-units tall (5U) and the
system control unit is two EIA-units (2U) tall. The Power E870 is housed in a 19-inch rack.
The Power E880 is a highly-scalable rack-mount system optimized for running AIX, IBM i, and
Linux workloads. The Power E880 is a modular-built system. The system is built of one, two,
three, or four system nodes together with a system control unit. Each system node is five
EIA-units tall (5U) and the system control unit is two EIA-units (2U) tall. The Power E880 is
housed in a 19-inch rack.
1

2 IBM Power Systems E870 and E880 Technical Overview and Introduction
1.1 Systems overview
The following sections provide detailed information about the Power E870 and Power E880
systems.
1.1.1 Power E870 server
The Power E870 (9119-MME) server is a powerful POWER8-based system that scales up to
eight sockets. Each socket contains a single 8-core or 10-core POWER8 processor. Thus, a
fully-configured Power E870 can scale up to 64 or 80 cores.
The Power E870 is a modular system that is built from a combination of the following four
building blocks:
System control unit
System node
PCIe Gen3 I/O expansion drawer
EXP24S SFF Gen2-bay drawer
The system control unit provides redundant system master clock and redundant system
master Flexible Service Processor (FSP) and support for the Operator Panel, the system
VPD, and the base DVD. The system control unit provides clock signals to the system nodes
with semi-flexible connectors. A system control unit is required for every E870 system.
Each system node provides four processor sockets for POWER8 processors and 32 CDIMM
slots supporting a maximum of 8 memory features. Using the 1024 GB memory features, the
system node can support a maximum of 8 TB of RAM. Thus a fully-configured Power E870
can support up to 16 TB of RAM. Each system node provides eight PCIe Gen3 x16 slots. One
or two system nodes can be included in an E870 system. All of the system nodes in the
server must be the same gigahertz and feature number.
Each optional 19-inch PCIe Gen3 4U I/O Expansion Drawer provides 12 PCIe Gen 3 slots.
The I/O expansion drawer connects to the system unit with a pair of PCIe x16 to optical CXP
converter adapters housed in the system unit. Each system node can support up to four I/O
expansion drawers. Thus, a fully configured Power E870 can support a maximum of eight I/O
expansion drawers.
Each EXP24S SFF Gen2-bay Drawer provides 24 x 2.5-inch form-factor (SFF) SAS bays.
The EXP24S connects to the Power E870 server using a SAS adapter in a PCIe slot in a
system node or in a I/O expansion drawer.

Chapter 1. General description 3
Figure 1-1 shows a single system node Power E870 in a T42 rack with two PCIe I/O drawers
and an EXP24S disk drawer.
Figure 1-1 A Power E870 in a T42 rack
1.1.2 Power E880 server
The Power E880 (9119-MHE) server is a powerful POWER8-based system that scales up to
16 sockets. Each socket contains a single 8-core or 12-core POWER8 processor. Thus, a
fully-configured Power E880 can scale up to 128 or 192 cores.
The Power E880 is a modular system that is built from a combination of the following four
building blocks:
System control unit
System node
PCIe Gen3 I/O expansion drawer
EXP24S SFF Gen2-bay drawer
The system control unit provides redundant system master clock and redundant system
master Flexible Service Processor (FSP) and support for the Operator Panel, the system
VPD, and the base DVD. The system control unit provides clock signals to the system nodes
with semi-flexible connectors. A system control unit is required for every E880 system.
Each system node provides four processor sockets for POWER8 processors and 32 CDIMM
slots supporting a maximum of 8 memory features. Using the 1024 GB memory features,
each system node can support a maximum of 8 TB of RAM. Thus a fully configured four-node
Power E880 can support up to 32 TB of memory. Each system node provides eight PCIe
Gen3 x16 slots. One, two, three, or four system nodes per server are supported. All of the
system nodes in the server must be the same gigahertz and feature number.

Each optional 19-inch PCIe Gen3 4U I/O Expansion Drawer provides twelve PCIe Gen 3
slots. The I/O expansion drawer connects to the system unit with a pair of PCIe x16 to Optical
CXP converter adapters housed in the system unit. Each system node can support up to four
I/O expansion drawers, for a total of 48 PCIe Gen 3 slots. A fully configured Power E880 can
support a maximum of 16 I/O expansion drawers.
Each EXP24S SFF Gen2-bay Drawer provides 24 x 2.5-inch form-factor (SFF) SAS bays.
The EXP24S is connected to the Power E880 server using a SAS adapter in a PCIe slot in a
system node or in a I/O expansion drawer.
Figure 1-2 shows a four system node Power E880 with two PCIe I/O drawers and an EXP24S
disk drawer.
Figure 1-2 A Power E880 in a T42 rack
1.1.3 System control unit
The system control unit (SCU) in a Power E870 and E880 is a new innovation to increase the
reliability, availability, and serviceability of the servers. The 2U unit provides redundant clock
and service processor capability to Power E870 and E880 systems even if they have only one
system unit installed. The SCU also provides a DVD option connected to a PCIe adapter with
a USB cable.
The SCU is powered from the system nodes using cables plugged into the first and second
system nodes in a two-, three-, or four-system node server. The SCU is powered from the
single system node in servers with only one.
The SCU provides redundant clock signalling to the system nodes using semi-flexible
connectors. These connectors are located at the rear of the system and route up and down

the middle of the system nodes. In this way they do not cause any restrictions on the allowed
rack specification.
The SCU provides redundant service processor function to the system nodes using flexible
service processor (FSP) cables. Each system node has two FSP connections to the system
control unit.
The SCU provides redundant connections to one or two HMCs using 1 Gbps RJ45 Ethernet
connections.
Figure 1-3 shows the front and rear view of a system control unit. The locations of the
connectors and features are indicated.
Figure 1-3 Front and rear view of the system control unit
1.1.4 System nodes
The system nodes in the Power E870 and E880 servers host the processors, memory, PCIe
slots, and power supplies for the system. Each system node is 5U high and the first and
second system nodes in a server connect to the system control unit with FSP, clock, and
power connectors. The system node connects to other system nodes using SMP connectors.
Each system node in a Power E870 or E880 server provides four POWER8 processors, 32
CDIMM slots, and eight PCIe Gen3 x16 slots.
Power connections
Backup
battery
Clock
connections
DVD USB
connection
FSP
connections
HMC
connections FSP
connections
HMC
connections
DVD Drive Operator panel
VPD card

Figure 1-4 shows the front view of the system node.
Figure 1-4 Front view of the system node
Figure 1-5 shows the rear view of the system node with notable features highlighted.
Figure 1-5 Rear view of the system node
PCIe Gen3 x16 slots
Clock connectors
Local clock and control units
Power sockets
FSP connector
SCU power
FSP connector
SCU power SMP connectors x6
Power sockets Power sockets

Unlike the inter-node flex connectors on the Power 770 and Power 780 servers, the SMP
cables on the Power E870 and E880 servers are fully flexible. These cables do not impose
restrictions on allowed rack specification. Figure 1-6 shows a diagram of how SMP cables can
be routed within a rack on a Power E880 with all four system nodes installed.
Figure 1-6 An example of SMP cable routing in a four-drawer system
Figure 1-7 shows SMP cable routing in a two-drawer Power E870 or E880.
Figure 1-7 SMP cable routing in a two-drawer Power E870 or E880

1.1.5 I/O drawers
I/O drawers to provide additional PCIe adapter slots and SAS disk slots are available for the
Power E870 and E880. More details can be found in 1.6, “I/O drawers” on page 22.
1.2 Operating environment
Table 1-1 details the operating environment for the Power E870 and E880 servers.
Table 1-1 Operating environment for Power E870 and Power E880
Power E870 and Power E880 operating environment
System Power E870 Power E880 Power E870 Power E880
Operating Non-operating
Temperature 5 to 35 degrees C (41 to 95 F) 5 to 45 degrees C (41 to 114F)
Relative humidity 20 - 80% 8 - 80%
Maximum dew
point
29 degrees C (84 F) Less than 27 degrees C (81 F)
Operating voltage 200 to 240 V AC
180 - 400 V DC
N/A
Operating
frequency
47 - 63 Hz AC N/A
Maximum power
consumption
4150 Watts per system node
510 Watts per system control unit
N/A
Maximum power
source loading
4.20 kVA per system node
0.52 kVA per system control unit
N/A
Maximum
thermal output
14,164 BTU/hour per system node
1,740 BTU/hour per system control
unit
N/A
Maximum altitude 3,048 m (10,000 ft.) N/A

Noise level
Declared A-weighted sound power
level, LWad (B) 1, 2, 3
Declared A-weighted sound pressure
level, LpAm (dB) 1, 2, 3
Operating Idle Operating Idle
Model 9119-MHE
or 9119-MME in a
single drawer
configuration.
7.7 7.7 60 60
Model 9119-MHE
or 9119-MME in a
single drawer
configuration.
The system is in
dynamic power
save (favor
performance)
mode with a
heavy workload.
9.55
9.55
795
795
Model 9119-MHE
or 9119-MME in a
single drawer
configuration.
The system has
acoustical doors
(FC 6248 or
6249) installed.
7.155
7.155
565
565
Model 9119-MHE
or 9119-MME in a
single drawer
configuration.
The system is in
dynamic power
save (favor
performance)
mode with a
heavy workload.
The system has
acoustical doors
(FC 6248 or
6249) installed.
8.855
8.855
735
735
Model 9119-MHE
or 9119-MME in a
2-drawer
configuration.
8.04 8.04 634 634

Noise level
Model 9119-MHE
or 9119-MME in a
2-drawer
configuration.
The system is in
dynamic power
save (favor
performance)
mode with a
heavy workload.
9.84,5
9.84,5
824,5
824,5
Model 9119-MHE
or 9119-MME in a
2-drawer
configuration.
The system has
acoustical doors
(FC 6248 or
6249) installed.
7.454,5
7.454,5
594,5
594,5
Model 9119-MHE
or 9119-MME in a
2-drawer
configuration.
The system is in
dynamic power
save (favor
performance)
mode with a
heavy workload.
The system has
acoustical doors
(FC 6248 or
6249) installed.
9.154,5
9.154,5
764,5
764,5

1.3 Physical package
Table 1-2 on page 12 lists the physical dimensions of the system control unit and of individual
system nodes. Both servers are available only in a rack-mounted form factor.
The Power E870 is a modular system that can be constructed from a single system control
unit and one or two system nodes.
The Power E880 is a modular system that can be constructed from a single system control
unit and one, two, three, or four system nodes.
The system control unit requires 2U and each system node requires 5U. Thus, a
single-enclosure system requires 7U, a two-enclosure system requires 12U, a
three-enclosure system requires 17U, and a four-enclosure system requires 22U.
Noise level
Notes:
1. Declared level LWad is the upper-limit A-weighted sound power level. Declared level LpAm is the
mean A-weighted emission sound pressure level that is measured at the 1-meter bystander
positions.
2. All measurements made in conformance with ISO 7779 and declared in conformance with ISO
9296.
3. 10 dB (decibel) equals 1 B (bel).
4. Estimated.
5. Notice: Government regulations (such as those prescribed by OSHA or European Community
Directives) might govern noise level exposure in the workplace and might apply to you and your
server installation. This IBM system is available in racks FC 7014-T00 and 7014-T42 with an
optional acoustical door feature that can help reduce the noise that is emitted from this system.
The actual sound pressure levels in your installation depend upon various factors, including the
number of racks in the installation; the size, materials, and configuration of the room where you
designate the racks to be installed; the noise levels from other equipment; the room ambient
temperature, and employees' location in relation to the equipment. Further, compliance with such
government regulations also depends upon various extra factors, including the duration of
employees' exposure and whether employees wear hearing protection. IBM recommends that
you consult with qualified experts in this field to determine whether you are in compliance with the
applicable regulations.
6. A 1.0 B increase in the LWad for a product equals a sound that is approximately twice as loud or
twice as noisy.
Environmental assessment: The IBM Systems Energy Estimator tool can provide more
accurate information about power consumption and thermal output of systems based on a
specific configuration, including adapter cards and I/O expansion drawers. The Energy
Estimator tool can be accessed online:
http://www-912.ibm.com/see/EnergyEstimator

Table 1-2 Physical dimensions of the Power E870 and Power E880
Figure 1-8 shows a picture of a Power E870 system control unit and system node from the
front.
Figure 1-8 Power 870 system control unit and system node
1.4 System features
Both Power E870 and Power E880 system nodes contain four processor modules with
512 KB L2 cache and 8 MB L3 cache per core.
1.4.1 Power E870 system features
The following features are available on the Power E870:
One or two 5U 19-inch rack mount system node drawers
One 2U 19-inch rack-mount system control unit drawer
Dimension System control unit Power E870 or
Power E880 system
node
PCIe Gen3 I/O
expansion drawer
Width 434 mm (17.1 in) 445 mm (17.5 in) 482 mm (19 in)
Depth 813 mm (32.0in) 902 mm (35.5 in) 802 mm (31.6in)
Height 86 mm (3.4 in) 2 EIA
units
219 mm (8.6 in) 5 EIA
units
173 mm 6.8 in) 4 EIA
units
Weight 23.6 kg (52.0 lb) 75.7 kg (167 lb) 54.4 kg (120 lb)

5U for a system with one system node drawer
12U for a system with two system node drawer
One processor feature per system node. All system nodes must have the same feature:
– 4.024 GHz, (4 X 0/8-core) 32-core POWER8 processor (#EPBA)
– 4.190 GHz, (4 X 0/10-core) 40-core POWER8 processor (#EPBC)
Static or mobile processor activation features available on a per core basis
POWER8 DDR3 or DDR4 memory CDIMMs (32 CDIMM slots per system node, 16 sites
populated per system node minimum - see 2.3.2, “Memory placement rules” on
page 55for #EM8Y restrictions):
– 0/64 GB (4 X 16 GB), 1600 MHz (#EM8J) DDR3
– 0/128 GB (4 X 32 GB), 1600 MHz (#EM8K) DDR3
– 0/256 GB (4 X 64 GB), 1600 MHz (#EM8L) DDR3
– 0/512 GB (4 x 128 GB), 1600MHz (#EM8M) DDR3
– 0/1024 GB (4 x 256 GB), 1600 MHz (#EM8Y) DDR4
IBM Active Memory™ Expansion - optimized onto the processor chip (#EM82)
Elastic CoD Processor Day - no-charge, temporary usage of inactive, Elastic CoD
resources on initial orders (#EPJ3 or #EPJ5)
90 Days Elastic CoD Temporary Processor Enablement (#EP9T)
Eight PCIe Gen3 x16 I/O expansion slots per system node drawer (maximum 16 with
2-drawer system)
One slim-line, SATA media bay per system control unit enclosure (DVD drive selected as
default with the option to de-select)
Redundant hot-swap AC or DC power supplies in each system node drawer
Two HMC 1 Gbps ports per Flexible Service Processor (FSP) in the system control unit
Dynamic logical partition (LPAR) support
Processor and memory Capacity Upgrade on Demand (CUoD)
Single root I/O virtualization (SR-IOV)
PowerVM virtualization (standard and optional):
– IBM Micro-Partitioning®
– Dynamic logical partitioning
– Shared processor pools
– Shared storage pools
– Live Partition Mobility
– Active Memory Sharing
– Active Memory Deduplication
– NPIV support
– IBM PowerVP™ Performance Monitor
Optional PowerHA for AIX and IBM i
Optional PCIe Gen3 I/O Expansion Drawer with PCIe Gen3 slots:
– Zero, one, two, three or four PCIe Gen3 Drawers per system node drawer (#EMX0)
– Each Gen3 I/O drawer holds two 6-slot PCIe3 Fan-out Modules (#EMXF)

– Each PCIe3 fan-out module attaches to the system node via two CXP Active Optical
Cables (#ECC6, #ECC8, or #ECC9) to one PCIe3 Optical CXP Adapter (#EJ07)
1.4.2 Power E880 system features
The following features are available on the Power E880:
One, two, three, or four 5U 19-inch rack-mount system node drawers
One 2U 19-inch rack-mount system control unit drawer
7U for a system with one system node drawer plus one system control unit
22U for a system with four system node drawers
One processor feature per system node:
– 4.35 GHz, (4 X 0/8-core) 32-core POWER8 processor (#EPBB)
– 4.19 GHz (4 X 0/10-core) 40-core POWER8 processor (#EPBS)
– 4.02 GHz, (4 X 0/12-core) 48-core POWER8 processor (#EPBD)
Static or mobile processor activation features available on a per core basis
POWER8 DDR3 or DDR4 memory CDIMMs (32 CDIMM slots per system node, 16 slots
populated per system node minimum - see 2.3.2, “Memory placement rules” on
page 55for #EM8Y restrictions):
– 0/64 GB (4 X 16 GB), 1600 MHz (#EM8J) DDR3
– 0/128 GB (4 X 32 GB), 1600 MHz (#EM8K) DDR3
– 0/256 GB (4 X 64 GB), 1600 MHz (#EM8L) DDR3
– 0/512 GB (4 X 128 GB), 1600 MHz (#EM8M) DDR3
– 0/1024 GB (4 x 256 GB), 1600 MHz (#EM8Y) DDR4
Active Memory Expansion - optimized onto the processor chip (#EM82)
90 Days Elastic CoD Temporary Processor Enablement (#EP9T)
Eight PCIe Gen3 x16 I/O expansion slots per system node drawer (maximum 16 with
2-drawer system)
One slim-line, SATA media bay per system control unit enclosure (DVD drive defaulted on
order, option to de-select)
Redundant hot-swap AC or DC power supplies in each system node drawer
Two HMC 1 Gbps ports per Flexible Service Processor (FSP) in the system control unit
Dynamic logical partition (LPAR) support
Processor and memory CUoD
Single root I/O virtualization (SR-IOV)
PowerVM virtualization:
– Micro-Partitioning
– Dynamic logical partitioning
– Shared processor pools
– Shared storage pools
– Live Partition Mobility
– Active Memory Sharing

– Active Memory Deduplication
– NPIV support
– PowerVP Performance Monitor
Optional PowerHA for AIX and IBM i
Optional PCIe Gen3 I/O Expansion Drawer with PCIe Gen3 slots:
– Zero, one, two, three, or four PCIe Gen3 Drawers per system node drawer (#EMX0)
– Each Gen3 I/O drawer holds two 6-slot PCIe3 Fan-out Modules (#EMXF)
– Each PCIe3 fan-out module attaches to the system node via two CXP Active Optical
Cables (#ECC6, #ECC8, or #ECC9) to one PCIe3 Optical CXP Adapter (#EJ07)
1.4.3 Minimum features
Each Power E870 or Power E880 initial order must include a minimum of the following items:
1 x system node with a choice of the following processors:
– 4.024 GHz, 32-core POWER8 processor module (#EPBA) for Power E870
– 4.190 GHz, 40-core POWER8 processor module (#EPBC) for Power E870
– 4.35 GHz, 32-core POWER8 processor module (#EPBB) for Power E880
– 4.02 GHz, 48-core POWER8 processor module (#EPBD) for Power E880
8 x 1 core processor activation:
– 1 core permanent Processor Activation for #EPBA (#EPBJ) for Power E870
– 1 core permanent Processor Activation for #EPBC (#EPBL) for Power E870
– 1 core permanent Processor Activation for #EPBB (#EPBK) for Power E880
– 1 core permanent Processor Activation for #EPBD (#EPBM) for Power E880
4 x 64 GB (4 x 16 GB) CDIMMs, 1600 MHz, 4 Gb DDR3 DRAM (#EM8J)
Total of 50% of installed memory must be activated with a combination of permanent or
mobile activation using features:
– 1 GB Memory Activation (#EMA5)
– 100 GB Memory Activations (#EMA6)
– 100 GB Mobile Activations (#EMA9)
A maximum of 75% of memory activations can be mobile activations
1 x 5U system node drawer (#EBA0)
2 x Service Processor (#EU0A)
1 x Load Source Specify:
– EXP24S SFF Gen2 (#5887 or #EL1S) Load Source Specify (#0728)
– SAN Load Source Specify (#0837)
1 x Rack-mount Drawer Bezel and Hardware:
– IBM Rack-mount Drawer Bezel and Hardware (#EBA2) for Power E870
– IBM Rack-mount Drawer Bezel and Hardware (#EBA3) for Power E880
– OEM Rack-mount Drawer Bezel and Hardware (#EBA4)
1 x System Node to System Control Unit Cable Set for Drawer 1 (#ECCA)

1 x Power Chunnels for routing power cables from back of machine to front (#EBAA)
1 x Language Group Specify (#9300/#97xx)
Optional 1 x Media:
– SATA Slimline DVD-RAM with write cache (#EU13)
– PCIe2 LP 4-Port USB 3.0 Adapter (#EC45)
– PCIe2 4-Port USB 3.0 Adapter (#EC46)
When either AIX or Linux are the primary operating systems, the order must include a
minimum of the following items also:
1 x Primary Operating System Indicator:
– Primary Operating System Indicator - AIX (#2146)
– Primary Operating System Indicator - Linux (#2147)
When IBM i is the primary operating system, the order must include a minimum of the
following items:
1 x Specify Code:
– Mirrored System Disk Level, Specify Code (#0040)
– Device Parity Protection-All, Specify Code (#0041)
– Mirrored System Bus Level, Specify Code (#0043)
– Device Parity RAID 6 All, Specify Code (#0047)
– Mirrored Level System Specify Code (#0308)
1 x System Console:
– Sys Console On HMC (#5550)
– System Console-Ethernet No IOP (#5557)
1 x Primary Operating System Indicator - IBM i (#2145)

1.4.4 Power supply features
This section describes how the system nodes and system control units are powered.
System node power
Four AC or DC power supplies are required for each system node enclosure. This
arrangement provides 2+2 redundant power with dual power sources for enhanced system
availability. A failed power supply can be hot swapped but must remain in the system until the
replacement power supply is available for exchange.
Four AC or DC power cords are used for each system node (one per power supply) and are
ordered using the AC Power Chunnel feature (#EBAA). Each #EBAA provides all four AC
power cords, therefore a single #EBAA should be ordered per system node. The chunnel
carries power from the rear of the system node to the hot swap power supplies located in the
front of the system node where they are more accessible for service.
Notes:
Additional optional features can be added, as desired.
IBM i systems require a DVD to be available to the system when required. This DVD
can be located in the system control unit (DVD feature #EU13) or it can be located
externally in an enclosure like the 7226-1U3. A USB PCIe adapter such as #EC45 or
#EC46 is required for #EU13. A SAS PCIe adapter such as #EJ11 is required to attach
a SATA DVD in the 7226-1U3. A virtual media repository may be used to substitute for a
DVD device.
Feature-coded racks are allowed for I/O expansion only.
A machine type/model rack, if desired, should be ordered as the primary rack.
A minimum number of eight processor activations must be ordered per system.
A minimum of four memory features per drawer are required.
At least 50% of available memory must be activated through a combination of feature
#EMA5 and #EMA6, and #EMA9. 25% of the available memory must be permanently
activated using #EMA5 and #EMA6. The remaining 75% of the active memory can be
enabled using either permanent or mobile options.
Memory sizes can differ across CPU modules, but the eight CDIMM slots connected to
the same processor module must be filled with identical CDIMMs.
If SAN Load Source Specify (#0837) is ordered #0040, #0041, #0043, #0047, #0308
are not supported.
The language group is auto-selected based on geographic rules.
No feature codes are assigned for the following items:
– Four AC power supplies are delivered as part of the system node. No features are
assigned to power supplies. Four line cords are auto-selected according to
geographic rules.
– There must be one system control unit on each system. The system control unit is
considered the system with the serial number.
An HMC is required for management of every Power E870 or Power E880; however, a
shared HMC is acceptable. HMCs supported on POWER8 hardware are 7042-C08 and
7042-CR5 through 7042-CR9. For more details see 1.12, “Management consoles” on
page 28

System control unit power
The system control unit is powered from the system nodes. UPIC cables provide redundant
power to the system control unit. Two UPIC cables attach to system node drawer 1 and two
UPIC cables attach to system node drawer 2. They are ordered with #ECCA and #ECCB. The
UPIC cords provides N+1 redundant power to the system control unit.
1.4.5 Processor card features
Each system node will deliver a set of four identical processors. All processors in the system
must be identical. Cable features are required to connect system nodes to the system control
unit and to other system nodes.
For a single system node configuration, #ECCA is required. #ECCA provides cables to
connect the system node with the system control unit.
For a dual system node configuration, #ECCB is required. #ECCB provides cables to
connect the system nodes with the system control unit and cables to connect the two
system nodes.
For three or four system node systems there are no additional cables required, as
redundant power is provided through system nodes 1 and 2.
Each system must have a minimum of eight active cores:
The Power E870 has two types of processors, offering the following features:
4.024 GHz, (4 X 0/8-core) 32-core POWER8 processor (#EPBA)
4.190 GHz, (4 X 0/10-core) 40-core POWER8 processor (#EPBC)
The Power E880 has three types of processors, offering the following features:
4.35 GHz, (4 X 0/8-core) 32-core POWER8 processor (#EPBB)
4.19 GHz (4 X 0/10-core) 40-core POWER8 processor (#EPBS)
4.02 GHz, (4 X 0/12-core) 48-core POWER8 processor (#EPBD)
Several types of capacity on demand (CoD) processor options are available on the
Power E870 and Power 880 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. The #EPJ3 provides no-charge elastic processor days for
Power E880. The #EMJ8 feature provides no-charge elastic memory days for Power E880.
1.4.6 Summary of processor features
Table 1-3 summarizes the processor feature codes for the Power E870.
Table 1-3 Summary of processor features for the Power E870
Feature Code Description
EPBA 4.02 GHz, 32-core POWER8 processor
EPBC 4.19 GHz, 40-core POWER8 processor
EB2R Single 5250 Enterprise Enablement, IBM i
EB30 Full 5250 Enterprise Enablement, IBM i
EP2S 1-Core Mobile Activation

Table 1-4 summarizes the processor feature codes for the Power E880.
Table 1-4 Summary of processor features for the Power E880
EP2U 1-Core Mobile Activation from POWER® 7
EP9T 90 Days Elastic CoD Processor Core Enablement
EPBJ 1 core Processor Activation for #EPBA
EPBL 1 core Processor Activation for #EPBC
EPBN 1 core Processor Activation for #EPBA, Mobile Enabled
EPBQ 1 core Processor Activation for #EPBC, Mobile Enabled
EPJ6 1 Proc-Day Elastic CoD Billing for #EPBA, AIX/Linux
EPJ7 1 Proc-Day Elastic CoD Billing for #EPBA, IBM i
EPJ8 100 Elastic CoD Proc-Days of Billing for Processor #EPBA. AIX/Linux
EPJ9 100 Elastic CoD Proc-Days of Billing for Processor #EPBA. IBM i
EPJA Proc CoD Utility Billing, 100 Proc-mins. for #EPBA, AIX/Linux
EPJB Proc CoD Utility Billing, 100 Proc-mins. for #EPBA, IBM i
EPJJ 1 Proc-Day Elastic CoD Billing for #EPBC, AIX/Linux
EPJK 1 Proc-Day Elastic CoD Billing for #EPBC, IBM i
EPJL 100 Elastic CoD Proc-Days of Billing for Processor #EPBC. AIX/Linux
EPJM 100 Elastic CoD Proc-Days of Billing for Processor #EPBC. IBM i
EPJN Proc CoD Utility Billing, 100 Proc-mins. for #EPBC, AIX/Linux
EPJP Proc CoD Utility Billing, 100 Proc-mins. for #EPBC, IBM i
ELJG Power Integrated Facility for Linux, adds #ELJ8 for 4 activationsa
a. This feature activates 4 processors (#ELJ8), as well as 32GB memory (#ELJH) and 4 licenses
for PowerVM for Linux (ELJJ)
EPBB 4.35 GHz, 32-core POWER8 processor
EPBS 4.19 GHz, 40-core POWER8 processor
EPBD 4.02 GHz, 48-core POWER8 processor
EB2R Single 5250 Enterprise Enablement
EB30 Full 5250 Enterprise Enablement
EP2T 1-Core Mobile Activation
EP2V 1-Core Mobile Activation from POWER 7
EP9T 90 Days Elastic CoD Processor Core Enablement
EPBK 1 core Processor Activation for #EPBB
EPBU 1 core Processor Activation for #EPBS
EPBM 1 core Processor Activation for #EPBD

1.4.7 Memory features
1600 MHz memory CDIMMs are available as 64 GB (#EM8J), 128 GB (#EM8K), 256 GB
(#EM8L), 512 GB (#EM8M), and 1024 (#EM8Y) memory features. Each memory feature
provides four CDIMMs. CDIMMs are custom DIMMs which enhance memory performance
and memory reliability. Each system node has 32 CDIMM slots which support a maximum of
eight memory features. Memory activations of 50% of the installed capacity are required. The
feature #EMJ8 provides no-charge elastic memory days for the Power E880.
Table 1-5 lists memory features that are available for the Power E870 and E880.
Table 1-5 Summary of memory features
EPBP 1 core Processor Activation for #EPBB, Mobile Enabled
EPBV 1 core Processor Activation for #EPBS, Mobile Enabled
EPBR 1 core Processor Activation for #EPBD, Mobile Enabled
EPJ3 48 Proc-Days of Elastic CoD Processor Resources for #EPBB
EPJ5 48 Proc-Days of Elastic CoD Processor Resources for #EPBD
EPJC 1 Proc-Day Elastic CoD Billing for #EPBB, AIX/Linux
EPJQ 1 Proc-Day Elastic CoD Billing for #EPBD, AIX/Linux
EPJD 1 Proc-Day Elastic CoD Billing for #EPBB, IBM i
EPQR 1 Proc-Day Elastic CoD Billing for #EPBD, IBM i
EPJE 100 Elastic CoD Proc-Days of Billing for Processor #EPBB. AIX/Linux
EPJS 100 Elastic CoD Proc-Days of Billing for Processor #EPBD. AIX/Linux
EPJF 100 Elastic CoD Proc-Days of Billing for Processor #EPBB. IBM i
EPJT 100 Elastic CoD Proc-Days of Billing for Processor #EPBD. IBM i
EPJG Proc CoD Utility Billing, 100 Proc-mins. for #EPBB, AIX/Linux
EPJU Proc CoD Utility Billing, 100 Proc-mins. for #EPBD, AIX/Linux
EPJH Proc CoD Utility Billing, 100 Proc-mins. for #EPBB, IBM i
EPJV Proc CoD Utility Billing, 100 Proc-mins. for #EPBD, IBM i
ELJG Power Integrated Facility for Linux, adds #ELJ8 for 4 activationsa
EM8J 64 GB (4X16 GB) CDIMMs, 1600 MHz, 4 GBIT DDR3 DRAM
EM8K 128 GB (4X32 GB) CDIMMs, 1600 MHz, 4 GBIT DDR3 DRAM
EM8L 256 GB (4X64 GB) CDIMMs, 1600 MHz, 4 GBIT DDR3 DRAM
EM8M 512 GB (4X128 GB) CDIMMs, 1600 MHz, 4 GBIT DDR3 DRAM
EM8Y 1024 GB (4X 256 GB) CDIMMs, 1600 MHz, 4 GBIT DDR4 DRAM

1.4.8 System node PCIe slots
Each Power E870 or Power E880 system node provides eight half-length, half-high x16 PCIe
Gen3 slots. These PCIe slots can be used for either low profile PCIe adapters or for attaching
a PCIe I/O drawer.
A new form factor blind swap cassette (BSC) is used to house the low profile adapters, which
go into these slots.
PCIe Gen1, Gen2, and Gen3 adapter cards are supported in these Gen3 slots.
Table 1-6 provides details of the PCI slots in the Power E870 and Power E880 system.
Table 1-6 PCIe slot locations and descriptions for the Power E870 and Power E880
EMB6 Bundle of eight #EM8M, 512 GB 1600 MHz Memory features
EM82 Active Memory Expansion Enablement
EM9T 90 Days Elastic CoD Memory Enablement
EMA5 1 GB Memory Activation
EMA6 Quantity of 100 1 GB Memory Activations (#EMA5)
EMA7 100 GB Mobile Memory Activations
EMA9 100 GB Mobile Enabled Memory Activations
EMB7 Bundle of forty, #EMA6 1600 MHz memory features
EMB8 Bundle of 512 #EMA5 1600 MHz memory features
EMJ4 1 GB-Day billing for Elastic CoD memory
EMJ8 384 GB-Days of Elastic CoD Memory Resources (E880 only)
ELJG Power Integrated Facility for Linux, adds #ELJH for 32 GB activationsa
Slot Locationcode Description PHB Adapter size
Slot 1 P1-C1 PCIe3, x16 Processor Module 1, PHB1 Short (low-profile)

All slots support enhanced error handling (EEH).
All PCIe slots are hot swappable and support concurrent maintenance.
1.5 Disk and media features
The Power E870 and Power E880 system control unit and system nodes do not support
internal disks. Any required disk must reside within a SAN disk subsystem or an external disk
drawer. The EXP24S SFF Gen2-bay Drawer (#5887) is the only supported disk drawer for
Power E870 or Power E880.
Each system control unit enclosure has one slim-line bay which can support one DVD drive
(#EU13). The #EU13 DVD is cabled to a USB PCIe adapter located in either a system node
(#EC45) or in a PCIe Gen3 I/O drawer (#EC46). A USB to SATA converter is included in the
configuration without a separate feature code.
IBM i systems require a DVD to be available to the system when required. This DVD can be
located in the system control unit (DVD feature #EU13) or it can be located externally in an
enclosure like the 7226-1U3. A USB PCIe adapter such as #EC45 or #EC46 is required for
#EU13. A SAS PCIe adapter such as #EJ11 is required to attach a SATA DVD in the
7226-1U3. A virtual media repository may be used to substitute for a DVD device if using
VIOS.
1.6 I/O drawers
If additional Gen3 PCIe slots beyond the system node slots are required, a system node x16
slot is used to attach a 6-slot expansion module in the I/O Drawer. A PCIe Gen3 I/O
expansion drawer (#EMX0) holds two expansion modules which are attached to any two x16
PCIe slots in the same system node or in different system nodes.
Disk-only I/O drawers (#5887) are also supported, providing storage capacity.
1.6.1 PCIe Gen3 I/O expansion drawer
The 19-inch 4 EIA (4U) PCIe Gen3 I/O expansion drawer (#EMX0) and two PCIe FanOut
Modules (#EMXF) provide twelve PCIe I/O full-length, full-height slots. One FanOut Module
provides six PCIe slots labeled C1 through C6. C1 and C4 are x16 slots and C2, C3, C5, and
C6 are x8 slots. PCIe Gen1, Gen2, and Gen3 full-high adapter cards are supported.
A blind swap cassette (BSC) is used to house the full-high adapters that go into these slots.
The BSC is the same BSC as used with the previous generation server's 12X attached I/O
drawers (#5802, #5803, #5877, #5873). The drawer is shipped with a full set of BSC, even if
the BSC is empty.
Concurrent repair and add/removal of PCIe adapter cards is done through HMC guided
menus or by operating system support utilities.
A PCIe x16 to Optical CXP converter adapter (#EJ07) and 2.0 M (#ECC6), 10.0 M (#ECC8),
or 20.0 M (#ECC9) CXP 16X Active Optical cables (AOC) connect the system node to a PCIe
FanOut module in the I/O expansion drawer. One feature #ECC6, #ECC8, or #ECC9 ships
two AOC cables. Each PCIe Gen3 I/O expansion drawer has two power supplies.

Each system node supports zero, one, two, three, or four PCIe Gen3 I/O expansion drawers.
A half drawer, consisting of just one PCIe fan-out module in the I/O drawer is also supported,
allowing a lower-cost configuration if fewer PCIe slots are required. Thus a system node
supports the following half drawer options: one half drawer, two half drawers, three half
drawers, or four half drawers. Because there is a maximum of four #EMX0 drawers per node,
a single system node cannot have more than four half drawers. A server with more system
nodes can support more half drawers up to four per system node. A system can also mix half
drawers and full PCIe Gen3 I/O expansion drawers. The maximum of four PCIe Gen3 drawers
per system node applies whether a full or half PCIe drawer.
Drawers can be added to the server at a later time, but system downtime must be scheduled
for adding a PCIe3 Optical Cable Adapter or a PCIe Gen3 I/O drawer (EMX0) or fan-out
module (#EMXF).
Figure 1-9 shows a PCIe Gen3 I/O expansion drawer.
Figure 1-9 PCIe Gen3 I/O expansion drawer
For more details on connecting PCIe Gen3 I/O expansion drawers to the Power E870 and
Power E880 servers, see 2.9.1, “PCIe Gen3 I/O expansion drawer” on page 85

1.6.2 I/O drawers and usable PCI slot
Figure 1-10 shows the rear view of the PCIe Gen3 I/O expansion drawer with the location
codes for the PCIe adapter slots in the PCIe3 6-slot fanout module.
Figure 1-10 Rear view of a PCIe Gen3 I/O expansion drawer with PCIe slots location codes
Table 1-7 provides details of the PCI slots in the PCIe Gen3 I/O expansion drawer.
Table 1-7 PCIe slot locations and descriptions for the PCIe Gen3 I/O expansion drawer
All slots support full-length, regular-height adapter or short (low-profile) with a
regular-height tailstock in single-wide, generation 3, blind-swap cassettes.
Slot Location code Description
Slot 1 P1-C1 PCIe3, x16
P1-C3
P1-C2
P1-C1
P1-C4
P1-C5
P1-C6
P2-C1
P2-C2
P2-C3
P2-C6
P2-C5
P2-C4

Slots C1 and C4 in each PCIe3 6-slot fanout module are x16 PCIe3 buses and slots C2,
C3, C5, and C6 are x8 PCIe buses.
All slots support enhanced error handling (EEH).
All PCIe slots are hot swappable and support concurrent maintenance.
Table 1-8 summarizes the maximum number of I/O drawers supported and the total number
of PCI slots that are available when expansion consists of a single drawer type.
Table 1-8 Maximum number of I/O drawers supported and total number of PCI slots
1.7 EXP24S SFF Gen2-bay Drawer
The EXP24S SFF Gen2-bay Drawer (#5887) is an expansion drawer with 24 2.5-inch
form-factor (SFF) SAS bays. The EXP24S supports up to 24 hot-swap SFF-2 SAS hard disk
drives (HDDs) or solid state drives (SSDs). It uses 2 EIA of space in a 19-inch rack. The
EXP24S includes redundant AC power supplies and uses two power cords.
With AIX, Linux, and VIOS, you can order the EXP24S with four sets of six bays, two sets of
12 bays, or one set of 24 bays (mode 4, 2, or 1). With IBM i, you can order the EXP24S as
one set of 24 bays (mode 1). Mode setting is done by IBM Manufacturing. If you need to
change the mode after installation, ask your IBM support rep. to refer to:
http://w3-03.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/PRS5121
The EXP24S SAS ports are attached to a SAS PCIe adapter or pair of adapters using SAS
YO or X cables.
System nodes Maximum #EMX0
drawers
Total number of slots
PCIe3, x16 PCIe3, x8 Total PCIe3
1 system node 4 16 32 48
2 system nodes 8 32 64 96
3 system nodes 12 48 96 144
4 system nodes 16 64 128 192

Figure 1-11 shows the EXP24S SFF Gen2-bay drawer.
Figure 1-11 EXP24S SFF Gen2-bay drawer
For more information on connecting the EXP24S Gen2-bay drawer to the Power E870 and
Power E880 servers, see 2.10.1, “EXP24S SFF Gen2-bay Drawer” on page 92.
1.8 Comparison between models
The Power E870 offers configuration options where the POWER8 processor can have one of
two processor speeds. Each system node is populated with four single chip modules (SCMs).
Each system node will contain one of the following processor configurations:
Four 4.02 GHz 8-core SCMs
A Power E870 with either of the processor configurations can have as few as eight cores
activated or up to 100% of the cores can be activated. Incrementing one core at a time is
available through built-in capacity on demand (CoD) functions to the full capacity of the
system. The Power E870 can be installed with one or two system nodes connected to the
system control unit. Each system node can have up to 8 TB of memory installed.
The Power E880 also offers system nodes populated with four SCMs. The Power E880
system node has two processor configuration:
A Power E880 with either of the processor configurations can have as few as eight cores
activated or up to 100% of the cores can be activated. Incrementing one core at a time is
available through built-in capacity on demand (CoD) functions to the full capacity of the
system. The Power E880 can be installed with one, two, three, or four system nodes
connected to the system control unit. Each system node can have up to 8 TB of memory
installed.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
FC5887
C1
C2
LNK LNK LNK LNK
T2 T3T1
LNK LNK
10/100 LNK
LNK LNK LNK LNK
T2 T3T1
LNK LNK
10/100 LNK
ESMESM
OK
IN
DC
OK
IN
DC
Front
Rear

Table 1-9 shows a summary of processor and memory maximums for the Power E870 and
Power E880.
Table 1-9 Summary of processor and memory maximums for the Power E870 and Power E880
1.9 Build to order
You can do a build to order (also called a la carte) configuration by using the IBM
Configurator for e-business (e-config). With it, you specify each configuration feature that you
want on the system.
This method is the only configuration method for the Power E870 and Power E880 servers.
1.10 IBM editions
IBM edition offerings are not available for the Power E870 and Power E880 servers.
1.11 Model upgrades
The following sections describe the various upgrades that are available.
1.11.1 Power E870
A model conversion from a Power 770 (9117-MMD) to a Power E870 (9119-MME) is
available. One-step upgrades from previous models of the Power 770 (9117-MMB and
9117-MMC) are not available. To upgrade from a 9117-MMB or 9117-MMC, an upgrade to a
9117-MMD is required first.
The existing components being replaced during a model or feature conversion become the
property of IBM and must be returned.
1.11.2 Power E880
A model conversion from a Power 780 (9179-MHD) to a Power E880 (9119-MHE) is available.
One-step upgrades from previous models of the Power 780 (9179-MHB and 9179-MHC) are
not available. To upgrade from a 9179-MHB or 9179-MHC, an upgrade to a 9179-MHD is
required first.
A model conversion from a Power 770 (9117-MMD) to a Power E880 (9119-MHE) is also
available. One-step upgrades from previous models of the Power 770 (9117-MMB and
9117-MMC) are not available. To upgrade from a 9117-MMB or 9117-MMC, an upgrade to a
9117-MMD is required first.
System Cores
per
SCM
Core
speed
System
node core
maximum
System
core
maximum
System
node
memory
maximum
System
memory
maximum
Power E870 8 4.02 GHz 32 64 8 TB 8 TB
Power E870 10 4.19 GHz 40 80 8 TB 16 TB
Power E880 8 4.35 GHz 32 128 8 TB 24 TB
Power E880 12 4.02 GHz 48 192 8 TB 32 TB

Upgrades to a Power E880 from a Power 795 (9119-FHB) are not available.
The existing components being replaced during a model or feature conversion become the
property of IBM and must be returned.
1.11.3 Upgrade considerations
Feature conversions are set up for the following items:
PCIe Crypto blind swap cassettes
Power IFL processor activations
Power IFL PowerVM for Linux
Active Memory Expansion Enablement
DDR3 memory DIMMS to CDIMMS
Static and mobile memory activations
5250 enterprise enablement
POWER7+™ processor cards to POWER8 processors
Static and mobile processor activations
System CEC enclosure and bezel to 5U system node drawer
PowerVM standard and enterprise
The following features that are present on the current system can be moved to the new
system if they are supported in the Power E870 and E880:
Disks (within an EXP24S I/O drawer)
SSDs (within an EXP24S I/O drawer)
PCIe adapters with cables, line cords, keyboards, and displays
Racks
Doors
EXP24S I/O drawers
For POWER7+ processor-based systems that have the Elastic CoD function enabled, you
must reorder the elastic CoD enablement features when placing the upgrade MES order for
the new Power E870 or E880 system to keep the elastic CoD function active. To initiate the
model upgrade, the on/off enablement features should be removed from the configuration file
before the MES order is started. Any temporary use of processors or memory owed to IBM on
the existing system must be paid before installing the Power E870 or E880.
1.12 Management consoles
This section discusses the supported management interfaces for the servers.
Statement of direction:
For clients with Power E880 servers that require additional system capacity and growth,
IBM plans to enhance the Power Systems' Enterprise system portfolio with greater
flexibility by providing a serial number protected upgrade path from the Power E880
32-core system node server to the Power E880 48-core system node server.

The Hardware Management Console (HMC) is required for managing the IBM
Power E870 and Power E880. It has a set of functions that are necessary to manage
the system:
Creating and maintaining a multiple partition environment
Displaying a virtual operating system terminal session 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
Multiple Power Systems servers can be managed by a single HMC. Each server can be
connected to multiple HMC consoles to build extra resiliency into the management platform.
In 2015, the 7042-CR9 was announced.
The IBM Power E870 and Power E880 are not supported by the Integrated Virtualization
Manager (IVM).
Several HMC models are supported to manage POWER8 based systems. The 7042-CR9 is
are available for ordering at the time of writing, but you can also use one of the withdrawn
models listed in Table 1-10.
Table 1-10 HMC models supporting POWER8 processor technology-based servers
HMC base Licensed Machine Code Version 8 Revision 8.2.0 or later is required to support
the Power E870 (9119-MME) and Power E880 (9119-MHE).
System firmware level 8.3 or later, is a prerequisite for the 48-core system node, Power E880
systems with three or four system nodes, or for system nodes supporting more than two I/O
expansion drawers or half drawers. This will require HMC firmware level 8.3.0 or higher.
Existing HMC models 7042 can be upgraded to Licensed Machine Code Version 8 to support
environments that might include IBM POWER6®, IBM POWER6+™, IBM POWER7®, IBM
POWER7+ and POWER8 processor-based servers.
If you want to support more than 254 partitions in total, the HMC will require a memory
upgrade to a minimum of 4 GB.
Type-model Availability Description
7042-C08 Withdrawn IBM 7042 Model C08 Deskside Hardware Management Console
7042-CR5 Withdrawn IBM 7042 Model CR5 Rack-Mounted Hardware Management Console
7042-CR6 Withdrawn IBM 7042 Model CR6 Rack mounted Hardware Management Console
7042-CR9 Available IBM 7042 Model CR9 Rack mounted Hardware Management Console
Fix Central: You can download or order the latest HMC code from the Fix Central website:
http://www.ibm.com/support/fixcentral

For further information on managing the Power E870 and Power E880 servers from an HMC,
see 2.11, “Hardware Management Console” on page 104.
1.13 System racks
The Power E870 and E880 and its I/O drawers are designed to be mounted in the following
existing IBM racks: 7014-T00, 7014-T42, 7965-94Y, #0553, #0551, and #ER05.
However, for initial system orders, the racks must be ordered as machine type 7014-T42,
#0553. The 36U (1.8-meter) rack (#0551) and the 42U (2.0-meter) rack (#ER05) are available
to order only on Miscellaneous Equipment Specification (MES) upgrade orders.
The Power E870 and Power E880 use a new type of connector between system drawers.
Therefore, the systems do not require wider racks, and an OEM rack or cabinet which meet
the requirements can be used.
1.13.1 IBM 7014 model T00 rack
The 1.8-meter (71-inch) model T00 is compatible with past and present IBM Power systems.
The features of the T00 rack are as follows:
36U (EIA units) of usable space.
Optional removable side panels.
Optional highly perforated front door.
Optional side-to-side mounting hardware for joining multiple racks.
Standard business black or optional white color in OEM format.
Increased power distribution and weight capacity.
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.
The #6068 feature provides a cost effective plain front door.
Weights are as follows:
– T00 base empty rack: 244 kg (535 lb.)
– T00 full rack: 816 kg (1795 lb.)
– Maximum Weight of Drawers is 572 kg (1260 lb.)
– Maximum Weight of Drawers in a zone 4 earthquake environment is 490 kg (1080 lb.).
This equates to 13.6 kg (30 lb.)/EIA.
Shipping without a rack: If you require the system to be shipped without an IBM rack,
feature code #ER21 must be used to remove the IBM rack from the order. The server will
then ship as separate packages for installation into an existing rack.
Installing in non-IBM racks: 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.

1.13.2 IBM 7014 model T42 rack
The 2.0-meter (79.3-inch) Model T42 addresses the requirement for a tall enclosure to house
the maximum amount of equipment in the smallest possible floor space. The following
features differ in the model T42 rack from the model T00:
The T42 rack has 42U (EIA units) of usable space (6U of additional space).
The model T42 supports AC power only.
Weights are as follows:
– T42 base empty rack: 261 kg (575 lb.)
– T42 full rack: 930 kg (2045 lb.)
The feature #ERG7 provides an attractive black full-height rack door. The door is steel,
with a perforated flat front surface. The perforation pattern extends from the bottom to the
top of the door to enhance ventilation and provide some visibility into the rack.
The feature #6069 provides a cost-effective plain front door.
The feature #6249 provides a special acoustic door
A minimum of one of the following is required:
– Feature #ER2B allows you to reserve 2U of space at the bottom of the rack.
– Feature #ER2T allows you to reserve 2U of space at the top of the rack.
A rear rack extension of 8 inches or 20.3 cm (#ERG0) provides space to hold cables on the
side of the rack and keep the center area clear for cooling and service access. Including this
extension is very, very strongly recommended where large numbers of thicker I/O cables are
present or may be added in the future. The definition of a large number depends on the type
of I/O cables used. Probably around 64 short-length SAS cables per side of a rack or around
50 longer-length (thicker) SAS cables per side of a rack is a good rule of thumb. Generally,
other I/O cables are thinner and easier to fit in the sides of the rack and the number of cables
can be higher. SAS cables are most commonly found with multiple EXP24S SAS drawers
(#5887) driven by multiple PCIe SAS adapters. For this reason, it can be a very good practice
to keep multiple EXP24S drawers in the same rack as the PCIe Gen3 I/O drawer or in a
separate rack close to the PCIe Gen3 I/O drawer, using shorter, thinner SAS cables. The
feature ERG0 extension can be good to use even with a smaller numbers of cables as it
enhances the ease of cable management with the extra space it provides.
1.13.3 IBM 7953 model 94Y rack
The 2.0-meter (79.3 inch) model 94Y rack has the following features:
42U (EIA units)
Weights:
– Base empty rack: 187 kg (221 lb.)
– Maximum load limit: 664 kg (1460 lb.)
Special door: The Power 780 logo rack door (#6250) is not supported.
Recommended Rack: The 7014-T42 System rack with additional rear extension (#ERG0)
is recommended for all initial Power E870 and Power E880 system orders. These are the
default options for new orders.

The IBM 42U Slim Rack (7953-94Y) differs from the IBM 42U enterprise rack (7014-T42) in
several aspects. Both provide 42U of vertical space, are 1100 mm deep, and have an interior
rail-to-rail depth of 715 mm. However, the IBM 42U Slim Rack is 600 mm wide; the T42 is
645 mm wide with side covers. For clients with 2-foot floor tiles, the extra 45 mm (1.77-inch)
width of the enterprise rack can sometimes cause challenges when cutting holes in the floor
tiles for cabling.
The 42U Slim Rack has a lockable perforated front steel door, providing ventilation, physical
security, and visibility of indicator lights in the installed equipment within. In the rear, either a
lockable perforated rear steel door (#EC02) or a lockable rear door heat exchanger (#EC15)
is used. Lockable optional side panels (#EC03) increase the rack’s aesthetics, help control
airflow through the rack, and provide physical security. Multiple 42U Slim Racks can be bolted
together to create a rack suite (indicated with feature #EC04).
1.13.4 Feature code #0551 rack
The 1.8-meter rack (#0551) is a 36U (EIA units) rack. The rack that is delivered as #0551 is
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 #0551 is not available for initial orders of Power E870 and E880 servers.
1.13.5 Feature code #0553 rack
The 2.0-meter rack (#0553) is a 42U (EIA units) rack. The rack that is delivered as #0553 is
the same rack delivered when you order the 7014-T42. The included features might differ.
Several features that are delivered as part of the 7014-T42 or must be ordered separately
with the #0553.
1.13.6 The AC power distribution unit and rack content
For rack models T00, T42, and the slim 94Y, 12-outlet PDUs are available. The PDUs
available include these:
PDUs Universal UTG0247 Connector (#7188)
Intelligent PDU+ Universal UTG0247 Connector (#7109)
When mounting the horizontal PDUs, it is a good practice to place them almost at the top or
almost at the bottom of the rack, leaving 2U or more of space at the very top or very bottom of
the rack for cable management. Mounting a horizontal PDU in the middle of the rack is
generally not optimal for cable management.
Weight calculation: Maximum weight limits must include everything that will be installed
in the rack. This must include servers, I/O drawers, PDUs, switches, and anything else
installed. In zone 4 earthquake environments, the rack should be configured starting with
the heavier drawers at the bottom of the rack.
PDU mounting: Only horizontal PDUs are allowed in racks hosting Power E870 and
Power E880 system. Vertically mounted PDUs limit access to the cable routing space on
the side of the rack and cannot be used.

For the Power E870 or E880 installed in IBM 7014 or FC 055x racks, the following PDU rules
apply:
For PDU #7109 and #7188 when using 24 Amp power cord #6654, #6655, #6656, #6657,
or #6658, each pair of PDUs can power one Power E870 or Power E880 system node and
two I/O expansion drawers, or eight I/O expansion drawers. 24A PDU cables are used to
supply 30A PDUs. In Figure 1-12 on page 33, you can see the rack configuration with two
pairs of 30A PDUs suppling a two system node configuration and two I/O expansion
drawers.
Figure 1-12 Two system node configuration and two I/O expansion drawers supplied by 30A PDUs
41
42
39
40
37
38
35
36
33
34
31
32
29
30
27
28
25
26
23
24
21
22
19
20
17
18
15
16
13
14
11
12
09
10
07
08
05
06
03
04
01
02
41
42
39
40
37
38
35
36
33
34
31
32
29
30
27
28
25
26
23
24
21
22
19
20
17
18
15
16
13
14
11
12
09
10
07
08
05
06
03
04
01
02
System Node
#1
Control Unit
I/O Expansion Drawer
P4P3P1 P2
30A PDU
30A PDU
P1 P2
P1 P2
System Node
#2
Control Unit
P4P3P1 P2
P4P3P1 P2
30A PDU
30A PDU
30A PDU
30A PDU
P2P1
P2P1

For PDU #7109 and #7188 when using three phase power cords or 48 Amp power cords
#6491 or #6492, each pair of PDUs can power up to two Power E870 or Power E880
system nodes and two I/O expansion drawers, or eight I/O expansion drawers. 48A PDU
cables are used to supply 60A PDU. In Figure 1-13 on page 34, you can see the rack
configuration with two pairs of 60A PDUs suppling a two system node configuration, four
I/O expansion drawers, and four EXP24S disk drawers.
Figure 1-13 Two system nodes configuration with I/O expansions and disk drawers supplied by 60A
PDUs
For detailed power cord requirements and power cord feature codes, see the IBM Power
Systems Hardware Knowledge Center website:
http://www-01.ibm.com/support/knowledgecenter/9119-MME/p8had/p8had_rpower.htm
Power cord: Ensure that the appropriate power cord feature is configured to support the
power being supplied.
41
42
39
40
37
38
35
36
33
34
31
32
29
30
27
28
25
26
23
24
21
22
19
20
17
18
15
16
13
14
11
12
09
10
07
08
05
06
03
04
01
02
41
42
39
40
37
38
35
36
33
34
31
32
29
30
27
28
25
26
23
24
21
22
19
20
17
18
15
16
13
14
11
12
09
10
07
08
05
06
03
04
01
02
System Node
#2
System Node
#1
Control Unit
EXP24S
EXP24S
EXP24S
EXP24S
P4P3P1 P2
P4P3P1 P2
60A PDU
60A PDU
60A PDU
60A PDU
P2P1
P2P1
P2P1
P2P1
P2P1
P2P1
P2P1
P2P1

For rack integrated systems, a minimum quantity of two PDUs (#7109 or #7188) are required.
The PDUs (#7109, #7188) 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 countries
and applications by selecting a 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 PDUs have 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 20 A circuit breaker.
The Universal PDUs are compatible with previous models.
Power cord and PDU: Based on the power cord that is used, the PDU can supply a range
of 4.8 - 21 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 node mounted in the rack requires four power cords. 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.

Chapter 2. Architecture and technical
overview
This chapter describes the overall system architecture for the IBM Power System E870
(9119-MME) and IBM Power System E880 (9119-MHE) servers. The bandwidths that are
provided throughout the section are theoretical maximums that are used for reference.
The speeds that are 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 by using real-world performance measurements and production workloads.
2

2.1 Logical diagrams
This section contains logical diagrams for the Power E870 and E880.
Figure 2-1 shows the logical system diagram for a single system node of a Power E870 or
Power E880.
Figure 2-1 Logical system diagram for a system node of a Power E870 or a Power E880
SCU Interface
P8 (1)Memory
Controller0
Memory
Controller1
PHB0
PHB1
P8 (2)
PHB0
PHB1
P8 (0)
PHB0
PHB1
P8 (3)
PHB0
PHB1
SMP
XBus
SMP
XBus
SMP
XBus
SMP
XBus
SMP
A Bus
SMP
A Bus
SMP
A Bus
SMP
A Bus
PCIeGen3x16slot
PCIeGen3x16slot
PCIeGen3x16slot
PCIeGen3x16slot
PCIeGen3x16slot
PCIeGen3x16slot
PCIeGen3x16slot
PCIeGen3x16slot
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
Memory
Controller0
Memory
Controller1
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
Memory
Controller0
Memory
Controller1
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
Memory
Controller0
Memory
Controller1
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
CDIMM
To other
enclosures
To other
enclosures
To other
enclosures
To other
enclosures
SCU Interface
card
12v power
to SCU
Service
Processor
Connection
Clock
Interface
To SCU
Clock
SCU Interface
SCU Interface
card
12v power
to SCU
Service
Processor
Connection
Clock
Interface
To SCU
Clock
25.6 GB/s
76.8 GB/s
15.75 GB/s
28.8 GB/s

Chapter 2. Architecture and technical overview 39
Figure 2-2 shows the logical system diagram for the system control unit of a Power E870 or a
Power E880.
Figure 2-2 Logical system diagram for the system control unit
Power in
Operator Panel
From
CEC
Power in
From
CEC
Power in
From
CEC
Power in
From
CEC
USB
cable
to
PCIe card
DVD
FSP 0 FSP 1Clock 0 Clock 1
VPD

Flexible symmetric multiprocessing (SMP) cables are used to connect system nodes when a
Power E870 or Power E880 is configured with more than one system node. Figure 2-3 shows
the SMP connection topology for a two-drawer Power E870 or Power E880 system.
Figure 2-3 SMP connection topology for a two-drawer Power E870 or Power E880
Drawer 1
Drawer 0
X3 X1 X0 X3 X1 X0 X3 X1 X0 X3 X1 X0
X0 X1 X3 X0 X1 X3 X0 X1 X3 X0 X1 X3
P0 P2 P3 P1
T1 T2 T5 T6T3 T4
T6 T5 T4 T3 T2 T1
A2 A1 A0 A2 A1 A0 A2 A1 A0 A2 A1 A0
A0 A1 A2 A0 A1 A2 A0 A1 A2 A0 A1 A2
P1 P3 P2 P0

Figure 2-4 shows the SMP connection topology for a three-drawer Power E880.
Figure 2-4 SMP connection topology for a three-drawer Power E880
T1T2T5T6T3T4
P3
X3
X1
X0
A2
A1
A0
Drawer3
P2
X3
X1
X0
A2
A1
A0
P0
X3
X1
X0
A2
A1
A0
P1
X3
X1
X0
A2
A1
A0
Drawer 2
T1 T2 T5 T6T3 T4
Drawer 1
X3 X1 X0 X3 X1 X0 X3 X1 X0 X3 X1 X0
P0 P2 P3 P1
A2 A1 A0 A2 A1 A0 A2 A1 A0 A2 A1 A0
T6 T5 T4 T3 T2 T1
A0 A1 A2A0 A1 A2 A0 A1 A2 A0 A1 A2
X0 X1 X3X0 X1 X3 X0 X1 X3 X0 X1 X3
P1 P3 P2 P0

Figure 2-5 shows the SMP connection topology for a four-drawer Power E880.
Figure 2-5 SMP connection topology for a four-drawer Power E880
2.2 The IBM POWER8 processor
This section introduces the latest processor in the IBM Power Systems product family, and
describes its main characteristics and features.
T1T2T5T3T4
P3
X3
X1
X0
A2
A1
A0
Drawer3
P2
X3
X1
X0
A2
A1
A0
P0
X3
X1
X0
A2
A1
A0
P1
X3
X1
X0
A2
A1
A0
Drawer 2
Drawer4
T1 T2 T5 T6T3 T4
Drawer 1
X3 X1 X0
P0 P2 P3 P1
X3 X1 X0 X3 X1 X0 X3 X1 X0
A2 A1 A0 A2 A1 A0 A2 A1 A0 A2 A1 A0
T6
T6T5T2T4T3T1
T6 T5 T4 T3 T2 T1
A2
A1
A0
A2
A1
A0
A2
A1
A0
A2
A1
A0
X3
X1
X0
X3
X1
X0
X3
X1
X0
X3
X1
X0
P0
P2
P3
P1
A0 A1 A2 A0 A1 A2 A0 A1 A2 A0 A1 A2
X0 X1 X3 X0 X1 X3 X0 X1 X3 X0 X1 X3
P1 P3 P2 P0

2.2.1 POWER8 processor overview
The POWER8 processor is manufactured using the IBM 22 nm Silicon-On-Insulator (SOI)
technology. Each chip is 649 mm2 and contains 4.2 billion transistors. As shown in Figure 2-6,
the chip contains 12 cores, two memory controllers, PCIe Gen3 I/O controllers, and an
interconnection system that connects all components within the chip. On some systems only
6, 8, 10, or 12 cores per processor may be available to the server. Each core has 512 KB of
L2 cache, and all cores share 96 MB of L3 embedded DRAM (eDRAM). The interconnect
also extends through module and board technology to other POWER8 processors in addition
to DDR3 memory and various I/O devices.
POWER8 systems use memory buffer chips to interface between the POWER8 processor
and DDR3 or DDR4 memory. Each buffer chip also includes an L4 cache to reduce the
latency of local memory accesses.
Figure 2-6 The POWER8 processor chip
The POWER8 processor is for system offerings from single-socket servers to multi-socket
Enterprise servers. It incorporates a triple-scope broadcast coherence protocol over local and
global SMP links to provide superior scaling attributes. Multiple-scope coherence protocols
reduce the amount of SMP link bandwidth that is required by attempting operations on a
limited scope (single chip or multi-chip group) when possible. If the operation cannot
complete coherently, the operation is reissued using a larger scope to complete the operation.
The following additional features can augment the performance of the POWER8 processor:
Support is provided for DDR3 and DDR4 memory through memory buffer chips that
offload the memory support from the POWER8 memory controller.
Each memory CDIMM has 16 MB of L4 cache within the memory buffer chip that reduces
the memory latency for local access to memory behind the buffer chip; the operation of the

L4 cache is not apparent to applications running on the POWER8 processor. Up to 128
MB of L4 cache can be available for each POWER8 processor.
Hardware transactional memory.
On-chip accelerators, including on-chip encryption, compression, and random number
generation accelerators.
Coherent Accelerator Processor Interface, which allows accelerators plugged into a PCIe
slot to access the processor bus using a low latency, high-speed protocol interface.
Adaptive power management.
There are two versions of the POWER8 processor chip. Both chips use the same building
blocks. The scale-out systems use a 6-core version of POWER8. The 6-core chip is installed
in pairs in a Dual Chip Module (DCM) that plugs into a socket in the system board of the
systems. Functionally, it works as a single chip module (SCM).
The Enterprise servers use a 12-core chip installed on a Single Chip module (SCM) that
plugs into a socket in the system board of the systems. SCMs are used in the Power E870
and Power E880 systems.
Figure 2-7 shows a graphic representation of the 6-core processor on a DCM. DCMs are only
available on the scale-out systems. It is shown here for informational purposes.
Figure 2-7 6-core POWER8 processor chip on a DCM

Figure 2-8 shows a graphic representation of the 12-core processor on a SCM. The SCM is
used in Enterprise systems. It is shown here for informational purposes.
Figure 2-8 12-core POWER8 processor chip on an SCM
Table 2-1 summarizes the technology characteristics of the POWER8 processor.
Table 2-1 Summary of POWER8 processor technology
2.2.2 POWER8 processor core
The POWER8 processor core is a 64-bit implementation of the IBM Power Instruction Set
Architecture (ISA) Version 2.07 and has the following features:
Multi-threaded design, which is capable of up to eight-way simultaneous multithreading
(SMT)
32 KB, eight-way set-associative L1 instruction cache
64 KB, eight-way set-associative L1 data cache
Technology POWER8 processor
Die size 649 mm2
Fabrication technology 22 nm lithography
Copper interconnect
SOI
eDRAM
Maximum processor cores 6 or 12
Maximum execution threads core/chip 8/96
Maximum L2 cache core/chip 512 KB/6 MB
Maximum On-chip L3 cache core/chip 8 MB/96 MB
Maximum L4 cache per chip 128 MB
Maximum memory controllers 2
SMP design-point 16 sockets with IBM POWER8 processors
Compatibility With prior generations of POWER processor

Enhanced prefetch, with instruction speculation awareness and data prefetch depth
awareness
Enhanced branch prediction, using both local and global prediction tables with a selector
table to choose the best predictor
Improved out-of-order execution
Two symmetric fixed-point execution units
Two symmetric load/store units and two load units, all four of which can also run simple
fixed-point instructions
An integrated, multi-pipeline vector-scalar floating point unit for running both scalar and
SIMD-type instructions, including the Vector Multimedia eXtension (VMX) instruction set
and the improved Vector Scalar eXtension (VSX) instruction set, and capable of up to
eight floating point operations per cycle (four double precision or eight single precision)
In-core Advanced Encryption Standard (AES) encryption capability
Hardware data prefetching with 16 independent data streams and software control
Hardware decimal floating point (DFP) capability
More information about Power ISA Version 2.07 can be found at the following website:
https://www.power.org/documentation/power-isa-v-2-07b/
Figure 2-9 shows a picture of the POWER8 core, with some of the functional units
highlighted.
Figure 2-9 POWER8 processor core
2.2.3 Simultaneous multithreading
POWER8 processor advancements in multi-core and multi-thread scaling are remarkable. A
significant performance opportunity comes from parallelizing workloads to enable the full
potential of the microprocessor, and the large memory bandwidth. Application scaling is
influenced by both multi-core and multi-thread technology.
Simultaneous Multithreading (SMT) allows a single physical processor core to simultaneously
dispatch instructions from more than one hardware thread context. With SMT, each POWER8

core can present eight hardware threads. Because there are multiple hardware threads per
physical processor core, additional instructions can run at the same time. SMT is primarily
beneficial in commercial environments where the speed of an individual transaction is not as
critical as the total number of transactions that are performed. SMT typically increases the
throughput of workloads with large or frequently changing working sets, such as database
servers and web servers.
Table 2-2 shows a comparison between the different POWER processors in terms of SMT
capabilities that are supported by each processor architecture.
Table 2-2 SMT levels that are supported by POWER processors
The architecture of the POWER8 processor, with its larger caches, larger cache bandwidth,
and faster memory, allows threads to have faster access to memory resources, which
translates in to a more efficient usage of threads. Because of that, POWER8 allows more
threads per core to run concurrently, increasing the total throughput of the processor and of
the system.
2.2.4 Memory access
On the Power E870 and Power E880, each POWER8 processor has two memory controllers,
each connected to four memory channels. Each memory channel operates at 1600 MHz and
connects to a DIMM. Each DIMM on a POWER8 system has a memory buffer that is
responsible for many functions that were previously on the memory controller, such as
scheduling logic and energy management. The memory buffer also has 16 MB of level 4 (L4)
cache.
On the Power E870, each memory channel can address up to 256 GB. Therefore a single
system node can address 8 TB of memory. A two node system can address up to 16 TB of
memory.
On the Power E880, each memory channel can address up to 256 GB. Therefore a single
system node can address 8 TB of memory. A two node system can address up to 16 TB of
memory and a four node system can address up to 32 TB of memory.
Technology Cores/system Maximum SMT mode Maximum hardware
threads per system
IBM POWER4 32 Single Thread (ST) 32
IBM POWER5 64 SMT2 128

Figure 2-10 gives a simple overview of the POWER8 processor memory access structure in
the Power E870 and Power E880.
Figure 2-10 Overview of POWER8 memory access structure
2.2.5 On-chip L3 cache innovation and Intelligent Cache
Similar to POWER7 and POWER7+, the POWER8 processor uses a breakthrough in material
engineering and microprocessor fabrication to implement the L3 cache in eDRAM and place it
on the 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 8 MB 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 that is associated with another core, depending on the
reference patterns. Data can also be cloned and stored in more than one core’s FLR-L3
cache, again depending on the reference patterns. This Intelligent Cache management
enables the POWER8 processor to optimize the access to L3 cache lines and minimize
overall cache latencies.
Figure 2-6 on page 43 and Figure 2-7 on page 44 show the on-chip L3 cache, and highlight
one fast 8 MB L3 region closest to a processor core.
The benefits of using eDRAM on the POWER8 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.
POWER8
Processor
Memory
Buffers
DRAM
Chips
Memory
Buffers
DRAM
Chips

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 that conventional SRAM uses, 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.
2.2.6 Level 4 cache and memory buffer
POWER8 processor-based systems introduce an additional level of memory hierarchy. The
Level 4 (L4) cache is implemented together with the memory buffer in the Custom DIMM
(CDIMM). Each memory buffer contains 16 MB of L4 cache. Figure 2-11 shows a picture of
the memory buffer, where you can see the 16 MB L4 cache, and processor links and memory
interfaces.
Figure 2-11 Memory buffer chip
Table 2-3 shows a comparison of the different levels of cache in the POWER7, POWER7+,
and POWER8 processors.
Table 2-3 POWER8 cache hierarchy
Cache POWER7 POWER7+ POWER8
L1 instruction cache:
Capacity/associativity
32 KB, 4-way 32 KB, 4-way 32 KB, 8-way
L1 data cache:
bandwidth
32 KB, 8-way
Two 16 B reads or
one 16 B writes per cycle
32 KB, 8-way
Two 16 B reads or
64 KB, 8-way
Two 16 B reads or

For more information about the POWER8 memory subsystem, see 2.3, “Memory subsystem”
on page 53.
2.2.7 Hardware transactional memory
Transactional memory is an alternative to lock-based synchronization. It attempts to simplify
parallel programming by grouping read and write operations and running them like a single
operation. Transactional memory is like database transactions where all shared memory
accesses and their effects are either committed all together or discarded as a group. All
threads can enter the critical region simultaneously. If there are conflicts in accessing the
shared memory data, threads try accessing the shared memory data again or are stopped
without updating the shared memory data. Therefore, transactional memory is also called a
lock-free synchronization. Transactional memory can be a competitive alternative to
lock-based synchronization.
Transactional Memory provides a programming model that makes parallel programming
easier. A programmer delimits regions of code that access shared data and the hardware
runs these regions atomically and in isolation, buffering the results of individual instructions,
and retrying execution if isolation is violated. Generally, transactional memory allows
programs to use a programming style that is close to coarse-grained locking to achieve
performance that is close to fine-grained locking.
Most implementations of transactional memory are based on software. The POWER8
processor-based systems provide a hardware-based implementation of transactional memory
that is more efficient than the software implementations and requires no interaction with the
processor core, therefore allowing the system to operate at maximum performance.
2.2.8 Coherent Accelerator Processor Interface
The Coherent Accelerator Interface Architecture (CAIA) defines a coherent accelerator
interface structure for attaching peripherals to Power Systems.
The Coherent Accelerator Processor Interface (CAPI) can attach accelerators that have
coherent shared memory access to the processors in the server and share full virtual address
translation with these processors, using a standard PCIe Gen3 bus.
Applications can have customized functions in Field Programmable Gate Arrays (FPGA) and
be able to enqueue work requests directly in shared memory queues to the FPGA, and using
the same effective addresses (pointers) it uses for any of its threads running on a host
L2 cache:
bandwidth
256 KB, 8-way
Private
32 B reads and 16 B writes
per cycle
256 KB, 8-way
Private
per cycle
512 KB, 8-way
Private
per cycle
L3 cache:
bandwidth
On-Chip
4 MB/core, 8-way
per cycle
On-Chip
10 MB/core, 8-way
per cycle
On-Chip
8 MB/core, 8-way
per cycle
L4 cache:
bandwidth
N/A N/A Off-Chip
16 MB/buffer chip, 16-way
Up to 8 buffer chips per
socket
Cache POWER7 POWER7+ POWER8

processor. From the practical perspective, CAPI allows a specialized hardware accelerator to
be seen as an additional processor in the system, with access to the main system memory,
and coherent communication with other processors in the system.
The benefits of using CAPI include the ability to access shared memory blocks directly from
the accelerator, perform memory transfers directly between the accelerator and processor
cache, and reduction in the code path length between the adapter and the processors. This is
because the adapter is not operating as a traditional I/O device, and there is no device driver
layer to perform processing. It also presents a simpler programming model.
Figure 2-12 on page 51 shows a high-level view of how an accelerator communicates with the
POWER8 processor through CAPI. The POWER8 processor provides a Coherent Attached
Processor Proxy (CAPP), which is responsible for extending the coherence in the processor
communications to an external device. The coherency protocol is tunneled over standard
PCIe Gen3, effectively making the accelerator part of the coherency domain.
The accelerator adapter implements the Power Service Layer (PSL), which provides address
translation and system memory cache for the accelerator functions. The custom processors
on the board, consisting of an FPGA or an Application Specific Integrated Circuit (ASIC) use
this layer to access shared memory regions, and cache areas as though they were a
processor in the system. This ability greatly enhances the performance of the data access for
the device and simplifies the programming effort to use the device. Instead of treating the
hardware accelerator as an I/O device, it is treated as a processor. This eliminates the
requirement of a device driver to perform communication, and the need for Direct Memory
Access that requires system calls to the operating system kernel. By removing these layers,
the data transfer operation requires fewer clock cycles in the processor, greatly improving the
I/O performance.
Figure 2-12 CAPI accelerator that is attached to the POWER8 processor
The implementation of CAPI on the POWER8 processor allows hardware companies to
develop solutions for specific application demands and use the performance of the POWER8
processor for general applications and custom acceleration of specific functions using a
hardware accelerator, with a simplified programming model and efficient communication with
the processor and memory resources.
Custom
Hardware
Application
CAPP
Coherence Bus
PSL
FPGA or ASIC
POWER8
PCIe Gen3
Transport for encapsulated messages

2.2.9 Power management and system performance
The POWER8 processor has power saving and performance enhancing features that can be
used to lower overall energy usage, while yielding higher performance when needed. The
following modes can be enabled and modified to use these features.
Dynamic Power Saver: Favor Performance
This mode is intended to provide the best performance. If the processor is being used even
moderately, the frequency is raised to the maximum frequency possible to provide the best
performance. If the processors are lightly used, the frequency is lowered to the minimum
frequency, which is potentially far below the nominal shipped frequency, to save energy. The
top frequency that is achieved is based on system type and is affected by environmental
conditions. Also, when running at the maximum frequency, more energy is being consumed,
which means this mode can potentially cause an increase in overall energy consumption.
Dynamic Power Saver: Favor Power
This mode is intended to provide the best performance per watt consumed. The processor
frequency is adjusted based on the processor usage to maintain the workload throughput
without using more energy than required to do so. At high processor usage levels, the
frequency is raised above nominal, as in the Favor Performance mode. Likewise, at low
processor usage levels, the frequency is lowered to the minimum frequency. The frequency
ranges are the same for the two Dynamic Power Saver modes, but the algorithm that
determines which frequency to set is different.
Dynamic Power Saver: Tunable Parameters
The Dynamic Power Saver: Favor Performance and Dynamic Power Saver: Favor Power
modes are tuned to provide both energy savings and performance increases. However, there
might be situations where only top performance is of concern, or, conversely, where peak
power consumption is an issue. The tunable parameters can be used to modify the setting of
the processor frequency in these modes to meet these various objectives. Modifying these
parameters should be done only by advanced users. If you must address any issues
concerning the Tunable Parameters, IBM support personal should be directly involved in the
parameter value selection.
Idle Power Saver
This mode is intended to save the maximum amount of energy when the system is nearly idle.
When the processors are found to be nearly idle, the frequency of all processors is lowered to
the minimum. Additionally, workloads are dispatched onto a smaller number of processor
cores so that the other processor cores can be put into a low energy usage state. When
processor usage increases, the process is reversed: The processor frequency is raised back
up to nominal, and the workloads are spread out once again over all of the processor cores.
There is no performance boosting aspect in this mode, but entering or exiting this mode might
affect overall performance. The delay times and usage levels for entering and exiting this
mode can be adjusted to allow for more or less aggressive energy savings.
The controls for all of these modes are available on the Advanced System Management
Interface (ASMI) and are described in more detail in a white paper that is found at the
following link:
http://public.dhe.ibm.com/common/ssi/ecm/po/en/pow03125usen/POW03125USEN.PDF
For more information, see 2.13, “Energy management” on page 113.

2.2.10 Comparison of the POWER7, POWER7+, and POWER8 processors
Table 2-4 shows comparable characteristics between the generations of POWER7,
POWER7+, and POWER8 processors.
Table 2-4 Comparison of technologies for the POWER8 processor and the prior generations
2.3 Memory subsystem
The Power E870 can have up to two system nodes per system with each system node having
32 CDIMM slots capable of supporting 16 GB, 32 GB, 64 GB, 128 GB, and 256 GB CDIMMs
running at speeds of 1600 MHz. This allows for a maximum system memory of 16 TB for the
256 CDIMM slots in a system comprised of two system nodes.
The Power E880 can have up to four system nodes per system with each system node having
32 CDIMM slots capable of supporting 16 GB, 32 GB, 64 GB, 128 GB, and 256 GB CDIMMs
running at speeds of 1600 MHz. This allows for a maximum system memory of 32 TB for the
256 CDIMM slots of a system comprised of four system nodes.
The memory on the systems are Capacity on Demand capable, allowing for the purchase of
additional physical memory capacity and dynamically activate it when needed. It is required
that at least 50% of the installed memory capacity is active.
Characteristics POWER7 POWER7+ POWER8
Technology 45 nm 32 nm 22 nm
Die size 567 mm2
567 mm2
649 mm2
Number of transistors 1.2 billion 2.1 billion 4.2 billion
Maximum cores 8 8 12
Maximum SMT
threads per core
4 threads 4 threads 8 threads
Maximum frequency 4.25 GHz 4.4 GHz 4.35 GHz
L2 Cache 256 KB per core 256 KB per core 512 KB per core
L3 Cache 4 MB or 8 MB of
FLR-L3 cache per core
with each core having
access to the full 32
MB of L3 cache,
on-chip eDRAM
10 MB of FLR-L3
cache per core with
each core having
access to the full 80
MB of L3 cache,
on-chip eDRAM
8 MB of FLR-L3 cache
per core with each
core having access to
the full 96 MB of L3
cache, on-chip
eDRAM
Memory support DDR3 DDR3 DDR3 and DDR4
I/O bus GX++ GX++ PCIe Gen3

The Power E870 and E880 servers support an optional feature called Active Memory
Expansion (#EM82). This allows the effective maximum memory capacity to be much larger
than the true physical memory. This feature runs innovative compression and decompression
of memory content by using a dedicated coprocessor present on each POWER8 processor to
provide memory expansion up to 125%, depending on the workload type and its memory
usage. As an example, a server with 256 GB of memory physically installed can effectively be
expanded over 512 GB of memory. This approach can enhance virtualization and server
consolidation by allowing a partition to do more work with the same physical amount of
memory or allowing a server to run more partitions and do more work with the same physical
amount of memory.
2.3.1 Custom DIMM
Custom DIMMs (CDIMMs) are innovative memory DIMMs that house industry-standard
DRAM memory chips and a set of components that allow for higher bandwidth, lower latency
communications and increased availability. These components include:
Memory Scheduler
Memory Management (RAS Decisions & Energy Management)
Memory Buffer
By adopting this architecture for the memory DIMMs, several decisions and processes
regarding memory optimizations are run internally into the CDIMM. This saves bandwidth and
allows for faster processor-to-memory communications. This also allows for a more robust
RAS. For more information, see Chapter 4, “Reliability, availability, and serviceability” on
page 159.
The CDIMMs exists in two different form factors, a 152 SDRAM design named the Tall
CDIMM and an 80 SDRAM design named the Short CDIMM. Each design will be comprised
of multiple 4 GB SDRAM devices depending on its total capacity. The CDIMM slots for the
Power E870 and Power E880 are tall CDIMMs slots. A filler is added to the short CDIMM
allowing it to properly latch into the same physical location of a tall CDIMM and allows for
proper airflow and ease of handling. Tall CDIMMs slots allow for larger DIMM sizes and the
adoption of future technologies more seamlessly.

A detailed diagram of the CDIMMs available for the Power E870 and Power E880 can be
seen in Figure 2-13.
Figure 2-13 Short CDIMM and Tall CDIMM details
The Memory Buffer is a L4 cache and is built on eDRAM technology (same as the L3 cache),
which has a lower latency than regular SRAM. Each CDIMM has 16 MB of L4 cache and a
fully populated Power E870 server has 1 GB of L4 Cache while a fully populated Power E880
has 2 GB of L4 Cache. The L4 Cache performs several functions that have direct impact on
performance and bring a series of benefits for the Power E870 and Power E880:
Reduces energy consumption by reducing the number of memory requests.
Increases memory write performance by acting as a cache and by grouping several
random writes into larger transactions.
Partial write operations that target the same cache block are gathered within the L4 cache
before being written to memory, becoming a single write operation.
Reduces latency on memory access. Memory access for cached blocks has up to 55%
lower latency than non-cached blocks.
2.3.2 Memory placement rules
For the Power E870 and Power E880, each memory feature code provides four CDIMMs.
Therefore a maximum of 8 memory feature codes per system node are allowed in order to fill
all the 32 CDIMM slots.
All the memory CDIMMs are capable of capacity upgrade on demand and must have a
minimum of 50% of their physical capacity activated. For example, the minimum installed
memory for both servers is 256 GB, which requires a minimum of 128 GB active.
Shot Form Factor+ Filler
DRAM Chips:
• 16 GB = 40 SDRAM Chips
1600 MHz
DRAM Sparing
Quad Interleave
16 MB L4 Cache eDRAM
Tall Form Factor
DRAM Chips:
• 128GB = 152 SDRAM Chips
• 256GB = 152 SDRAM Chips
1600 MHz
DRAM Sparing
QuadInterleave
16 MB L4 Cache eDRAM

For the Power E870 and Power E880, the following memory options are orderable:
64 GB (4 X 16 GB) CDIMMs, 1600 MHz DDR3 DRAM (#EM8J)
128 GB (4 X 32 GB) CDIMMs, 1600 MHz DDR3 DRAM (#EM8K)
256 GB (4 X 64 GB) CDIMMs, 1600 MHz DDR3 DRAM (#EM8L)
512 GB (4 X 128 GB) CDIMMs, 1600 MHz DDR3 DRAM (#EM8M)
1024_GB (4 X 256 GB) CDIMMs, 1600 Mhz DDR4 DRAM (#EM8Y)
Each processor has two memory controllers. These memory controllers must have at least a
pair of CDIMMs attached to it. This set of mandatory four CDIMMs is called a memory quad.
A logical diagram of a POWER8 processor with its two memory quads can be seen in
Figure 2-14.
Figure 2-14 Logical diagram of a POWER8 processor and its two memory quads
The basic rules for memory placement follow:
Each feature code equals a set of four physical CDIMMs; a memory quad.
Each installed processor must have at least one memory quad populated, which equals to
at least one feature code per installed processor.
A given processor can only have four or eight CDIMMs attached to it.
All the CDIMMs connected to the same POWER8 processor must be identical. However, it
is permitted to mix different CDIMM sizes between different POWER8 processors on a
system.
At least 50% of the installed memory must be activated via memory activation features.
The suggested approach is to install memory evenly across all processors in the system and
across all system nodes in a system. Balancing memory across the installed processors
allows memory access in a consistent manner and typically results in the best possible
performance for your configuration. You should account for any plans for future memory
upgrades when you decide which memory feature size to use at the time of the initial system
order.
Note: The DDR4 #EM8Y CDIMM has the following usage guidelines:
No mixing #EM8Y with another size CDIMM on the same system node
At the time of writing, no mixing #EM8Y with another size CDIMM on the same server
100% of the memory slots must be filled.
Firmware 840.12 is the required minimum level
Development may change these usage guidelines at any time.

A physical diagram with the location codes of the memory CDIMMs of a system node, as well
as their grouping as memory quads, can be seen in Figure 2-15.
Figure 2-15 System node physical diagram with location codes for CDIMMs
Each system node has eight memory quads identified by the different colors in Figure 2-15.
These are the location codes for the slots on each memory quad:
Quad 1: P1-C45, P1-C46, P1-C51, and P1-C52
POWER8
Processor
1
MC
1
MC
0
POWER8
Processor
3
MC
1
MC
0
POWER8
Processor
2
MC
1
MC
0
POWER8
Processor
0
MC
1
MC
0
P1-C52 (Quad 1)
P1-C51 (Quad 1)
P1-C50 (Quad 5)
P1-C49 (Quad 5)
P1-C48 (Quad 5)
P1-C47 (Quad 5)
P1-C46 (Quad 1)
P1-C45 (Quad 1)
P1-C44 (Quad 2)
P1-C43 (Quad 2)
P1-C42 (Quad 6)
P1-C41 (Quad 6)
P1-C40 (Quad 6)
P1-C39 (Quad 6)
P1-C38 (Quad 2)
P1-C37 (Quad 2)
P1-C36 (Quad 3)
P1-C35 (Quad 3)
P1-C34 (Quad 7)
P1-C33 (Quad 7)
P1-C32 (Quad 7)
P1-C31 (Quad 7)
P1-C30 (Quad 3)
P1-C29 (Quad 3)
P1-C28 (Quad 4)
P1-C27 (Quad 4)
P1-C26 (Quad 8)
P1-C25 (Quad 8)
P1-C24 (Quad 8)
P1-C23 (Quad 8)
P1-C22 (Quad 4)
P1-C21 (Quad 4)
Location (Memory Quad)
Front
Rear

Table 2-5 shows the CDIMM plugging order for a Power E870 or Power E880 with a single
system node.
Table 2-5 Optimal CDIMM memory quad placement for a single system node
Table 2-6 shows the CDIMM plugging order for a Power E870 or Power E880 with two system
nodes.
Table 2-6 Optimal CDIMM memory quad placement for two system nodes
System Node 1
Processor 0 Processor 2 Processor 3 Processor 1
Quad 1 Quad 5 Quad 2 Quad 6 Quad 3 Quad 7 Quad 4 Quad 8
1 5 2 6 3 7 4 8
Notes:
Memory quads 1-4 must be populated.
Memory quads on the same processor must be populated with CDIMMs of the same capacity.
System Node 1
1 9 2 11 3 13 4 15
System Node 2
5 10 6 12 7 14 8 16
Notes:

Table 2-7 shows the CDIMM plugging order for a Power E880 with a three system nodes.
Table 2-7 Optimal CDIMM memory quad placement for three system nodes
System Node 1
1 13 2 16 3 19 4 22
System Node 2
5 14 6 17 7 20 8 23
System Node 3
9 15 10 18 11 21 12 24
Notes:

Table 2-8 shows the CDIMM plugging order for a Power E880 with a four system nodes.
Table 2-8 Optimal CDIMM memory quad placement for four system nodes
2.3.3 Memory activation
All the memory CDIMMs are capable of capacity upgrade on demand and must have a
minimum of 50% of their physical capacity activated. For example, the minimum physical
installed memory for both Power E870 and Power E880 is 256 GB, which requires a minimum
of 128 GB activated.
There are two activation types that can be used to accomplish this:
Static memory activations: Memory activations that are exclusive for a single server
Mobile memory activations: Memory activations that can be moved from server to server
in a power enterprise pool.
Both types of memory activations can co-reside in the same system, as long as at least 25%
of the memory activations are static. This leads to a maximum of 75% of the memory
activations as mobile.
System Node 1
1 17 2 21 3 25 4 29
System Node 2
5 18 6 22 7 26 8 30
System Node 3
9 19 10 23 11 27 12 31
System Node 4
13 20 14 24 15 28 16 32
Notes:

In Figure 2-16, there is an example of the minimum required activations for a system with
1 TB of installed memory.
Figure 2-16 Example of the minimum required activations for a system with 1 TB of installed memory
The granularity for static memory activation is 1 GB, while for mobile memory activation, the
granularity is 100 GB. In Table 2-9, there is a list of the feature codes that can be used to
achieve the desired number of activations:
Table 2-9 Static and mobile memory activation feature codes
Static memory activations can be converted to mobile memory activations after system
installation. In order to enable mobile memory activations, the systems must be part of a
power enterprise pool and have feature code #EB35 configured. For more information about
power enterprise pools, see 2.4.2, “Power enterprise pools and mobile capacity on demand
(Mobile CoD)” on page 68.
2.3.4 Memory throughput
The peak memory and I/O bandwidths per system node have increased over 300% compared
to previous POWER7 servers, providing the next generation of data intensive applications
with a platform capable of handling the needed amount of data.
DDR4 256 GB CDIMMs have similar performance as 128 GB CDIMMs. They both operate
1600 MHz and have the same memory bandwidth considerations as with the 512 GB memory
feature.
Power E870
Table 2-10 on page 62 shows the maximum bandwidth estimates for a single core on the
Power E870 system.
Feature code Feature description Amount of memory Type of activation
EMA5 1 GB Memory activation 1 GB Static
EMA6 100 GB Memory activation 100 GB Static
EMA7 100 GB Mobile memory activation 100 GB Mobile
1024 GB
512 GB
768 GB
256 GB
Installed memory
Minimum active
Minimum Static active
Maximum Mobile active
50% of total memory
75% of installed memory
25% of installed memory
Without Power Enterprise Pools
With Power Enterprise Pools

Table 2-10 Power E870 single core bandwidth estimates
The bandwidth figures for the caches are calculated as follows:
L1 cache: In one clock cycle, two 16-byte load operations and one 16-byte store operation
can be accomplished. The value varies depending on the clock of the core and the formula
is as follows:
– 4.024 GHz Core: (2 * 16 B + 1 * 16 B) * 4.024 GHz = 193.15 GBps
L2 cache: In one clock cycle, one 32-byte load operation and one 16-byte store operation
is as follows:
is as follows:
For each system node of a Power E870 populated with four processors and all its memory
CDIMMs filled, the overall bandwidths are shown in Table 2-11.
Table 2-11 Power E870 system node bandwidth estimates
PCIe Interconnect: Each POWER8 processor has 32 PCIe lanes running at 7.877Gbps
full-duplex. The bandwidth formula is calculated as follows:
32 lanes * 4 processors * 7.877 Gbps * 2 = 252.064 GBps
Single core Power E870 Power E870
1 core @ 4.024 GHz 1 core @ 4.190 GHz
L1 (data) cache 193.15 GBps 201.12 GBps
L2 cache 193.15 GBps 201.12 GBps
System node bandwidths Power E870 Power E870
32 cores @ 4.024 GHz 40 cores @ 4.190 GHz
L1 (data) cache 6,181 GBps 8,045 GBps
L2 cache 6,181 GBps 8,045 GBps
Total Memory 922 GBps 922 GBps
PCIe Interconnect 252.064 GBps 252.064 GBps
Intra-node buses
(two system nodes)
922 GBps 922 GBps
Rounding: The bandwidths listed here may appear slightly differently in other materials
due to rounding of some figures.

For the entire Power E870 system populated with two system nodes, the overall bandwidths
are shown in Table 2-12.
Table 2-12 Power E870 total bandwidth estimates
Power E880
Table 2-13 shows the maximum bandwidth estimates for a single core on the Power E880
system.
Table 2-13 Power E880 single core bandwidth estimates
The bandwidth figures for the caches are calculated as follows:
L1 cache: In one clock cycle, two 16-byte load operations and one 16-byte store operation
is as follows:
is as follows:
is as follows:
Total bandwidths Power E870 Power E870
Total Memory 1,844 GBps 1,844 GBps
Inter-node buses (two
system nodes)
307 GBps 307 GBps
Intra-node buses (two
system nodes)
1,844 GBps 1,844 GBps
Single core Power E880 Power E880
1 core @ 4.024 GHz 1 core @ 4.350 GHz
L1 (data) cache 193.15 GBps 208.80 GBps

For each system node of a Power E880 populated with four processors and all its memory
CDIMMs filled, the overall bandwidths are shown in Table 2-14.
Table 2-14 Power E880 system node bandwidth estimates
PCIe Interconnect: Each POWER8 processor has 32 PCIe lanes running at 7.877 Gbps
full-duplex. The bandwidth formula is calculated as follows:
32 lanes * 4 processors * 7.877Gbps * 2 = 252.064 GBps
For the entire Power E880 system populated with four system nodes, the overall bandwidths
are shown in Table 2-15.
Table 2-15 Power E880 total bandwidth estimates
2.3.5 Active Memory Mirroring
The Power E870 and Power E880 systems have the ability to provide mirroring of the
hypervisor code across multiple memory CDIMMs. If a CDIMM that contains the hypervisor
code develops an uncorrectable error, its mirrored partner will enable the system to continue
to operate uninterrupted.
System node bandwidths Power E880 Power E880
Total Memory 922 GBps 922 GBps
Intra-node buses
(two system nodes)
922 GBps 922 GBps
Rounding: The bandwidths listed here may appear slightly differently in other materials
due to rounding of some figures.
Total bandwidths Power E880 Power E880
Total Memory 3,688 GBps 3,688 GBps
Inter-node buses (four
system nodes)
307 GBps 307 GBps
Intra-node buses (four
system nodes)
3,688 GBps 3,688 GBps

Active Memory Mirroring (AMM) is included with all Power E870 and Power E880 systems at
no additional charge. It can be enabled, disabled, or re-enabled depending on the user’s
requirements.
The hypervisor code logical memory blocks will be mirrored on distinct CDIMMs to allow for
more usable memory. There is no specific CDIMM that hosts the hypervisor memory blocks
so the mirroring is done at the logical memory block level, not at the CDIMM level. To enable
the AMM feature it is mandatory that the server has enough free memory to accommodate
the mirrored memory blocks.
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
It is possible to check whether the Active Memory Mirroring option is enabled and change its
current status through HMC, under the Advanced Tab on the CEC Properties panel
(Figure 2-17).
Figure 2-17 CEC Properties panel on an HMC
After a failure on one of the CDIMMs containing hypervisor data occurs, all the server
operations remain active and flexible service processor (FSP) will isolate the failing CDIMMs.
Systems stay in the partially mirrored state until the failing CDIMM is replaced.
There are components that are not mirrored because they are not vital to the regular server
operations and require a larger amount of memory to accommodate its data:
Advanced Memory Sharing Pool
Memory used to hold the contents of platform dumps

With AMM, uncorrectable errors in data that are owned by a partition or application are
handled by the existing Special Uncorrectable Error handling methods in the hardware,
firmware, and operating system.
2.3.6 Memory Error Correction and Recovery
The memory has error detection and correction circuitry is designed such that the failure of
any one specific memory module within an ECC word can be corrected without any other
fault.
In addition, a spare DRAM per rank on each memory port provides for dynamic DRAM device
replacement during runtime operation. Also, dynamic lane sparing on the DMI link allows for
repair of a faulty data lane.
Other memory protection features include retry capabilities for certain faults detected at both
the memory controller and the memory buffer.
Memory is also periodically scrubbed to allow for soft errors to be corrected and for solid
single-cell errors reported to the hypervisor, which supports operating system deallocation of
a page associated with a hard single-cell fault.
For more details on Memory RAS, see 4.3.10, “Memory protection” on page 168.
2.3.7 Special Uncorrectable Error handling
Special Uncorrectable Error (SUE) handling prevents an uncorrectable error in memory or
cache from immediately causing the system to terminate. Rather, the system tags the data
and determines whether it will 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 freeze of the I/O adapters controlled by an I/O hub controller if data is to
be transferred to an I/O device.
2.4 Capacity on Demand
Several types of Capacity on Demand (CoD) offerings are optionally available on the
Power 870 and Power E880 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)
Power E870 and Power E880 systems include a number of active processor cores and
memory units. They can also include inactive processor cores and memory units. Active
processor cores or memory units are processor cores or memory units that are already
available for use on your server when it comes from the manufacturer. Inactive processor
cores or memory units are processor cores or memory units that are included with your
server, but not available for use until you activate them. Inactive processor cores and memory
units can be permanently activated by purchasing an activation feature called Capacity
Upgrade on Demand (CUoD) and entering the provided activation code on your server.
Partition data: 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

With the CUoD offering, you can purchase additional static processor or memory capacity
and dynamically activate them when needed, without requiring you to restart your server or
interrupt your business. All the static processor or memory activations are restricted to a
single server.
Capacity Upgrade on Demand can have several applications to allow for a more flexible
environment. One of its benefits is to allow for a given company to reduce the initial
investment on a system. Traditional projects using other technologies require that the a
system is acquired with all the resources available to support the whole lifecycle of the
project. This might incur in costs that would only be necessary on later stages of the project,
usually with impacts on software licensing costs and software maintenance.
By using Capacity Upgrade on Demand a company could start with a system with enough
installed resources to support the whole project lifecycle but only with enough active
resources necessary for the initial project phases. More resources could be added along with
the project, adjusting the hardware platform with the project needs. This would allow for a
company to reduce the initial investment in hardware and only acquire software licenses that
are needed on each project phase, reducing the Total Cost of Ownership and Total Cost of
Acquisition of the solution. Figure 2-18 shows a comparison between two scenarios: a fully
activated system versus a system with CUoD resources being activated along with the project
timeline.
Figure 2-18 Active cores scenarios comparison during a project lifecycle
0
10
20
30
40
50
60
70
80
Without CuOD
With CuOD
Time
Cores
Projected
Core Activations

Table 2-16 lists the static processor activation features that are available for the Power E870
and Power E880.
Table 2-16 Power Systems CUoD static processor activation features
Table 2-17 lists the static memory activation features that are available for the Power E870
and Power E880.
Table 2-17 Power Systems CUoD static memory activation features
2.4.2 Power enterprise pools and mobile capacity on demand (Mobile CoD)
While static activations are valid for a single system, some customers could benefit from
moving processor and memory activations among different servers due to workload
rebalance or disaster recovery.
IBM power enterprise pools, is a technology for dynamically sharing processor and memory
activations among a group (or pool) of IBM Power Systems servers. Using Mobile Capacity on
Demand (CoD) activation codes, the systems administrator can perform tasks without
contacting IBM.
Two types of power enterprise pools are available:
Power 770 (9117-MMD) and Power E870 (9119-MME) class systems
Power 780 (9117-MHD), Power 795 (9119-FHB), and Power E880 (9119-MHE) class
systems
Each pool type can support systems with different clock speeds or processor generations.
The basic rules for the Mobile Capacity on Demand follows:
The Power 770 and Power 780 systems require a minimum of four static processor
activations.
The Power 870 and Power 880 require a minimum of eight static processor activations.
The Power 795 requires a minimum of 24 static processor activations or 25% of the
installed processor capacity whichever is bigger.
For all systems, 25% of the installed processor capacity must have static activations.
All the systems in a pool must be managed by the same HMC or by the same pair of
redundant HMCs. If redundant HMCs are used, the HMCs must be connected to a network so
that they can communicate with each other. The HMCs must have at least 2 GB of memory.
An HMC can manage multiple power enterprise pools and can also manage systems that are
not part of a power enterprise pool. Systems can belong to only one power enterprise pool at
System Processor feature Processor core static activation
feature
Power E870 #EPBA (4.02 GHz Processor Card) EPBJ
Power E870 #EPBC (4.19 GHz Processor Card) EPBL
Power E880 #EPBB (4.35 GHz Processor Card) EPBK
Power E880 #EPBD (4.02 GHz Processor Card) EPBM
System Description Feature code
Power E870, Power E880 Activation of 1 GB DDR3 POWER8 memory EMA5
Power E870, Power E880 Activation of 100 GB DDR3 POWER8 memory EMA6

a time. Powering down an HMC does not limit the assigned resources of participating
systems in a pool but does limit the ability to perform pool change operations.
After a power enterprise pool is created, the HMC can be used to perform the following
functions:
Mobile CoD processor and memory resources can be assigned to systems with inactive
resources. Mobile CoD resources remain on the system to which they are assigned until
they are removed from the system.
New systems can be added to the pool and existing systems can be removed from the
pool.
New resources can be added to the pool or existing resources can be removed from the
pool.
Pool information can be viewed, including pool resource assignments, compliance, and
history logs.
In order for the Mobile activation features to be configured, it is necessary that a power
enterprise pool is registered with IBM as well as the systems that are going to be included as
members of the pool. Also, it is necessary that the systems have the feature code #EB35 for
mobile enablement configured, and the required contracts must be in place.
Table 2-18 lists the mobile processor activation features that are available for the Power E870
and Power E880.
Table 2-18 Mobile processor activation features
Table 2-19 lists the mobile memory activation features that are available for the Power E870
and Power E880.
Table 2-19 Mobile memory activation features
For more information about power enterprise pools, see the Redpaper publication, Power
Enterprise Pools on IBM Power Systems, REDP5101:
http://www.redbooks.ibm.com/Redbooks.nsf/RedbookAbstracts/redp5101.html?Open
2.4.3 Elastic Capacity on Demand (Elastic CoD)
With the Elastic CoD offering, you can 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. Elastic CoD was previously called On/Off CoD. When you
order an Elastic CoD feature, you receive an enablement code that allows a system operator
System Description CUoD mobile processor core
activation feature
Power E870 1-Core Mobile activation #EP2S
Power E880 1-Core Mobile activation #EP2T
System Description Feature code
Power E870, Power E880 100 GB Mobile memory activation #EMA7
Note: Some web sites or documents still refer to Elastic Capacity on Demand as On/Off
Capacity on Demand.

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.
Charges are based on usage reporting that is collected monthly. Processors and memory
may 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.
New to both Power E870 and Power E880 are 90-day temporary Elastic CoD processor and
memory enablement features. These features enable a system to temporarily activate all
inactive processor and memory CoD resources for a maximum of 90 days before ordering
another temporary elastic enablement. A feature code is required.
Before using temporary capacity on your server, you must enable your server. To enable, an
enablement feature (MES only) must be ordered and the required contracts must be in place.
If a Power E870 or Power E880 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 Elastic 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 E870 and Power E880 are described in 1.4, “System features” on page 12 and 1.4.7,
“Memory features” on page 20.
The Elastic CoD process consists of three steps, enablement, activation, and billing.
Enablement
Before requesting temporary capacity on a server, you must enable it for Elastic 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 reach the limit of 999 memory days, place an order for another memory
enablement code to reset the number of allowable days you can request back to 999.
Activation requests
When Elastic CoD temporary capacity is needed, use the HMC menu for On/Off CoD.
Specify how many inactive processors or gigabytes of memory are required to be
temporarily activated for some number of days. You are billed for the days requested,
whether the capacity is assigned to partitions or remains 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.

Billing
The contract, signed by the client before receiving the enablement code, requires the
Elastic 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 about registration, enablement, and usage of Elastic CoD, visit the
following location:
http://www.ibm.com/systems/power/hardware/cod
2.4.4 Utility Capacity on Demand (Utility CoD)
Utility CoD automatically provides additional processor performance on a temporary basis
within the shared processor pool.
With Utility CoD, you can 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 that are 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 E870 or Power E880 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 the
following location:
http://www.ibm.com/systems/support/planning/capacity/index.html
2.4.5 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 E870 or Power E880 system with inactive CoD
resources.
A standard request activates two processors or 64 GB of memory (or 8 processor cores and
64 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 E870 or Power E880 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 the following location:
https://www-912.ibm.com/tcod_reg.nsf/TrialCod?OpenForm
2.4.6 Software licensing and CoD
For software licensing considerations with the various CoD offerings, see the most recent
revision of the Power Systems Capacity on Demand User’s Guide:
http://www.ibm.com/systems/power/hardware/cod
2.5 System bus
This section provides additional information related to the internal buses.

2.5.1 PCI Express Gen3
The internal I/O subsystem on the Power E870 and Power E880 is connected to the PCIe
Controllers on a POWER8 processor in the system. Each POWER8 processor module has
two buses that have 16 PCIe lanes each (for a total of 32 PCIe lanes) running at 7.877 Gbps
full-duplex and provides 31.508 GBps of I/O connectivity to the PCIe slots. A diagram with the
connections can be seen in Figure 2-19.
Figure 2-19 System nodes PCIe slots directly attached to PCIe controllers on POWER8 chips
Besides the slots directly attached to the processors PCI Gen3 controllers, the systems also
allow for additional PCIe adapters on external PCIe Expansion Drawers and disks on external
drawers connected through PCIe SAS adapters.
Figure 2-20 on page 74 shows a diagram with the I/O connectivity options available for the
Power E870 and Power E880. The system nodes allow for eight PCIe Gen3 x16 slots.
Additional slots can be added by attaching PCIe Expansion Drawers and SAS disks can be
attached to EXP24S SFF Gen2 Drawers. The EXP24S can be either attached to SAS
adapters on the system nodes or on the PCIe Expansion Drawer.
I/OBackplane
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
P8 (0)
PHB 1
PHB 0
SMP
X Bus
SMP
ABus
Memory
Controller 0
Memory
Controller 1
P8 (2)
PHB 1
PHB 0
SMP
X Bus
SMP
ABus
Memory
Controller 0
Memory
Controller 1
P8 (3)
PHB 1
PHB 0
SMP
X Bus
SMP
ABus
Memory
Controller 0
Memory
Controller 1
P8 (1)
Memory
Controller 0
Memory
Controller 1
PHB 1
PHB 0
SMP
X BusSMP
ABus

For a list of adapters and their supported slots, see 2.7, “PCI adapters” on page 77.
I
Figure 2-20 I/O connectivity options available for Power E870 and Power E880
For more information about PCIe Expansion Drawers, see 2.9.1, “PCIe Gen3 I/O expansion
drawer” on page 85.
2.5.2 Service Processor Bus
The redundant service processor bus connectors are located on the rear of the control unit
and the system nodes. All of the service processor (SP) communication between the control
unit and the system nodes flows though these cables.
Unlike the previous generations where a given pair of enclosures would host the service
processors, on Power E870 and Power E880 as a standard, redundant service processor
cards are installed on the control unit as well as redundant clock cards.
The cables used to provide communications between the control units and system nodes
depend on the amount of system nodes installed. When a system node is added, a new set of
cables is also added.
The cables necessary for each system node have been grouped under a single feature code,
allowing for an easier configuration. Each cable set includes a pair of FSP cables, a pair of
Disk support: There is no support for disks directly installed on the system nodes and
PCIe Expansion Drawers. If directly attached SAS disk are required, they need to be
installed in a SAS disk drawer and connected to a supported SAS controller in one of the
PCIe slots.
EXP24S SFF Gen2 Drawer
Disks
PCIe Adapters
EXP24S SFF Gen2 Drawer
System Node
PCIe Gen3 x16
PCIe Adapter for
Expansion Drawer
PCIe Gen3 x16
PCIe Gen3 x16
PCIe Gen3 x16
PCIe Gen3 x16
PCIe Gen3 x16
SAS Adapter
CPU
1
CPU
0
CPU
2
CPU
3

clock cables, and when applicable SMP cables and UPIC cables. Table 2-20 shows a list of
the feature codes available.
Table 2-20 Features for cable sets
Cable sets feature codes are incremental and depend on the number of installed drawers as
follows:
1 system node: #ECCA
2 system nodes: #ECCA and #ECCB
3 system nodes: #ECCA, #ECCB, and #ECCC
4 system nodes: #ECCA, #ECCB, #ECCC, and #ECCD
The system connection topology is shown in 2.1, “Logical diagrams” on page 38.
2.6 Internal I/O subsystem
The internal I/O subsystem resides on the I/O planar, which supports eight PCIe Gen3 x16
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. For more information about RAS on the I/O buses, see 4.3.11, “I/O subsystem
availability and Enhanced Error Handling” on page 169.
Table 2-21 lists the slot configuration of Power E870 and Power E880 system nodes.
Table 2-21 Slot configuration and capabilities
Feature code Description
ECCA System node to system control unit cable set for drawer 1
ECCB System node to system control unit cable set for drawer 2
ECCC System node to system control unit cable set for drawer 3
ECCD System node to system control unit cable set for drawer 4
Slot Location code Slot type CAPI capablea
a. At the time of writing, there are no supported CAPI adapters for the Power E870 and E880
system unit.
SRIOV capable
Slot 1 P1-C1 PCIe Gen3 x16 No Yes
Slot 2 P1-C2 PCIe Gen3 x16 Yes Yes

The physical location of the slots can be seen in Figure 2-21.
Figure 2-21 System node top view and PCIe slot location codes
2.6.1 Blind-swap cassettes
The Power E870 and Power E880 use a next generation blind-swap cassette to manage the
installation and removal of PCIe adapters. This mechanism requires an interposer card that
allows the PCIe adapters to plug in vertically to the system, allows more airflow through the
cassette, and allows for faster hot swap procedures. Cassettes can be installed and removed
without removing the system nodes or PCIe expansion drawers from the rack.
2.6.2 System ports
The system nodes do not have integrated ports. All networking and storage for the virtual
machines must be provided via PCIe adapters installed in standard PCIe slots.
I/OBackplane
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
PCIe Gen3 x16 slot
P8 (0)
PHB 1
PHB 0
SMP
X Bus
SMP
ABus
Memory
Controller 0
Memory
Controller 1
P8 (2)
PHB 1
PHB 0
SMP
X Bus
SMP
ABus
Memory
Controller 0
Memory
Controller 1
P8 (3)
PHB 1
PHB 0
SMP
X Bus
SMP
ABus
Memory
Controller 0
Memory
Controller 1
P8 (1)
Memory
Controller 0
Memory
Controller 1
PHB 1
PHB 0
SMP
X Bus
SMP
ABus
P1-C1
P1-C2
P1-C3
P1-C4
P1-C5
P1-C6
P1-C7
P1-C8
Front
Rear
System Node top view

The system control unit has one USB Port dedicated to the DVD drive and 4 Ethernet ports
used for HMC communications. There is no serial port so an HMC is mandatory for system
management. The FSP’s virtual console will be on the HMC.
The location of the USB and HMC Ethernet ports can be seen in Figure 2-22.
Figure 2-22 Physical location of the USB and HMC ports on the system control unit
The connection and usage of the DVD can be seen in detail in 2.8.1, “DVD” on page 84.
2.7 PCI adapters
This section covers the types and functions of the PCI cards supported by Power E870 and
Power E880 systems.
2.7.1 PCI Express (PCIe)
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 labelled as
x1, x2, x8, x12, x16, or x32, where the number is effectively the number of lanes.
The PCIe interfaces supported on this server are PCIe Gen3, capable of 16 GBps simplex
(32 GBps duplex) on a single x16 interface. PCIe Gen3 slots also support previous
generations (Gen2 and Gen1) adapters, which operate at lower speeds, according to the
following rules:
Place x1, x4, x8, and x16 speed adapters in same connector size slots first, before mixing
adapter speed with connector slot size.
Adapters with smaller speeds are allowed in larger sized PCIe connectors but larger
speed adapters are not compatible in smaller connector sizes (i.e. a x16 adapter cannot
go in an x8 PCIe slot connector).
IBM POWER8 processor-based servers can support two different form factors of PCIe
adapters:
PCIe low profile (LP) cards, which are used with system node PCIe slots.
FSP 1 FSP 2
HMC1
HMC2
HMC1
HMC2
USB

PCIe full height and full high cards are used in the PCIe Gen3 I/O expansion drawer
(#EMX0)
Low-profile PCIe adapter cards are supported only in low-profile PCIe slots, and full-height
and full-high cards are supported only in full-high slots.
Before adding or rearranging adapters, use the System Planning Tool to validate the new
adapter configuration. For more information, 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 that is 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 describe the supported adapters and provide tables of orderable
feature numbers. The tables indicate operating system support (AIX, IBM i, and Linux) for
each of the adapters.
2.7.2 LAN adapters
To connect the Power E870 and Power E880 servers to a local area network (LAN), you can
use the LAN adapters that are supported in the PCIe slots of the system. Table 2-22 lists the
available LAN adapters. Information about FCoE adapters can be found in Table 2-26 on
page 82.
Table 2-22 Available LAN adapters
Feature Code CCIN Description Placement OS support
5260 576F PCIe2 LP 4-port 1 GbE Adapter CEC AIX, IBM i,
Linux
5274 5768 PCIe LP 2-Port 1 GbE SX Adapter CEC AIX, IBM i,
Linux
5744 2B44 PCIe2 4-Port 10 GbE&1 GbE SR&RJ45
Adapter
I/O drawer Linux
5767 5767 2-Port 10/100/1000 Base-TX Ethernet
PCI Express Adapter
I/O drawer AIX, IBM i,
Linux
5768 5768 2-Port Gigabit Ethernet-SX PCI Express
Adapter
Linux
5769 5769 10 Gigabit Ethernet-SR PCI Express
Adapter
Linux
5772 576E 10 Gigabit Ethernet-LR PCI Express
Adapter
Linux
5899 576F PCIe2 4-port 1 GbE Adapter I/O drawer AIX, IBM i,
Linux
EC27 EC27 PCIe2 LP 2-Port 10 GbE RoCE SFP+
Adapter
CEC AIX, Linux
EC28 EC27 PCIe2 2-Port 10 GbE RoCE SFP+
Adapter
Linux

EC29 EC29 PCIe2 LP 2-Port 10 GbE RoCE SR
Adapter
CEC AIX, IBM i,
Linux
EC2G PCIe2 LP 2-port 10 GbE SFN6122F
Adapter
CEC Linux
EC2J PCIe2 2-port 10 GbE SFN6122F
Adapter
I/O drawer Linux
EC2M 57BE PCIe3 LP 2-port 10 GbE NIC&RoCE SR
Adapter
CEC AIX, IBM i,
Linux
EC2N PCIe3 2-port 10 GbE NIC&RoCE SR
Adapter
Linux
EC30 EC29 PCIe2 2-Port 10 GbE RoCE SR Adapter I/O drawer AIX, IBM i,
Linux
EC37 57BC PCIe3 LP 2-port 10 GbE NIC&RoCE
SFP+ Copper Adapter
CEC AIX, IBM i,
Linux
EC38 PCIe3 2-port 10 GbE NIC&RoCE SFP+
Copper Adapter
Linux
EC3A 57BD PCIe3 LP 2-Port 40 GbE NIC RoCE
QSFP+ Adapter
CEC AIX, IBM i,
Linux
EC3B 57B6 PCIe3 2-Port 40 GbE NIC RoCE
QSFP+ Adapter
Linux
EN0M 2CC0 PCIe2 4-port(10Gb FCoE & 1 GbE)
LR&RJ45 Adapter
Linux
EN0N 2CC0 PCIe2 LP 4-port(10Gb FCoE & 1 GbE)
LR&RJ45 Adapter
CEC AIX, IBM i,
Linux
EN0S 2CC3 PCIe2 4-Port (10Gb+1 GbE) SR+RJ45
Adapter
Linux
EN0T 2CC3 PCIe2 LP 4-Port (10Gb+1 GbE)
SR+RJ45 Adapter
CEC AIX, IBM i,
Linux
EN0U 2CC3 PCIe2 4-port (10Gb+1 GbE) Copper
SFP+RJ45 Adapter
Linux
EN0V 2CC3 PCIe2 LP 4-port (10Gb+1 GbE) Copper
SFP+RJ45 Adapter
CEC AIX, IBM i,
Linux
EN0W 2CC4 PCIe2 2-port 10/1 GbE BaseT RJ45
Adapter
Linux
EN0X 2CC4 PCIe2 LP 2-port 10/1 GbE BaseT RJ45
Adapter
CEC AIX, IBM i,
Linux
EN15 2CE3 PCIe3 4-port 10 GbE SR Adapter I/O drawer AIX, IBM i,
Linux
EN16 PCIe3 LPX 4-port 10 GbE SR Adapter CEC AIX, IBM i,
Linux
EN17 2CE4 PCIe3 4-port 10 GbE SFP+ Copper
Adapter
AIX, IBM i,
Linux

2.7.3 Graphics accelerator adapters
Table 2-23 lists the available graphics accelerator adapters. An adapter can be configured to
operate in either 8-bit or 24-bit color modes. The adapter supports both analog and digital
monitors.
Table 2-23 Available graphics accelerator adapters
2.7.4 SAS adapters
Table 2-24 lists the SAS adapters that are available for Power E870 and Power E880
systems.
Table 2-24 Available SAS adapters
2.7.5 Fibre Channel adapter
The systems support direct or SAN connection to devices that use Fibre Channel adapters.
Table 2-25 summarizes the available Fibre Channel adapters, which all have LC connectors.
EN18 PCIe3 LPX 4-port 10 GbE SFP+ Copper
Adapter
AIX, IBM i,
Linux
Feature
Code
CCIN Description Placement OS support
5269 5269 PCIe LP POWER GXT145 Graphics
Accelerator
CEC AIX, Linux
EC41 PCIe2 LP 3D Graphics Adapter x1 CEC AIX, Linux
5901 57B3 PCIe Dual-x4 SAS Adapter I/O drawer AIX, Linux
5913 57B5 PCIe2 1.8 GB Cache RAID SAS
Adapter Tri-port 6Gb
I/O drawer AIX, Linux
ESA3 57BB PCIe2 1.8 GB Cache RAID SAS
Adapter Tri-port 6Gb CR
EJ0J 57B4 PCIe3 RAID SAS Adapter Quad-port
6Gb x8
EJ0L 57CE PCIe3 12 GB Cache RAID SAS Adapter
Quad-port 6Gb x8
EJ0M PCIe3 LP RAID SAS ADAPTER CEC AIX, Linux
EJ10 57B4 PCIe3 SAS Tape/DVD Adapter
Quad-port 6Gb x8
EJ11 57B4 PCIe3 LP SAS Tape/DVD Adapter
Quad-port 6Gb x8
CEC AIX, Linux
EJ14 57B1 PCIe3 12 GB Cache RAID PLUS SAS
Adapter Quad-port 6Gb x8
Linux

If you are attaching a device or switch with an SC type fiber connector, then 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
2.7.6 Fibre Channel over Ethernet
Fibre Channel over Ethernet (FCoE) allows for the convergence of Fibre Channel and
Ethernet traffic onto a single adapter and a converged fabric.
Figure 2-23 compares existing Fibre Channel and network connections and FCoE
connections.
Figure 2-23 Comparison between existing Fibre Channel and network connections and FCoE
connections
Feature
Code
5273 577D PCIe LP 8Gb 2-Port Fibre Channel Adapter CEC AIX, IBM i,
Linux
5276 5774 PCIe LP 4 Gb 2-Port Fibre Channel Adapter CEC AIX, IBM i,
Linux
5729 5729 PCIe2 8Gb 4-port Fibre Channel Adapter I/O drawer AIX, Linux
5735 577D Gigabit PCI Express Dual Port Fibre Channel
Adapter
Linux
5774 5774 4 Gigabit PCI Express Dual Port Fibre
Channel Adapter
Linux
EN0A 577F PCIe2 16Gb 2-port Fibre Channel Adapter I/O drawer AIX, Linux
EN0B 577F PCIe2 LP 16Gb 2-port Fibre Channel Adapter CEC AIX, IBM i,
Linux
EN0F 578D PCIe2 LP 8Gb 2-Port Fibre Channel Adapter CEC AIX, Linux
EN0G PCIe2 8Gb 2-Port Fibre Channel Adapter I/O drawer AIX, Linux
EN0Y EN0Y PCIe2 LP 8Gb 4-port Fibre Channel Adapter CEC AIX, IBM i,
Linux
EN12 PCIe2 8Gb 4-port Fibre Channel Adapter I/O drawer AIX, Linux
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

Table 2-26 lists the available FCoE adapters. They are high-performance Converged Network
Adapters (CNAs) using SR optics. Each port can simultaneously provide network interface
card (NIC) traffic and Fibre Channel functions.
Table 2-26 Available FCoE adapters
For more information about FCoE, see An Introduction to Fibre Channel over Ethernet, and
Fibre Channel over Convergence Enhanced Ethernet, REDP-4493.
2.7.7 USB adapters
Each system control unit enclosure can have one slim-line bay that can support one DVD
drive (#EU13). The DVD drive is cabled to a USB PCIe adapter located in either the system
node or in a PCIe Gen3 I/O drawer.
Table 2-27 lists the available USB adapters.
Table 2-27 Available asynchronous and USB adapters
2.7.8 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.
Feature
Code
EN0H 2B93 PCIe2 4-port (10Gb FCoE & 1 GbE) SR&RJ45 I/O drawer AIX, IBM i,
Linux
EN0J 2B93 PCIe2 LP 4-port (10Gb FCoE & 1 GbE)
SR&RJ45
CEC AIX, IBM i,
Linux
EN0K 2CC1 PCIe2 4-port (10Gb FCoE & 1 GbE)
SFP+Copper&RJ45
Linux
EN0L 2CC1 PCIe2 LP 4-port(10Gb FCoE & 1 GbE)
SFP+Copper&RJ45
CEC AIX, IBM i,
Linux
EN0M PCIe3 4-port(10Gb FCoE & 1GbE) LR&RJ45
Adapter
CEC AIX, Linux
EN0N PCIe3 LP 4-port(10Gb FCoE & 1GbE)
LR&RJ45 Adapter
CEC AIX, Linux
Note: Adapters EN0J and EN0L support SR-IOV when minimum firmware and software
levels are met. See 3.4, “Single root I/O virtualization (SR-IOV)” on page 126 for more
information.
Feature
Code
EC45 PCIe2 LP 4-Port USB 3.0 Adapter CEC AIX, IBM i, Linux
EC46 PCIe2 4-Port USB 3.0 Adapter I/O drawer AIX, IBM i, Linux

InfiniBand (IB) is an open set of interconnect standards and specifications. The main IB
specification is published by the InfiniBand Trade Association and is available at the following
location:
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, and also 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.
Table 2-28 lists the available InfiniBand adapters.
Table 2-28 Available InfiniBand adapters
2.7.9 Cryptographic Coprocessor
The Cryptographic Coprocessor cards provide both cryptographic coprocessor and
cryptographic accelerator functions in a single card.
The IBM PCIe Cryptographic Coprocessor adapter highlights the following features:
Integrated Dual processors that operate in parallel for higher reliability
Supports IBM Common Cryptographic Architecture or PKCS#11 standard
Ability to configure adapter as coprocessor or accelerator
Support for smart card applications using Europay, MasterCard, and Visa
Cryptographic key generation and random number generation
PIN processing: generation, verification, translation
Encrypt and Decrypt using AES and DES keys
See the following site for the most recent firmware and software updates:
http://www.ibm.com/security/cryptocards/
Table 2-29 lists the cryptographic adapter that is available for the server.
Table 2-29 Available cryptographic adapters
Feature
Code
5283 58E2 PCIe2 LP 2-Port 4X IB QDR Adapter 40Gb CEC AIX, Linux
5285 58E2 PCIe2 2-Port 4X IB QDR Adapter 40Gb I/O drawer AIX, Linux
Feature
Code
EJ28 PCIe Crypto Coprocessor Gen3 BSC
4765-001
Linux

2.7.10 CAPI adapters
The available CAPI adapters for JAVA and WebSphere acelleration are shown in Table 2-30.
Table 2-30 Available CAPI adapters
2.8 Internal storage
The system nodes for Power E870 and Power E880 do not allow for conventional physical
storage. All storage must be provided externally via I/O expansion drawers or SAN. At the
time of writing, the only external I/O expansion drawer that can be used in Power E870 and
Power E880 is EXP24S, attached via SAS ports to a SAS PCIe adapter installed either in a
system node drawer or in a PCIe expansion slot located in an I/O expansion drawer.
The system control unit has a DVD drive which is connected to an external USB port on the
rear of the unit. In order to use the DVD drive, at least on PCIe USB adapter must be installed
and connected via an USB cable to the DVD drive.
The following NVMe Flash Adapters are available for internal storage
PCIe3 LP 1.6 TB NVMe Flash Adapter (#EC54)
PCIe3 LP 3.2 TB NVMe Flash Adapter (#EC56)
2.8.1 DVD
There is one DVD media bay per system, located on the front of the system control unit. This
media bay allows for a DVD (#EU13) to be installed on the system control unit and it enables
the USB port on the rear of the control unit.
The USB port must be connected via a USB cable to a USB PCIe adapter, installed in one of
the available PCIe slots on the system nodes or I/O expansion drawers. A diagram of the DVD
connection can be seen in Figure 2-24 on page 85.
The basic rules for DVD drives on these system are as follows:
Include a DVD drive with #EC13.
Include the PCI USB adapter #EC45 or #EC46.
Include a USB cable male-male with the proper length. As a suggestion, #EBK4 is a 1.6 m
cable that allows enough length for the adapters on the first two system nodes to be
connected to the USB DVD port.
This architecture allows for a more flexible infrastructure where the DVD drive is completely
independent of another components and can be freely moved between partition.
Feature
code
CCIN Description One
Processor
Two
Processors
OS support
EJ18 PCIe3 CAPI
FlashSystem
Accelerator Adapter
1 1 AIX

It is also important to note that several daily functions where a DVD drive was needed can
now be executed by using other methods such as Virtual Media Repository on the Virtual I/O
Server or the Remote Virtual I/O Server install on HMC.
Figure 2-24 DVD drive physical location on control unit and suggested cabling
2.9 External I/O subsystems
This section describes the PCIe Gen3 I/O expansion drawer that can be attached to the
Power E870 and Power E880.
2.9.1 PCIe Gen3 I/O expansion drawer
The PCIe Gen3 I/O expansion drawer is a 4U high, PCI Gen3-based and rack mountable I/O
drawer. It offers two PCIe Fan Out Modules (#EMXF) each of them providing six PCIe slots.
The physical dimensions of the drawer are 444.5 mm (17.5 in.) wide by 177.8 mm (7.0 in.)
high by 736.6 mm (29.0 in.) deep for use in a 19-inch rack.
A PCIe x16 to Optical CXP converter adapter (#EJ07) and 2.0 m (#ECC6), 10.0 m (#ECC8),
or 20.0 m (#ECC9) CXP 16X Active Optical cables (AOC) connect the system node to a PCIe
Fan Out module in the I/O expansion drawer. One feature #ECC6, one #ECC8, or one #ECC9
ships two AOC cables.
Control Unit – Front View
DVD Drive (#EU13)
Control Unit – Rear View USB Port
PCIe USB
Adapter (#EC45)
USB Cable 1.6m
Male-Male (#EBK4)
System Node
PCIe Expansion
Drawer

Concurrent repair and add/removal of PCIe adapter cards is done by HMC guided menus or
by operating system support utilities.
A blind swap cassette (BSC) is used to house the full high adapters which go into these slots.
The BSC is the same BSC as used with the previous generation server's
#5802/5803/5877/5873 12X attached I/O drawers.
Figure 2-25 shows the back view of the PCIe Gen3 I/O expansion drawer.
Figure 2-25 Rear view of the PCIe Gen3 I/O expansion drawer
2.9.2 PCIe Gen3 I/O expansion drawer optical cabling
I/O drawers are connected to the adapters in the system node with data transfer cables:
2.0 m Optical Cable Pair for PCIe3 Expansion Drawer (#ECC6)
A minimum of one PCIe3 Optical Cable Adapter for PCIe3 Expansion Drawer (#EJ07) is
required to connect to the PCIe3 6-slot Fan Out module in the I/O expansion drawer. The top
port of the fan out module must be cabled to the top port of the #EJ07 port. Likewise, the
bottom two ports must be cabled together.
1. Connect an active optical cable to connector T1 on the PCIe3 optical cable adapter in your
server.
Fan-out Module
6 PCIe G3 Slots
4 x8 & 2 x164 x8 & 2 x16
Fan-out Module
6 PCIe G3 Slots
4 x8 & 2 x164 x8 & 2 x16 Dual Power Cords
PCIe Optical Interface to CEC PCIe Optical Interface to CEC
Cable lengths: Use the 2.0 m cables for intra-rack installations. Use the 10.0 m or 20.0 m
cables for inter-rack installations.

2. Connect the other end of the optical cable to connector T1 on one of the PCIe3 6-slot Fan
Out modules in your expansion drawer.
3. Connect another cable to connector T2 on the PCIe3 optical cable adapter in your server.
4. Connect the other end of the cable to connector T2 on the PCIe3 6-slot Fan Out module in
your expansion drawer.
5. Repeat the four steps above for the other PCIe3 6-slot Fan Out module in the expansion
drawer, if required.
Figure 2-26 shows connector locations for the PCIe Gen3 I/O expansion drawer.
Figure 2-26 Connector locations for the PCIe Gen3 I/O expansion drawer
Drawer connections: Each Fan Out module in a PCIe3 Expansion Drawer can only be
connected to a single PCIe3 Optical Cable Adapter for PCIe3 Expansion Drawer (#EJ07).
However the two Fan Out modules in a single I/O expansion drawer can be connected to
different system nodes in the same server.

Figure 2-27 shows typical optical cable connections.
Figure 2-27 Typical optical cable connection
General rules for the PCI Gen3 I/O expansion drawer configuration
The PCIe3 optical cable adapter can be in any of the PCIe adapter slots in the Power E870
and Power E880 system node. However, we advise that you use the PCIe adapter slot priority
information while selecting slots for installing PCIe3 Optical Cable Adapter (#EJ07).
Table 2-31 shows PCIe adapter slot priorities in the Power E870 and Power E880 system.
Table 2-31 PCIe adapter slot priorities
The following figures show several examples of supported configurations. For simplification
we have not shown every possible combination of the I/O expansion drawer to server
attachments.
Figure 2-28 on page 89 shows an example of a single system node and two PCI Gen3 I/O
expansion drawers.
Feature code Description Slot priorities
EJ07 PCIe3 Optical Cable Adapter for PCIe3 Expansion Drawer 1, 7, 3, 5, 2, 8, 4, 6
T1
T2
Optical cable adapter
T1
T2
Optical cable adapter
T1
T2
Fanout module
T1
T2
Fanout module
Optical cables
Optical cables
Server I/O Expansion Drawer

Figure 2-28 Example of a single system node and two I/O drawers
System node
I/O Drawer
I/O Drawer

Figure 2-29 shows an example of two system nodes and two PCI Gen3 I/O expansion
drawers.
Figure 2-29 Example of two system nodes and two I/O drawers
System node
System node
I/O Drawer
I/O Drawer

Figure 2-30 shows an example of two system nodes and four PCI Gen3 I/O expansion
drawers.
Figure 2-30 Example of two system nodes and four I/O drawers
2.9.3 PCIe Gen3 I/O expansion drawer SPCN cabling
There is no system power control network (SPCN) used to control and monitor the status of
power and cooling within the I/O drawer. SPCN capabilities are integrated in the optical
cables.
System node
System node
I/O Drawer
I/O Drawer
I/O Drawer
I/O Drawer

2.10 External disk subsystems
This section describes the following external disk subsystems that can be attached to the
Power E870 and Power E880 system:
EXP24S SFF Gen2-bay Drawer for high-density storage (#5887)
IBM System Storage®
2.10.1 EXP24S SFF Gen2-bay Drawer
The EXP24S SFF Gen2-bay Drawer (#5887) is an expansion drawer with twenty-four 2.5-inch
form-factor SAS bays. The EXP24S supports up to 24 hot-swap SFF-2 SAS hard disk drives
(HDDs) or solid-state drives (SSDs). It uses only 2 EIA of space in a 19-inch rack. The
EXP24S includes redundant ac power supplies and uses two power cords.
To maximize configuration flexibility and space utilization, the system node of Power E870
and Power E880 system does not have integrated SAS bays or integrated SAS controllers.
PCIe SAS adapters and the EXP24S can be used to provide direct access storage.
To further reduce possible single points of failure, EXP24S configuration rules consistent with
previous Power Systems are used. IBM i configurations require the drives to be protected
(RAID or mirroring). Protecting the drives is highly advised, but not required for other
operating systems. All Power operating system environments that are using SAS adapters
with write cache require the cache to be protected by using pairs of adapters.
Note: The EXP30 Ultra SSD Drawer (#EDR1 or #5888), the EXP12S SAS Disk Drawer
(#5886), and the EXP24 SCSI Disk Drawer (#5786) are not supported on the Power E870
and Power E880 server.

With AIX, Linux, and VIOS, you can order the EXP24S with four sets of six bays, two sets of
12 bays, or one set of 24 bays (mode 4, 2, or 1). With IBM i, you can order the EXP24S as
one set of 24 bays (mode 1). Figure 2-31 shows the front of the unit and the groups of disks
on each mode.
Figure 2-31 EXP24S front view with location codes and disk groups depending on its mode of
operation
Mode setting is done by IBM manufacturing. If you need to change the mode after installation,
ask your IBM support representative to refer to:
http://w3-03.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/PRS5121
The stickers indicate whether the enclosure is set to mode 1, mode 2, or mode 4. They are
attached to the lower-left shelf of the chassis (A) and the center support between the
enclosure services manager modules (B).
GROUP 1
GROUP 1 GROUP 2
GROUP 1 GROUP 2 GROUP 3 GROUP 4
MODE 1
MODE 2
MODE 4
1 x 24 disks
2 x 12 disks
4 x 6 disks
IBM EXP24S front view

Figure 2-32 shows the mode stickers.
Figure 2-32 Mode sticker locations at the rear of the 5887 disk drive enclosure
The EXP24S SAS ports are attached to a SAS PCIe adapter or pair of adapters using SAS
YO or X cables. Cable length varies depending on the feature code, and proper length should
be calculated considering routing for proper airflow and ease of handling. A diagram of both
types of SAS cables can be seen in Figure 2-33.
Figure 2-33 Diagram of SAS cables types X and YO

The following SAS adapters support the EXP24S:
PCIe 380MB Cache Dual-x4 3Gb SAS RAID Adapter (#5805)
PCIe Dual-x4 SAS Adapter (#5901)
PCIe2 1.8GB Cache RAID SAS Adapter Tri-port 6Gb (#5913)
PCIe2 1.8GB Cache RAID SAS Adapter Tri-port 6Gb CR (#ESA3)
PCIe3 RAID SAS Adapter Quad-port 6Gb x8 (#EJ0J)
PCIe3 12GB Cache RAID SAS Adapter Quad-port 6Gb x8 (#EJ0L)
PCIe3 12GB Cache RAID Plus SAS Adapter Quad-port 6Gb x8 (#EJ14)
PCIe3 LP RAID SAS ADAPTER (#EJ0M)
The EXP24S drawer can support up to 24 SAS SFF Gen-2 disks. Table 2-32 lists the
available disk options.
Table 2-32 Available disks for the EXP24S
Feature
Code
CCIN Description OS support
ES0G 775 GB SFF-2 SSD for AIX/Linux AIX, Linux
ES0H 775 GB SFF-2 SSD for IBM i IBM i
ES0Q 387 GB SFF-2 4 K SSD for AIX/Linux AIX, Linux
ES0R 387 GB SFF-2 4 K SSD for IBM i IBM i
ES0S 775 GB SFF-2 4 K SSD for AIX/Linux AIX, Linux
ES0T 775 GB SFF-2 4 K SSD for IBM i IBM i
ES19 387 GB SFF-2 SSD for AIX/Linux AIX, Linux
ES1A 387 GB SFF-2 SSD for IBM i IBM i
ES2C 387 GB SFF-2 SSD for AIX/Linux AIX, Linux
ES2D 387 GB SFF-2 SSD for IBM i IBM i
ES78 387 GB SFF-2 SSD 5xx eMLC4 for AIX/Linux AIX, Linux
ES79 387 GB SFF-2 SSD 5xx eMLC4 for IBM i IBM i
ES7E 775 GB SFF-2 SSD 5xx eMLC4 for AIX/Linux AIX, Linux
ES7F 775 GB SFF-2 SSD 5xx eMLC4 for IBM i IBM i
ES80 1.9 TB Read Intensive SAS 4k SFF-2 SSD for AIX/Linux AIX, Linux
ES81 1.9 TB Read Intensive SAS 4k SFF-2 SSD for IBM i IBM i
ES85 387 GB SFF-2 SSD 4k eMLC4 for AIX/Linux AIX, Linux
ES86 387 GB SFF-2 SSD 4k eMLC4 for IBM i IBM i
ES8C 775 GB SFF-2 SSD 4k eMLC4 for AIX/Linux AIX, Linux
ES8D 775 GB SFF-2 SSD 4k eMLC4 for IBM i IBM i
ES8F 1.55 TB SFF-2 SSD 4k eMLC4 for AIX/Linux AIX, Linux
ES8G 1.55 TB SFF-2 SSD 4k eMLC4 for IBM i IBM i

There are six SAS connectors on the rear of the EXP24S drawer to which two SAS adapters
or controllers are attached. They are labeled T1, T2, and T3; there are two T1, two T2, and
two T3 connectors. While configuring the drawer, special configuration feature codes will
indicate for the plant the mode of operation in which the disks and ports will be split:
In mode 1, two or four of the six ports are used. Two T2 ports are used for a single SAS
adapter, and two T2 and two T3 ports are used with a paired set of two adapters or dual
adapters configuration.
In mode 2 or mode 4, four ports are used, two T2 and two T3, to access all SAS bays.
1738 19B4 856 GB 10 K RPM SAS SFF-2 Disk Drive (IBM i) IBM i
1752 900 GB 10 K RPM SAS SFF-2 Disk Drive (AIX/Linux) AIX, Linux
ESD2 59CD 1.1 TB 10 K RPM SAS SFF-2 Disk Drive (IBM i) IBM i
ESD3 0 1.2 TB 10 K RPM SAS SFF-2 Disk Drive (AIX/Linux) AIX, Linux
ESDN 571 GB 15 K RPM SAS SFF-2 Disk Drive - 528 Block (IBM i) IBM i
ESDP 600 GB 15 K RPM SAS SFF-2 Disk Drive - 5xx Block
(AIX/Linux)
AIX, Linux
ESEU 59D2 571 GB 10 K RPM SAS SFF-2 Disk Drive 4 K Block - 4224 IBM i
ESEV 0 600 GB 10 K RPM SAS SFF-2 Disk Drive 4 K Block - 4096 AIX, Linux
ESEY 0 283 GB 15 K RPM SAS SFF-2 4 K Block - 4224 Disk Drive IBM i
ESEZ 0 300 GB 15 K RPM SAS SFF-2 4 K Block - 4096 Disk Drive AIX, Linux
ESF2 58DA 1.1 TB 10 K RPM SAS SFF-2 Disk Drive 4 K Block - 4224 IBM i
ESF3 0 1.2 TB 10 K RPM SAS SFF-2 Disk Drive 4 K Block - 4096 AIX, Linux
ESFN 0 571 GB 15 K RPM SAS SFF-2 4 K Block - 4224 Disk Drive IBM i
ESFP 0 600 GB 15 K RPM SAS SFF-2 4 K Block - 4096 Disk Drive AIX, Linux
ESFS 0 1.7 TB 10 K RPM SAS SFF-2 Disk Drive 4 K Block - 4224 IBM i
ESFT 0 1.8 TB 10 K RPM SAS SFF-2 Disk Drive 4 K Block - 4096 AIX, Linux
Feature
Code
CCIN Description OS support

Figure 2-34 shows the rear connectors of the EXP24S drawer, how they relate with the
modes of operation, and disk grouping.
Figure 2-34 Rear view of EXP24S with the 3 modes of operation and the disks assigned to each port
An EXP24S drawer in mode 4 can be attached to two or four SAS controllers and provide
high configuration flexibility. An EXP24S in mode 2 has similar flexibility. Up to 24 HDDs can
be supported with any of the supported SAS adapters or controllers.
The most common configurations for EXP24S with Power Systems are detailed in 2.10.2,
“EXP24S common usage scenarios” on page 97. Not all possible scenarios are included. For
more information about SAS cabling and cabling configurations, search “Planning for
serial-attached SCSI cables” in the IBM Knowledge Center, which can be accessed at:
http://www-01.ibm.com/support/knowledgecenter/9119-MME/p8had/p8had_sascabling5887.
htm?cp=9119-MME
2.10.2 EXP24S common usage scenarios
The EXP24S drawer is very versatile in the ways that it can be attached to Power Systems.
This section describes the most common usage scenarios for EXP24S and Virtual I/O
Servers, using standard PCIe SAS adapters #5901.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
T3
T1
T2
T3
T1
T2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
T1
T3
T1
T2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
T3
T1
T2
T3
T1
T2
T3
T2
MODE 1 MODE 2 MODE 4
SharedPorts
XCablesRequired
ESM1 ESM2 ESM1 ESM2 ESM1 ESM2
IBM EXP24S rear view
Note: Not all possible scenarios are included. Refer to the “Planning for serial-attached
SCSI cables” guide in the IBM Knowledge Center to see more supported scenarios.

Scenario 1: Basic non-redundant connection
This scenario assumes a single Virtual I/O Server with a single PCIe SAS adapter #5901 and
an EXP24S set on mode 1, allowing for up to 24 disks to be attached to the server.
Figure 2-35 shows the connection diagram and components of the solution.
Figure 2-35 Scenario 1 - Basic non-redundant connection
For this scenario, these are the required feature codes:
One EXP24S drawer #5887 with indicator feature #9359 (mode 1 with single #5901)
One PCIe SAS adapter #5901
One SAS YO cable 3 Gbps with proper length
EXP24S – Mode 1 with 1 x #5901 (config indicator #9359)
SAS adapter #5901
SAS YO Cable
Virtual
I/O
Server
GROUP 1

Scenario 2: Basic redundant connection
This scenario assumes a single Virtual I/O Server with two PCIe SAS adapters #5901 and an
EXP24S set on mode 1, allowing for up to 24 disks to be attached to the server. Figure 2-36
shows the connection diagram and components of the solution.
Figure 2-36 Scenario 2- Basic redundant connection
One EXP24S drawer #5887 with indicator feature #9360 (mode 1 with dual #5901)
Two PCIe SAS adapter #5901
Two SAS YO cables 3 Gbps with proper length
The ports used on the SAS adapters must be the same for both adapters of the pair. There is
no SSD support on this scenario.
#5901
SAS YO Cable
Virtual
I/O
Server
#5901
#5901
Use
same
port on
the pair
GROUP 1

Scenario 3: Dual Virtual I/O Servers sharing a single EXP24S
This scenario assumes a dual Virtual I/O Server with two PCIe SAS adapters #5901 each
and an EXP24S set on mode 2, allowing for up to 12 disks to be attached to the each Virtual
I/O Server. Figure 2-37 shows the connection diagram and components of the solution.
Figure 2-37 Dual Virtual I/O Servers sharing a single EXP24S
One EXP24S drawer #5887 with indicator feature #9366 (mode 2 with quad #5901)
Four PCIe SAS adapter #5901
Two SAS X cables 3 Gbps with proper length
The ports used on the SAS adapters must be the same for both adapters of the pair. There is
no SSD support on this scenario.
#5901
2 x SAS X Cables
Virtual
I/O
Server
#1
#5901
#5901
Use
same
port on
the pair
Virtual
I/O
Server
#2
#5901
#5901 Use
same
port on
the pair
GROUP 1 GROUP 2

Scenario 4: Dual Virtual I/O Servers sharing two EXP24S
This scenario assumes a dual Virtual I/O Server with two PCIe SAS adapters #5901 each
and two EXP24S set on mode 2, allowing for up to 24 disks to be attached to the each Virtual
I/O Server (2 per drawer). If compared to scenario 3, this scenario has the benefit to allow
disks from different EXP24S drawers to be mirrored, allowing for hot maintenance of the
whole EXP24S drawers if all data is properly mirrored. Figure 2-38 shows the connection
diagram and components of the solution.
Figure 2-38 Dual Virtual I/O Servers sharing two EXP24S
Two EXP24S drawers #5887 with indicator feature #9361 (mode 2 with dual #5901)
Four PCIe SAS adapter #5901
Four SAS YO cables 3 Gbps with proper length.
There is no SSD support on this scenario.
#5901
2 x SAS YO Cables
Virtual
I/O
Server
#1
#5901
Virtual
I/O
Server
#2
#5901
#5901
2 x SAS YO Cables
GROUP 1 GROUP 2
GROUP 1 GROUP 2

Scenario 5: Four Virtual I/O Servers sharing two EXP24S
This scenario assumes four Virtual I/O Servers with two PCIe SAS adapters #5901 each and
two EXP24S set on mode 4, allowing for up to 12 disks to be attached to the each Virtual I/O
Server (6 per drawer). This scenario has the benefit to allow disks from different EXP24S
drawers to be mirrored, allowing for hot maintenance of the whole EXP24S drawers if all data
is properly mirrored. Figure 2-39 shows the connection diagram and components of the
solution.
Figure 2-39 Four Virtual I/O Servers sharing two EXP24S
Two EXP24S drawers #5887 with indicator feature #9365 (mode 4 with four #5901)
Eight PCIe SAS adapter #5901
Four SAS X cables 3 Gbps with proper length.
There is no SSD support on this scenario.
Other scenarios
For direct connection to logical partitions, different adapters, and cables, see “5887 disk drive
enclosure” in the IBM Knowledge Center:
http://www-01.ibm.com/support/knowledgecenter/POWER8/p8hdx/POWER8welcome.htm
2 x SAS X Cables
Virtual I/O
Server #1 #5901
2 x SAS X Cables
#5901
Virtual I/O
Server #2 #5901#5901
Virtual I/O
Server #3 #5901#5901
Virtual I/O
Server #4 #5901#5901
G1 G2 G3 G4
G1 G2 G3 G4

2.10.3 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 to high-end storage
systems. For more information about the various offerings, see the following website:
http://www.ibm.com/systems/storage/disk
The following section highlights a few of the offerings.
IBM Network-Attached Storage
IBM Network-Attached Storage (NAS) products provide a wide range of network attachment
capabilities to a broad range of host and client systems, such as IBM Scale Out Network
Attached Storage and the IBM System Storage Nxxx series. For more information about the
hardware and software, see the following website:
http://www.ibm.com/systems/storage/network
IBM Storwize family
The IBM Storwize® family is the ideal solution to optimize the data architecture for business
flexibility and data storage efficiency. Different models, such as the Storwize V3700, V5000,
and V7000, offer storage virtualization, IBM Real-time Compression™, IBM Easy Tier®, and
many other functions. For more information, see the following website:
http://www.ibm.com/systems/storage/storwize
IBM Flash Storage
IBM Flash Storage delivers extreme performance to derive measurable economic value
across the data architecture (servers, software, applications, and storage). IBM offers a
comprehensive flash portfolio with the IBM FlashSystem™ family. For more information, see
the following website:
http://www.ibm.com/systems/storage/flash
IBM XIV Storage System
IBM XIV® is a high-end disk storage system, helping thousands of enterprises meet the
challenge of data growth with hotspot-free performance and ease of use. Simple scaling, high
service levels for dynamic, heterogeneous workloads, and tight integration with hypervisors
and the OpenStack platform enable optimal storage agility for cloud environments.
XIV extends ease of use with integrated management for large and multi-site XIV
deployments, reducing operational complexity and enhancing capacity planning. For more
information, see the following website:
http://www.ibm.com/systems/storage/disk/xiv/index.html
IBM System Storage DS8000
The IBM System Storage DS8000 is a high-performance, high-capacity, and secure storage
system that is designed to deliver the highest levels of performance, flexibility, scalability,
resiliency, and total overall value for the most demanding, heterogeneous storage
environments. The system is designed to manage a broad scope of storage workloads that
exist in today’s complex data center, doing it effectively and efficiently.

Additionally, the IBM System Storage DS8000 includes a range of features that automate
performance optimization and application quality of service, and also provide the highest
levels of reliability and system uptime. For more information, see the following website:
http://www.ibm.com/systems/storage/disk/ds8000/index.html
2.11 Hardware Management Console
The Hardware Management Console (HMC) is a dedicated appliance that allows
administrators to configure and manage system resources on IBM Power Systems servers
that use IBM POWER6, POWER6+ POWER7, POWER7+, and POWER8 processors and the
PowerVM hypervisor. The HMC provides basic virtualization management support for
configuring logical partitions (LPARs) and dynamic resource allocation, including processor
and memory settings for selected Power Systems servers. The HMC also supports advanced
service functions, including guided repair and verification, concurrent firmware updates for
managed systems, and around-the-clock error reporting through IBM Electronic Service
Agent™ for faster support.
The HMC management features help improve server usage, simplify systems management,
and accelerate provisioning of server resources by using the PowerVM virtualization
technology.
The Power E870 and Power E880 platforms support two main service environments:
Attachment to one or more HMCs. This environment is the common configuration for
servers supporting logical partitions with dedicated or virtual I/O. In this case, all servers
have at least one logical partition.
Hardware support for customer-replaceable units. This support comes standard along
with the HMC. In addition, users can upgrade this support level to IBM onsite support to be
consistent with other Power Systems servers.
2.11.1 HMC code level
HMC V8R8.3.0 contains the following new features:
Support for Power E880 servers with the new 48-core system node option (#EPBD)
Support for Power E880 servers with the new option for a third and fourth I/O drawer
Support for Power E850 servers
Service Management:
New access points for Call Home functionality in Europe
Virtualization Management:
– Trial / temporary Enterprise Enablement support
– 52 sessions for local 5250 consoles (currently limited to 26)
Console Management:
Requirements:
When using the HMC with the Power E870 and Power E880 servers with a maximum of
two system nodes, the HMC code must be running at V8R8.2.0 level, or later.
When using the HMC with the Power E880 servers with a 12-core chip option or a
maximum of three or four system nodes, the HMC code must be running at V8R8.3.0
level, or later.

– Full release of the new enhanced user interface
– Browser currency
– Improved log retention (through file system resizing, rotation changes & content
reduction)
– Call Home support for modem (dial-in via AT&T Global Network) and VPN will be
removed
If you are attaching an HMC to a new server or adding a function to an existing server that
requires a firmware update, the HMC machine code might need to be updated to support the
firmware level of the server. In a dual HMC configuration, both HMCs must be at the same
version and release of the HMC code.
To determine the HMC machine code level that is required for the firmware level on any
server, go to the following website to access the Fix Level Recommendation Tool (FLRT) on
or after the planned availability date for this product:
https://www14.software.ibm.com/webapp/set2/flrt/home
FLRT identifies the correct HMC machine code for the selected system firmware level.
2.11.2 HMC RAID 1 support
HMCs now offer a high availability feature. The 7042-CR9, by default, includes two HDDs with
RAID 1 configured. RAID 1 is also offered on the 7042-CR6, 7042-CR7, 7042-CR8 and
7042-CR9 models (if the feature was removed from the initial order) as an MES upgrade
option.
RAID 1 uses data mirroring. Two physical drives are combined into an array, and the same
data is written to both drives. This makes the drives mirror images of each other. If one of the
drives experiences a failure, it is taken offline and the HMC continues operating with the other
drive.
HMC models
To use an existing HMC to manage any POWER8 processor-based server, the HMC must be
a model CR5, or later, rack-mounted HMC, or model C08, or later, deskside HMC. The latest
HMC model is the 7042-CR9. For your reference, Table 2-33 lists a comparison between the
7042-CR8 and the 7042-CR9 HMC models.
Table 2-33 Comparison between 7042-CR7 and 7042-CR8 models
Note: Access to firmware and machine code updates is conditional on entitlement and
license validation in accordance with IBM policy and practice. IBM may verify entitlement
through customer number, serial number electronic restrictions, or any other means or
methods that are employed by IBM at its discretion.
Note: The 7042-CR9 ships with 16 GB of memory, and is expandable to 192 GB with an
upgrade feature. 16 GB is advised for large environments or where external utilities, such
as PowerVC and other third party monitors, are to be implemented.
Feature CR8 CR9
IBM System x model x3550 M4 7914 PCH x3550 M5 5463 AC1
HMC model 7042-CR8 7042-CR9

2.11.3 HMC connectivity to the POWER8 processor-based systems
POWER8 processor-based servers, and their predecessor systems, 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.
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 may attach a second HMC to the remaining HMC port of the
server for redundancy. The two HMC ports must be addressed by two separate subnets.
Figure 2-40 shows a simple network configuration to enable the connection from the HMC to
the server and to allow for dynamic LPAR operations. For more information about HMC and
the possible network connections, see IBM Power Systems HMC Implementation and Usage
Guide, SG24-7491.
Figure 2-40 Network connections from the HMC to service processor and LPARs
Processor Intel 8-Core Xeon v2 2.00 GHz Intel 18-core Xeon v3 2.4 GHz
Memory max: 16 GB (when featured) 16 GB DDR4 expandable to 192 GB
DASD 500 GB 500 GB
RAID 1 Default Default
USB ports Two front, four back Two front, four rear
Integrated network Four 1 Gb Ethernet Four 1 Gb Ethernet
I/O slots One PCI Express 3.0 slot One PCI Express 3.0 slot
Feature CR8 CR9
Power Systems server
LPAR
n
LPAR
...
LPAR
2
LPAR
1
entxentxentxentx
Service
Processor
HMC2
eth0
HMC1eth1
Management LAN
HMC

By default, the service processor HMC ports are configured for dynamic IP address
allocation. The HMC can be configured as a DHCP server, providing an IP address at the
time that the managed server is powered on. In this case, the flexible service processor (FSP)
is allocated an IP address from a set of address ranges that are predefined in the HMC
software.
If the service processor of the managed server does not receive a DHCP reply before
timeout, predefined IP addresses are set up on both ports. Static IP address allocation is also
an option and can be configured using the ASMI menus.
2.11.4 High availability HMC configuration
The HMC is an important hardware component. Although Power Systems servers and their
hosted partitions can continue to operate when the managing HMC becomes unavailable,
certain operations, such as dynamic LPAR, partition migration using PowerVM Live Partition
Mobility, or the creation of a new partition, cannot be performed without the HMC. To avoid
such situations, consider installing a second HMC, in a redundant configuration, to be
available when the other is not (during maintenance, for example).
To achieve HMC redundancy for a POWER8 processor-based server, the server must be
connected to two HMCs:
The HMCs must be running the same level of HMC code.
The HMCs must use different subnets to connect to the service processor.
The HMCs must be able to communicate with the server’s partitions over a public network
to allow for full synchronization and functionality.
Notes: The two service processor HMC ports have the following features:
Run at a speed of 1 Gbps
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 directly from a client web browser
Use the following network configuration if no IP addresses are set:
– Service processor eth0 (HMC1 port): 169.254.2.147 with netmask 255.255.255.0
– Service processor eth1 (HMC2 port): 169.254.3.147 with netmask 255.255.255.0
For more information about the service processor, see 2.5.2, “Service Processor Bus” on
page 74.

Figure 2-41 shows one possible highly available HMC configuration that is managing two
servers. Each HMC is connected to one FSP port of each managed server.
Figure 2-41 Highly available HMC networking example
For simplicity, only the hardware management networks (LAN1 and LAN2) are highly
available (Figure 2-41). However, the open network (LAN3) can be made highly available by
using a similar concept and adding a second network between the partitions and HMCs.
For more information about redundant HMCs, see IBM Power Systems HMC Implementation
and Usage Guide, SG24-7491.
2.12 Operating system support
The IBM Power E870 and Power E880 systems support the following 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.
For details about the software available on IBM Power Systems, visit the IBM Power Systems
Software™ website:
http://www.ibm.com/systems/power/software/index.html
LAN 1 LAN 2
SYSTEM A SYSTEM B
LPAR A1
LPAR A2
LPAR A3
LPAR B1
LPAR B2
LPAR B3
FSPFSP
1 2 1 2
HMC 1
eth0eth1
HMC 2
eth0eth1
LAN 3
LAN 1 - Hardware management
network for first FSP ports (private)
LAN 2 - Hardware management
network for second FSP ports (private),
different network from LAN 1
LAN 3 – Open network for HMC access
and DLPAR operations

2.12.1 Virtual I/O Server
The minimum required levels of Virtual I/O Server for both the Power E870 and Power E880
are:
VIOS 2.2.3.4 with ifix IV63331
VIOS 2.2.2.6
IBM regularly updates the Virtual I/O Server code. To find information about the latest
updates, visit the Fix Central website:
http://www-933.ibm.com/support/fixcentral/
2.12.2 IBM AIX operating system
The following sections discuss the various levels of AIX operating system support.
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 that are
included on CD-ROM.
The Service Update Management Assistant (SUMA), 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 the following website:
http://www14.software.ibm.com/webapp/set2/sas/f/genunix/suma.html
IBM AIX Version 6.1
A partition that uses AIX 6.1 can run in POWER6, POWER6+, or POWER7 mode. This will
limit the partition to SMT-4 among other hardware capabilities.
The minimum level of AIX Version 6.1 supported on the Power E870 and Power E880
depends on the partition having 100% virtualized resources or not.
For partitions that have all of their resources virtualized via Virtual I/O Server, the minimum
levels of AIX Version 6.1 supported on the Power E870 and Power E880 are as follows:
AIX Version 6.1 with the 6100-08 Technology Level and Service Pack 1 or later
For all other partitions, the minimum levels of AIX Version 6.1 supported on the Power E870
and Power E880 are as follows:
AIX Version 6.1 with the 6100-08 Technology Level and Service Pack 6, or later
AIX Version 6.1 with the 6100-09 Technology Level and Service Pack 4, and APAR
IV63331, or later

IBM AIX Version 7.1
A partition that uses AIX 7.1 can run in POWER6, POWER6+, POWER7, or POWER8 mode.
This allows for an easier migration from previous systems and full exploitation of the
POWER8 features.
The minimum level of AIX Version 7.1 supported on the Power E870 and Power E880
depends on the partition having 100% virtualized resources or not.
For partitions that have all of their resources virtualized via Virtual I/O Server, the minimum
levels of AIX Version 7.1 supported on the Power E870 and Power E880 are as follows:
For all other partitions, the minimum levels of AIX Version 7.1 supported on the Power E870
and Power E880 are as follows:
AIX Version 7.1 with the 7100-02 Technology Level Service Pack 6, or later
AIX Version 7.1 with the 7100-03 Technology Level Service Pack 4, and APARs IV63332
and IV69116, or later
2.12.3 IBM i operating system
The IBM i operating system is supported on the Power E870 and Power E880 with the
following minimum required levels:
IBM i 7.1 TR9 or later
IBM i 7.2 TR1 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:
Visit the IBM Prerequisite website for compatibility information for hardware features and the
corresponding AIX and IBM i Technology Levels:
http://www-912.ibm.com/e_dir/eserverprereq.nsf
2.12.4 Linux operating systems
Linux is an open source operating system that runs on numerous platforms from embedded
systems to mainframe computers. It provides an implementation like UNIX across many
computer architectures.
The supported versions of Linux on Power E870 and Power E880 are as follows:
Big endian:
– SUSE Linux Enterprise Server 11 Service Pack 3, or later
– Red Hat Enterprise Linux 6.5 for POWER, or later
Note: The minimum supported levels for the Power E880 with the new processor or I/O
drawer options are:
IBM i 7.1 TR10
IBM i 7.2 TR2

Little endian:
– Red Hat Enterprise Linux 7.1, or later
– SUSE Linux Enterprise Server 12 and later Service Packs
– Ubuntu 15.04
If you want to configure Linux partitions in virtualized Power Systems, be aware of the
following 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 POWER8
processor-based servers, or from other Linux distributors.
For information about features and external devices that are supported by Linux, see this site:
http://www.ibm.com/systems/p/os/linux/index.html
For information about SUSE Linux Enterprise Server, see this site:
http://www.novell.com/products/server
For information about Red Hat Enterprise Linux Advanced Server, see this site:
http://www.redhat.com/rhel/features
2.12.5 Java versions that are supported
Java is supported on POWER8 servers. For best exploitation of the performance capabilities
and most recent improvements of POWER8 technology, upgrade Java-based applications to
Java 7 or Java 6. For more information, visit these websites:
http://www.ibm.com/developerworks/java/jdk/aix/service.html
http://www.ibm.com/developerworks/java/jdk/linux/download.html
2.12.6 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
POWER8 processor architecture. Release after release, these compilers continue to deliver
application performance improvements and additional 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. Compilers help you unleash the full power of your IT investment. With the
XL compilers you can create and maintain critical business and scientific applications, while
maximizing application performance and improving developer productivity. The performance
gain from years of compiler optimization experience is seen in the continuous
release-to-release compiler improvements that support the POWER and POWERPC families
of processors. XL compilers support POWER4, POWER4+, POWER5, POWER5+,
POWER6, POWER7, and POWER7+ processors, and now add support for the new POWER8
processors. With the support of the latest POWER8 processor chip, IBM advances a more
than 20-year investment in the XL compilers for POWER series and IBM PowerPC® series
architectures.
XL C, XL C/C++ and XL Fortran features introduced to exploit the latest POWER8 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.
XL compilers support application development on big endian distributions. XL C/C++ for
Linux, V13.1.1 and XL Fortran for Linux, V15.1.1 deliver new compilers that support
application development on the IBM POWER8 servers that run the little endian Linux
distributions. With these two releases, compiler support on the Linux distributions Ubuntu
14.04 for IBM POWER8, Ubuntu 14.10 for IBM POWER8, and SUSE Linux Enterprise Server
12 for Power, includes exploitation of the little endian architecture on the POWER8 processor.
IBM COBOL for AIX enables you to selectively target code generation of your programs to
either exploit a particular POWER 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 IBM 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. The PL/I compiler produces code that is intended to
perform well across all hardware levels on AIX.
IBM Rational® Developer for AIX and Linux, C/C++ Edition provides a rich set of integrated
development tools that support XL C for AIX, XL C/C++ for AIX, and XL C/C++ for Linux
compiler. It also supports the GNU compiler and debugger on Linux on x86 architectures to
make it possible to do development on other infrastructures and then easily port and optimize
the resultant workloads to run on POWER and fully exploit the POWER platform's unique
qualities of service. Tool capabilities include file management, searching, smart assistive
editing, application analysis, unit test automation, code coverage analysis, a unique expert
system Migration and Porting Assistant, a unique expert system Performance Advisor, local
build, and cross-language/cross-platform debugger, all integrated into an Eclipse workbench.
This solution can greatly improve developers' productivity and initial code quality with
resultant benefits to downstream disciplines such as QA and Operations.
IBM Rational Developer for AIX and Linux, AIX COBOL Edition provides a rich set of
integrated development tools that support the COBOL for AIX compiler. Capabilities include
file management, searching, smart assistive editing, application analysis, local build, and
cross-language/cross-platform debugger, all integrated into an Eclipse workbench. This
solution can boost developers' productivity by moving from older, text-based, command-line
development tools to a rich set of integrated development tools and unlike competing
distributed COBOL IDEs, is not dependent upon an expensive companion COBOL runtime
environment.
IBM Rational Development Studio for IBM i 7.2 provides programming languages for creating
modern business applications such as these:
The ILE RPG, ILE COBOL, C, and C++ compilers
The heritage RPG and COBOL compilers
Application Development ToolSet (ADTS), the text-based development tools (such as SEU
and PDM)
The latest release includes support for free-form RPG and better Unicode handling and
continues to support Open Access: RPG Edition as an included technology.

IBM Rational Developer for i provides a rich set of integrated development tools that support
the IBM i ILE compilers. Tool capabilities include file management, searching, smart assistive
editing, application analysis, code coverage analysis, and cross-language/cross-platform
debugger, all integrated into an Eclipse workbench. This solution can greatly improve
developers' productivity as compared to using text-based and command-line development
tools, and is more appealing than such tools when you need to recruit new IBM i development
talent.
2.13 Energy management
The Power E870 and Power E880 systems have features to help clients become more energy
efficient. EnergyScale technology enables advanced energy management features to
conserve power dramatically and dynamically and further improve energy efficiency.
Intelligent Energy optimization capabilities enable the POWER8 processor to operate at a
higher frequency for increased performance and performance per watt, or dramatically
reduce frequency to save energy.
2.13.1 IBM EnergyScale technology
IBM EnergyScale technology provides functions to help the user understand and dynamically
optimize processor performance versus processor energy consumption, and system
workload, to control IBM Power Systems power and cooling usage.
EnergyScale uses power and thermal information that is collected from the system to
implement policies that can lead to better performance or better energy usage. IBM
EnergyScale has the following features:
Power trending
EnergyScale provides continuous collection of real-time server energy consumption. It
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 data center energy consumption at various times of the day, week,
or month.
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 almost 50%
down from nominal frequency (the actual value depends on the server type and
configuration).
Power saver mode is not supported during system start, although it is a persistent
condition that is sustained after the boot when the system starts running instructions.

Dynamic power saver mode
Dynamic power saver mode varies processor frequency and voltage based on the usage
of the POWER8 processors. Processor frequency and usage are inversely proportional for
most workloads, implying that as the frequency of a processor increases, its usage
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 usage.
When a system is idle, the system firmware lowers the frequency and voltage to power
energy saver mode values. When fully used, the maximum frequency varies, depending
on whether the user favors power savings or system performance. If an administrator
prefers energy savings and a system is fully used, the system is designed to reduce the
maximum frequency to about 95% of nominal values. If performance is favored over
energy consumption, the maximum frequency can be increased above the nominal
frequency for extra performance. Table 2-34 shows the maximum frequency boost
available for different speed processors in the Power E870 and E880.
Table 2-34 Maximum frequency boosts for Power E870 and E880 processors
The frequency boost figures are maximums and will depend on the environment where the
servers are installed. Maximum boost frequencies may not be reached if the server is
installed in higher temperatures or at altitude.
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 extra power that was 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 data center. 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.
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 ensured 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.
System Cores per chip Nominal speed Maximum boost
speed
Power E870 8 4.024 GHz 4.123 GHz
Power E870 10 4.190 GHz 4.456 GHz
Power E880 8 4.356 GHz 4.522 GHz
Power E880 12 4.024 GHz 4.256 GHz

Processor core nap mode
IBM POWER8 processor uses a low-power mode that is called nap that stops processor
execution when there is no work to do on that processor core. The latency of exiting nap
mode is small, typically not generating any impact on applications running. Therefore, the
IBM 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 off the clock on most of the circuits
in 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
and causes a cumulative effect. Nap mode saves 10 - 15% of power consumption in the
processor core.
Processor core sleep mode
To save even more energy, the POWER8 processor has an even lower power mode
referred to as sleep. Before a core and its associated private L2 cache enter sleep mode,
the cache is flushed, transition lookaside buffers (TLB) are invalidated, and the hardware
clock is turned off in the core and in the cache. Voltage is reduced to minimize leakage
current. Processor cores that are inactive in the system (such as capacity on demand
(CoD) processor cores) are kept in sleep mode. Sleep mode saves about 80% power
consumption in the processor core and its associated private L2 cache.
Processor chip winkle mode
The most amount of energy can be saved when a whole POWER8 chiplet enters the
winkle mode. In this mode, the entire chiplet is turned off, including the L3 cache. This
mode can save more than 95% power consumption.
Fan control and altitude input
System firmware dynamically adjusts 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), the fan
speed varies based on power consumption and ambient temperature.
Processor folding
Processor folding is a consolidation technique that dynamically adjusts, over the short
term, the number of processors that are available for dispatch to match the number of
processors that are demanded by the workload. As the workload increases, the number of
processors made available increases. As the workload decreases, the number of
processors that are 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 POWER8 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 POWER8
processor-based servers and the expansion units that they support.

Server power down
If overall data center processor usage is low, workloads can be consolidated on fewer
numbers of servers so that some servers can be turned off completely. Consolidation
makes sense when there are long periods of low usage, such as weekends. Live Partition
Mobility can be used to move workloads to consolidate partitions onto fewer systems,
reducing the number of servers that are powered on and therefore reducing the power
usage.
On POWER8 processor-based systems, several EnergyScale technologies are
embedded in the hardware and do not require an operating system or external
management component. Fan control, environmental monitoring, and system energy
management are controlled by the On Chip Controller (OCC) and associated components.
The power mode can also be set up without external tools, by using the ASMI interface, as
shown in Figure 2-42.
Figure 2-42 Setting the power mode in ASMI
2.13.2 On Chip Controller
To maintain the power dissipation of POWER7+ with its large increase in performance and
bandwidth, POWER8 invested significantly in power management innovations. A new On
Chip Controller (OCC) using an embedded IBM PowerPC core with 512 KB of SRAM runs
real-time control firmware to respond to workload variations by adjusting the per-core
frequency and voltage based on activity, thermal, voltage, and current sensors.
The on-die nature of the OCC allows for approximately 100× speedup in response to
workload changes over POWER7+, enabling reaction under the timescale of a typical OS
time slice and allowing for multi-socket, scalable systems to be supported. It also enables
more granularity in controlling the energy parameters in the processor, and increases
reliability in energy management by having one controller in each processor that can perform
certain functions independently of the others.
POWER8 also includes an internal voltage regulation capability that enables each core to run
at a different voltage. Optimizing both voltage and frequency for workload variation enables
better increase in power savings versus optimizing frequency only.

2.13.3 Energy consumption estimation
Often, for Power Systems, various energy-related values are important:
Maximum power consumption and power source loading values
These values are important for site planning and are described in the IBM Knowledge
Center, found at the following website:
http://pic.dhe.ibm.com/infocenter/powersys/v3r1m5/index.jsp
Search for type and model number and “server specifications”. For example, for the
Power E870 and Power E880 system, search for “9119-MME and 9119-MHE server
specifications”.
An estimation of the energy consumption for a certain configuration
The calculation of the energy consumption for a certain configuration can be done in the
IBM Systems Energy Estimator, found at the following website:
http://www-912.ibm.com/see/EnergyEstimator/
In that tool, select the type and model for the system, and enter some details about the
configuration and wanted CPU usage. As a result, the tool shows the estimated energy
consumption and the waste heat at the wanted usage and also at full usage.

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 offer key capabilities that can help
you consolidate and simplify your IT environment:
Improve server usage and share I/O resources to reduce total cost of ownership (TCO)
and make better usage of IT assets.
Improve business responsiveness and operational speed by dynamically reallocating
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, so you can make business-driven policies to deliver resources based on time,
cost, and service-level requirements.
Single Root I/O Virtualization (SR-IOV) is now also supported on the Power E870 and
Power E880 server. For more information about SR-IOV see chapter 3.4, “Single root I/O
virtualization (SR-IOV)” on page 126.
3
PowerVM license: Enterprise edition is default on Power E870 and Power E880.
Note: PowerKVM is not supported on the Power E870 and Power E880. PowerVM is the
only virtualization technology available for these systems.

3.1 IBM POWER Hypervisor
Combined with features in the POWER8 processors, the IBM 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 132).
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 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 that is required by
the POWER Hypervisor firmware varies according to several factors:
Number of logical partitions
Number of physical and virtual I/O devices that are used by the logical partitions
Maximum memory values that are specified in the logical partition profiles
The minimum amount of physical memory that is required to create a partition is the size of
the system’s logical memory block (LMB). The default LMB size varies according to the
amount of memory that is configured in the system (see Table 3-1).
Table 3-1 Configured system memory-to-default logical memory block size
In most cases, however, the actual minimum requirements and preferences for the supported
operating systems are greater than 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 Central Electronics Complex
memory
Default logical memory block
Up to 32 GB 128 MB
Greater than 32 GB 256 MB

Chapter 3. Virtualization 121
3.1.1 Virtual SCSI
The POWER Hypervisor provides a virtual SCSI mechanism for the virtualization of storage
devices. The storage virtualization is accomplished by using two paired adapters:
A virtual SCSI server adapter
A virtual SCSI client adapter
A Virtual I/O Server (VIOS) partition or an IBM i partition can define virtual SCSI server
adapters. Other partitions are client partitions. The VIOS partition is a special logical partition,
which is described in 3.5.4, “Virtual I/O Server” on page 134. The VIOS software is included
on all PowerVM editions. When using the PowerVM Standard Edition and PowerVM
Enterprise Edition, dual VIOSes can be deployed to provide maximum availability for client
partitions when performing VIOS maintenance.
3.1.2 Virtual Ethernet
The POWER Hypervisor provides a virtual Ethernet switch function that allows partitions on
the same server to use fast and secure communication without any need for physical
interconnection. The virtual Ethernet allows a transmission speed up to 20 Gbps, depending
on the maximum transmission unit (MTU) size, type of communication, and CPU entitlement.
Virtual Ethernet support began with IBM AIX Version 5.3, IBM i V5R3, Red Hat Enterprise
Linux 4, and SUSE Linux Enterprise Server, 9, and it is supported on all later versions.
(For more information, see 3.5.8, “Operating system support for PowerVM” on page 144).
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 (or 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 VIOS partition, also known as Shared Ethernet
Adapter. For more information about shared Ethernet, see 3.5.4, “Virtual I/O Server” on
page 134.
Adapter and access: 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.

3.1.3 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 VIOS logical partition. The
VIOS logical partition provides the connection between the virtual Fibre Channel adapters on
the VIOS logical partition and the physical Fibre Channel adapters on the managed system.
Figure 3-1 shows the connections between the client partition virtual Fibre Channel adapters
and the external storage. For more information, see 3.5.8, “Operating system support for
PowerVM” on page 144.
Figure 3-1 Connectivity between virtual Fibre Channels adapters and external SAN devices
3.1.4 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 (HMC) virtual TTY, IVM virtual TTY, or from a terminal
emulator that is connected to a system port.
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

3.2 POWER processor modes
Although they are not virtualization features, the POWER processor modes are described
here because they affect various virtualization features.
On Power System servers, partitions can be configured to run in several modes, including the
following modes:
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 website:
http://power.org/wp-content/uploads/2012/07/PowerISA_V2.05.pdf
POWER6+ compatibility mode
This mode is similar to POWER6, with eight more storage protection keys.
POWER7 mode
This is the mode for POWER7+ and POWER7 processors, implementing Version 2.06 of
the Power Instruction Set Architecture. For more information, visit the following website:
http://power.org/wp-content/uploads/2012/07/PowerISA_V2.06B_V2_PUBLIC.pdf
POWER8 mode
This is the native mode for POWER8 processors implementing Version 2.07 of the Power
Instruction Set Architecture. For more information, visit the following address:
https://www.power.org/documentation/power-isa-version-2-07/
The selection of the mode is made on a per-partition basis, from the managed console, by
editing the partition profile.

Figure 3-2 shows the compatibility modes within the LPAR profile.
Figure 3-2 Configuring partition profile compatibility mode using the HMC
Table 3-2 lists the differences between the processor modes.
Table 3-2 Differences between POWER6, POWER7, and POWER8 compatibility modes
POWER6 and
POWER6+ mode
POWER7 mode POWER8 mode Customer value
2-thread SMT 4-thread SMT 8-thread SMT Throughput
performance, and
processor core usage
Vector Multimedia
Extension/ AltiVec
(VMX)
Vector scalar
extension (VSX)
VSX2
In-Core Encryption
Acceleration
High-performance
computing

3.3 Active Memory Expansion
Active Memory Expansion is an optional feature for the enterprise servers Power E870 and
Power E880.
This feature enables memory expansion on the system. By using compression and
decompression of memory, content can effectively expand the maximum memory capacity,
providing additional server workload capacity and performance.
Active Memory Expansion is a technology that allows the effective maximum memory
capacity to be much larger than the true physical memory maximum. Compression and
decompression of memory content can allow memory expansion up to 125% for AIX
partitions, which in turn enables a partition to perform 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 uses the CPU resource of a partition to compress and 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 and decompression.
The POWER8 processor includes Active Memory Expansion on the processor chip to provide
dramatic improvement in performance and greater processor efficiency. To take advantage of
the hardware compression offload, AIX 6.1 Technology Level 9 or later is required, for node1
and node2 with HMCV8R8.2.0 and for node2 and node3 with HMCV8R8.3.0+FW830.
Affinity off by default 3-tier memory,
micropartition affinity,
and dynamic platform
optimizer
HW memory
affinity tracking
assists
Micropartition
prefetch
Concurrent LPARs
per core
Improved system
performance for
system images
spanning sockets and
nodes
64-core and
128-thread scaling
32-core and
128-thread scaling
64-core and
256-thread scaling
128-core and
512-thread scaling
256-core and
1024-thread
scaling
192-core and
1536-thread
scaling
Hybrid threads
Transactional
memory
Active system
optimization
hardware assists
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
WINKLE, NAP, SLEEP,
and Idle power saver
Improved energy
efficiency
POWER6 and
POWER6+ mode
POWER7 mode POWER8 mode Customer value
Note: The Active Memory Expansion feature is not supported by IBM i and the Linux
operating system.

Tests in IBM laboratories, using sample work loads, showed excellent results for many
workloads in terms of memory expansion per additional CPU used. Other test workloads had
more modest results. The ideal scenario is when there are many cold pages, that is,
infrequently referenced pages. However, if many memory pages are referenced frequently,
Active Memory Expansion might not be a preferred choice.
For more information about Active Memory Expansion, download the document Active
Memory Expansion: Overview and Usage Guide, found at:
http://public.dhe.ibm.com/common/ssi/ecm/en/pow03037usen/POW03037USEN.PDF
3.4 Single root I/O virtualization (SR-IOV)
Single root I/O virtualization (SR-IOV) is an extension to the PCI Express (PCIe) specification
that allows multiple operating systems to simultaneously share a PCIe adapter with little or no
runtime involvement from a hypervisor or other virtualization intermediary.
SR-IOV is PCI standard architecture that enables PCIe adapters to become self-virtualizing. It
enables adapter consolidation, through sharing, much like logical partitioning enables server
consolidation. With an adapter capable of SR-IOV, you can assign virtual slices of a single
physical adapter to multiple partitions through logical ports; all of this is done without the need
for a Virtual I/O Server (VIOS).
Initial SR-IOV deployment supports up to 48 logical ports per adapter, depending on the
adapter. You can provide additional fan-out for more partitions by assigning a logical port to a
VIOS, and then using that logical port as the physical device for a Shared Ethernet Adapter
(SEA). VIOS clients can then use that SEA through a traditional virtual Ethernet configuration.
Overall, SR-IOV provides integrated virtualization without VIOS and with greater server
efficiency as more of the virtualization work is done in the hardware and less in the software.
The following are the hardware requirements to enable SR-IOV:
One of the following pluggable PCIe adapters:
– PCIe2 4-port (10Gb FCoE and 1GbE) SR&RJ45 Adapter (#EN0H)
– PCIe2 4-port (10Gb FCoE and 1GbE) SFP+Copper and RJ4 Adapter (#EN0K)
– PCIe2 LP 4-port(10Gb FCoE & 1GbE) SFP+Copper&RJ45 (#EN0L)
– PCIe2 4-port (10Gb FCoE and 1GbE) LR&RJ45 Adapter (#EN0M)
– PCIe2 LP 4-port(10Gb FCoE & 1GbE) LR&RJ45 Adapter (#EN0N)
– PCIe3 4-port (10Gb) SR Adapter (#EN15)
– PCIe3 LP 4-port (10Gb) SR Adapter (#EN16)
– PCIe3 4-port (10Gb) SFP+Copper Adapter (#EN17)
– PCIe3 LP 4-port (10Gb) SFP+Copper Adapter (#EN18)
The minimum operating system requirements, related to SR-IOV functions, are as follows:
VIOS
– Virtual I/O Server Version 2.2.3.51
AIX
– AIX 6.1 Technology Level 9 with Service Pack 5 and APAR IV68443 or later
– AIX 7.1 Technology Level 3 with Service Pack 5 and APAR IV68444 or later
Tip: If the workload is based on Java, the garbage collector must be tuned so that it does
not access the memory pages so often, that is, turning cold pages to hot.

Linux
– SUSE Linux Enterprise Server 11 SP3, or later
– SUSE Linux Enterprise Server 12, or later
– Red Hat Enterprise Linux 6.5, or later
– Red Hat Enterprise Linux 7, or later
IBM i
– IBM i 7.1 TR10 or later
– IBM i 7.2 TR2 or later
The entire adapter (all four ports) is configured for SR-IOV or none of the ports is. (FCoE not
supported when using SR-IOV).
SR-IOV provides significant performance and usability benefits, as described in the following
sections.
3.4.1 Direct access I/O and performance
The primary benefit of allocating adapter functions directly to a partition, as opposed to using
a virtual intermediary (VI) like VIOS, is performance. The processing overhead involved in
passing client instructions through a VI, to the adapter and back, are substantial.
With direct access I/O, SR-IOV capable adapters running in shared mode allow the operating
system to directly access the slice of the adapter that has been assigned to its partition, so
there is no control or data flow through the hypervisor. From the partition perspective, the
adapter appears to be physical I/O. With respect to CPU and latency, it exhibits the
characteristics of physical I/O; and because the operating system is directly accessing the
adapter, if the adapter has special features, like multiple queue support or receive side
scaling (RSS), the partition can leverage those also, if the operating system has the capability
in its device drivers.
3.4.2 Adapter sharing
The current trend of consolidating servers to reduce cost and improve efficiency is increasing
the number of partitions per system, driving a requirement for more I/O adapters per system
to accommodate them. SR-IOV addresses and simplifies that requirement by enabling the
sharing of SR-IOV capable adapters. Because each adapter can be shared and directly
accessed by up to 48 partitions, depending on the adapter, the partition to PCI slot ratio can
be significantly improved without adding the overhead of a virtual intermediary.
3.4.3 Adapter resource provisioning (QoS)
Power Systems SR-IOV provides QoS controls to specify a capacity value for each logical
port, improving the ability to share adapter ports effectively and efficiently. The capacity value
Firmware level: SR-IOV is supported from firmware level SC820_067 (FW820.10) for
POWER8 processor-based servers. Check the Fix Central portal to verify the specific
firmware level for your type of the machine at:
https://www.ibm.com/support/fixcentral/

determines the desired minimum percentage of the physical port’s resources that should be
applied to the logical port.
The exact resource represented by the capacity value can vary based on the physical port
type and protocol. In the case of Ethernet physical ports, capacity determines the minimum
percentage of the physical port’s transmit bandwidth that the user desires for the logical port.
For example, consider Partitions A, B, and C, with logical ports on the same physical port. If
Partition A is assigned an Ethernet logical port with a capacity value of 20%, Partitions B and
C cannot use more than 80% of the physical port’s transmission bandwidth unless Partition A
is using less than 20%. Partition A can use more than 20% if bandwidth is available. This
ensures that, although the adapter is being shared, the partitions maintain their portion of the
physical port resources when needed.
3.4.4 Flexible deployment
Power Systems SR-IOV enables flexible deployment configurations, ranging from a simple,
single-partition deployment, to a complex, multi-partition deployment involving VIOS
partitions and VIOS clients running different operating systems.
In a single-partition deployment, the SR-IOV capable adapter in shared mode is wholly owned
by a single partition, and no adapter sharing takes place. This scenario offers no practical
benefit over traditional I/O adapter configuration, but the option is available.
In a more complex deployment scenario, an SR-IOV capable adapter could be shared by both
VIOS and non-VIOS partitions, and the VIOS partitions could further virtualize the logical
ports as shared Ethernet adapters for VIOS client partitions. This scenario leverages the
benefits of direct access I/O, adapter sharing, and QoS that SR-IOV provides, and also the
benefits of higher-level virtualization functions, such as Live Partition Mobility (for the VIOS
clients), that VIOS can offer.

3.4.5 Reduced costs
SR-IOV facilitates server consolidation by reducing the number of physical adapters, cables,
switch ports, and I/O slots required per system. This translates to reduced cost in terms of
physical hardware required, and also reduced associated energy costs for power
consumption, cooling, and floor space. You may save additional cost on CPU and memory
resources, relative to a VIOS adapter sharing solution, because SR-IOV does not have the
resource overhead inherent in using a virtualization intermediary to interface with the
adapters.
3.5 PowerVM
The PowerVM platform is the family of technologies, capabilities, and offerings that delivers
industry-leading virtualization on the IBM Power Systems. It is the umbrella branding term for
Power Systems virtualization (logical partitioning, IBM Micro-Partitioning, POWER
Hypervisor, VIOS, Live Partition Mobility, and more). As with Advanced Power Virtualization in
the past, PowerVM is a combination of hardware enablement and software. The licensed
features of each of the two separate editions of PowerVM are described here.
3.5.1 PowerVM edition
Power VM Enterprise Edition (#5228) is standard on Power E870 and Power E880 servers.
Power VM Standard Edition is not supported on the Power E870 and E880. PowerVM
Enterprise Edition offers all PowerVM features including Advanced Memory Sharing (AMS)
and Live Partition Mobility (LPM) for specific operating systems. To verify the details about
what is specifically supported by the operating system, see Table 3-3 on page 144.
3.5.2 Logical partitions
Logical partitions (LPARs) and virtualization increase the usage of system resources and add
a level of configuration possibilities.
Logical partitioning
Logical partitioning was introduced with the POWER4 processor-based product line and AIX
Version 5.1, Red Hat Enterprise Linux 3.0, and SUSE Linux Enterprise Server 9.0 operating
systems. This technology was able to divide an IBM eServer™ pSeries (now IBM System p)
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 running. AIX Version 5.2, with all the necessary
enhancements to enable dynamic LPAR, was introduced in 2002. At the same time, Red Hat
Enterprise Linux 5 and SUSE Linux Enterprise 9.0 were also able to support dynamic logical
partitioning. 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.

Micro-Partitioning
When you use the Micro-Partitioning technology, you can 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
micropartition. Micropartitions run over a set of processors that are called a shared processor
pool, and virtual processors are used to let the operating system manage the fractions of
processing power that are assigned to the logical partition. From an operating system
perspective, a virtual processor cannot be distinguished from a physical processor, unless the
operating system is enhanced to determine 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.
When defining a shared processor partition, several options must be defined:
The minimum, wanted, 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
Select a pool from the list with the names of each configured shared processor pool. This
list also shows, in parentheses, the pool ID of each configured shared processor pool. If
the name of the wanted shared processor pool is not available here, you must first
configure the shared processor pool by using the shared processor pool Management
window. Shared processor partitions use the default shared processor pool, called
DefaultPool by default. For more information about multiple shared processor pools, see
3.5.3, “Multiple shared processor pools” on page 133.
Whether the partition can access extra processing power to “fill up” its virtual processors
above its capacity entitlement (you select either to cap or uncap your partition)
If spare processing power is 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) if there is an uncapped partition
The minimum, wanted, and maximum number of virtual processors
The POWER Hypervisor calculates partition processing power based on minimum, wanted,
and maximum values, processing mode, and the requirements of other active partitions. The
actual entitlement is never smaller than the processing unit’s wanted value, but can exceed
that value if it is an uncapped partition and up to the number of virtual processors that are
allocated.
On the POWER8 processors, a partition can be defined with a processor capacity as small as
0.05 processing units. This number represents 0.05 of a physical core. Each physical core
can be shared by up to 20 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 Power E870 supports up to 80 cores in a single system. Here are the maximum
numbers:
80 dedicated partitions
1024 micropartitions
The Power E880 has statement of direction to support up to 192 cores in a single system.
Here are the maximum numbers:
192 dedicated partitions
1024 micropartitions
An important point is that the maximum amounts are supported by the hardware, but the
practical limits depend on application workload demands.
Consider the following additional information about virtual processors:
A virtual processor can be running (dispatched) either on a physical core or as standby
waiting for a physical core 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 a 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 3-3 shows a diagram of the concepts that are
described in this section.
Figure 3-3 Logical partitioning concepts
Dedicated mode
In dedicated mode, physical processors are assigned as a whole to partitions. The
simultaneous multithreading feature in the POWER8 processor core allows the core to run
instructions from two, four, or eight independent software threads simultaneously.
To support this feature, consider the concept of logical processors. The operating system
(AIX or Linux) sees one physical core as two, four, or eight logical processors if the
simultaneous multithreading feature is on. It can be turned off and on dynamically while the
operating system is running (for AIX, run smtctl, for Linux, run ppc64_cpu --smt, and for
IBM i, use QPRCMLTTSK system value). If simultaneous multithreading is off, each physical core
is presented as one logical processor in AIX or Linux, and thus only one thread.
Shared dedicated mode
On POWER8 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
increase system usage without compromising the computing power for critical workloads in a
dedicated processor.
Set of micro-partitions
KEY:
vp Virtual processor
lp Logical processor
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
AIXV7.1
AIXV6.1
Dedicated processors Dedicated processors
POWER Hypervisor
10-core SMP System

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 20 times the number of processing units that are assigned to the partition). From
the POWER Hypervisor perspective, 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 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 on whether simultaneous multithreading is
turned on or off.
3.5.3 Multiple shared processor pools
Multiple shared processor pools (MSPPs) are supported on POWER8 processor-based
servers. This capability allows a system administrator to create a set of micropartitions with
the purpose of controlling the processor capacity that can be consumed from the physical
shared processor pool.
Implementing MSPPs depends on a set of underlying techniques and technologies.
Figure 3-4 shows an overview of the architecture of multiple shared processor pools.
Figure 3-4 Overview of the architecture of multiple shared processor pools
POWER Hypervisor
p1p0
Physical Shared-Processor Pool
p2 p3 p4 p5 p6 p7
Shared Processor Pool0
IBM i V7.2
EC 1.6
AIX V7.1
EC 0.8
Linux
EC 0.5
vp0
vp1
vp2 vp3
vp4
AIX V7.1
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

Micropartitions 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 micropartitions 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 micropartition within a shared
processor pool is ensured its processor entitlement plus any capacity that it might be
allocated from the reserved pool capacity if the micropartition is uncapped.
If certain micropartitions in a shared processor pool do not use their capacity entitlement, the
unused capacity is ceded and other uncapped micropartitions 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
micropartitions within that shared processor pool.
All Power Systems servers that support the multiple shared processor pools capability have a
minimum of one (the default) shared processor pool and up to a maximum of 64 shared
processor pools.
3.5.4 Virtual I/O Server
The virtual I/O Server (VIOS) 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
usage (for example, consolidation). In this case, the VIOS 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 VIOS eliminates the requirement that every partition owns a dedicated network adapter,
disk adapter, and disk drive. The VIOS supports OpenSSH for secure remote logins. It also
provides a firewall for limiting access by ports, network services, and IP addresses.
Figure 3-5 shows an overview of a VIOS configuration.
Figure 3-5 Architectural view of the VIOS
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

Because the VIOS is an operating system-based appliance server, redundancy for physical
devices that are attached to the VIOS can be provided by using capabilities such as Multipath
I/O and IEEE 802.3ad Link Aggregation.
Installation of the VIOS 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
VIOS, and is supported only in special VIOS partitions. Three major virtual devices are
supported by the VIOS:
Shared Ethernet Adapter
Virtual SCSI
Virtual Fibre Channel adapter
The Virtual Fibre Channel adapter is used with the NPIV feature, as described in 3.5.8,
“Operating system support for PowerVM” on page 144.
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 POWER 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 SEA also provides the ability for several client partitions to share one physical adapter.
With an SEA, you can connect internal and external VLANs by using a physical adapter. The
Shared Ethernet Adapter service can be hosted only in the VIOS, 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 VIOS.

Figure 3-6 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 VIOS that
shares the physical access.
Figure 3-6 Architectural view of a Shared Ethernet Adapter
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 VIOS 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.
Virtual SCSI
Virtual SCSI is used to see a virtualized implementation of the SCSI protocol. Virtual SCSI is
based on a client/server relationship. The VIOS 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 that are provided by the VIOS as clients.
VIOS Client 1
Ethernet
switch
VLAN=2 PVID=1
ent3
(sea)
en3
(if.)
en0
(if.)
Client 2
ent0
(virt.)
VLAN=2
PVID=2
PVID=99
VID=2
PVID=1
PVID=1
PVID=1
VLAN=1
Hypervisor
External
Network
ent0
(phy.)
en0
(if.)
ent0
(virt.)
Client 3
en0
(if.)
ent0
(virt.)
ent1
(virt.)
ent2
(virt.)

The virtual I/O adapters (virtual SCSI server adapter and a virtual SCSI client adapter) are
configured by using a managed console or through the Integrated Virtualization Manager on
smaller systems. The virtual SCSI server (target) adapter is responsible for running any SCSI
commands that it receives. It is owned by the VIOS 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 VIOS.
Physical disks that are presented to the Virtual/O Server can be exported and assigned to a
client partition in various 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 VIOS 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.
For more information about specific storage devices that are supported for VIOS, see the
following website:
http://www14.software.ibm.com/webapp/set2/sas/f/vios/documentation/datasheet.html
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 VIOS partition that acts only as a pass-through, managing the data transfer
through the POWER Hypervisor.
Each partition that uses 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 more information and requirements for NPIV, see the following resources:
PowerVM Migration from Physical to Virtual Storage, SG24-7825
IBM PowerVM Virtualization Managing and Monitoring, SG24-7590
Virtual I/O Server functions
The Virtual I/O Server (VIOS) has many features, including monitoring solutions and the
following capabilities:
Support for Live Partition Mobility starting on POWER6 processor-based systems with the
PowerVM Enterprise Edition. For more information about Live Partition Mobility, see 3.5.6,
“Active Memory Sharing” on page 142.
Support for virtual SCSI devices that are 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 function.

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 3.6, “System Planning Tool” on page 152.
IBM Systems Director agent and several preinstalled IBM Tivoli® agents, such as the
following examples:
– IBM Tivoli Identity Manager, which allows easy integration into an existing Tivoli
Systems Management infrastructure
– IBM Tivoli Application Dependency Discovery Manager (ADDM), which creates and
automatically maintains application infrastructure maps, including dependencies,
change histories, and deep configuration values
vSCSI enterprise reliability, availability, and serviceability (eRAS).
Additional CLI statistics in svmon, vmstat, fcstat, and topas.
The VIOS Performance Advisor tool provides advisory reports based on key performance
metrics for various partition resources that are collected from the VIOS environment.
Monitoring solutions to help manage and monitor the VIOS and shared resources.
Commands and views provide additional metrics for memory, paging, processes, Fibre
Channel HBA statistics, and virtualization.
For more information about the VIOS and its implementation, see IBM PowerVM
Virtualization Introduction and Configuration, SG24-7940
3.5.5 Live Partition Mobility
Live Partition Mobility (LPM) is a technique that allows a partition running on one server to be
migrated dynamically to another server.
This feature can be extremely useful in a situation where a system needs to be evacuated for
maintenance but its partitions do not allow for downtime. LPM allows for all the partitions to be
moved while running to another server so the system can be properly shut down without
impacts for the applications.
In simplified terms, LPM typically works in an environment where all of the I/O from one
partition is virtualized through PowerVM and VIOS and all partition data is stored in a Storage
Area Network (SAN) accessed by both servers.
To migrate a partition from one server to another, a partition is identified on the new server
and configured to have the same virtual resources as the primary server including access to
the same logical volumes as the primary using the SAN.
When an LPM migration is initiated on a server for a partition, PowerVM in the first system
starts copying the state of memory in the first partition over to a destination partition in
another server through each system PowerVM. This is done across a LAN while the initial
partition continues to run. PowerVM has control of I/O operations through I/O virtualization
and keeps track of memory changes that occur throughout the process.
At some point, when all of the memory state is copied from the primary partition, the primary
partition is paused. PowerVM in the second server takes over control of the shared storage
resources and allows the partition now running in that server to resume processing at the
point where the first server left off.
Thinking in terms of using LPM for hardware repairs, if all of the workloads on a server are
migrated by LPM to other servers, then after all have been migrated, the first server could be
turned off to repair components.

LPM can also be used for doing firmware upgrades or adding additional hardware to a server
when the hardware cannot be added concurrently in addition to software maintenance within
individual partitions.
In successful LPM situations, while there may be a short time when applications are not
processing new workload, the applications do not fail or crash and do not need to be
restarted. Roughly speaking then, LPM allows for planned outages to occur on a server
without suffering downtime that would otherwise be required.
Minimum configuration
For LPM to work, it is necessary that the system containing a partition to be migrated, and the
system being migrated to, both have a local LAN connection using a virtualized LAN adapter.
The LAN adapter should be high speed for better migration performance. The LAN used
should be a local network and should be private and have only two uses. The first is for
communication between servers; the second is for communication between partitions on
each server and the HMC for resource monitoring and control functions (RMC.)
LPM also needs all systems in the LPM cluster to be attached to the same SAN (when using
SAN for required common storage), which typically requires use of Fibre Channel adapters.
If a single HMC is used to manage both systems in the cluster, connectivity to the HMC also
needs to be provided by an Ethernet connection to each service processor.
The LAN and SAN adapters used by the partition must be assigned to a Virtual I/O server and
the partitions access to these would be by virtual LAN (VLAN) and virtual SCSI (vSCSI)
connections within each partition to the VIOS.
Each partition to be migrated must only use virtualized I/O through a VIOS; there can be no
non-virtualized adapters assigned to such partitions.
A diagram with the minimum requirements can be seen in Figure 3-7.
Figure 3-7 Minimum Live Partition Mobility requirements
InfrastructureInfrastructure
Computational
Hardware
Virtual I/O
Servers
Fibre
Channel
Adapters
LAN
Adapters
PowerVM Enterprise Edition
Dynamic Logical Partitions
Service Processor
Fibre
Channel
Adapters
LAN
Adapters
Virtual I/O
Servers
Computational
Hardware
Service Processor
HMC
SAN
Switch
LAN
Switch
Shared
Storage

Suggested configuration
LPM connectivity in the minimum configuration discussion is vulnerable to a number of
different hardware and firmware faults that would lead to the inability to migrate partitions.
Multiple paths to networks and SANs are therefore advised. To accomplish this, a VIOS
server can be configured to use dual Fibre Channel and LAN adapters.
Externally to each system, redundant hardware management consoles (HMCS) can be
utilized for greater availability. There can also be options to maintain redundancy in SANs and
local network hardware. A diagram with the suggested scenario can be seen in Figure 3-8.
Figure 3-8 Redundant infrastructure for LPM
PCIe slot selection
POWER8 processor has PCIe controllers integrated on the chip allowing for a single
processor to have two or more PCIe gen3 slots directly attached to it. In case of a complete
processor failure, these slots might become unusable.
When this affects availability of PCIe slots, it must be taken into consideration while selecting
the slot placement for the adapters on the VIrtual I/O Servers.
For scale-out class systems, each processor module directly controls two 16 lane (x16) PCIe
slots and additional I/O capabilities are provided by x8 connections to a PCIe switch
integrated in the processor board. All the slots connected to the PCIe switch belong to a
single processor.
InfrastructureInfrastructure
Computational
Hardware
Virtual I/O
Servers
Fibre
Channel
Adapters
LAN
Adapters
Service Processor
Fibre
Channel
Adapters
LAN
Adapters
Virtual I/O
Servers
Computational
Hardware
Service Processor
HMC
HMC
SAN
Switch
SAN
Switch
LAN
Switch
LAN
Switch

Figure 3-9 illustrates how such a system could be configured to maximize redundancy in a
VIOS environment, presuming that the rootvgs for each VIOS are accessed from storage
area networks.
Figure 3-9 I/O Subsystem of a POWER8 2-socket system
DCM DCM
Physical
Processor Chip Processor Chip Processor Chip Processor Chip
PCle Ctrl PCle Ctrl PCle Ctrl PCle Ctrl
PHB PHB PHB PHB PHB PHB PHB PHB
PCIe
Switch
PCIe
Switch
LAN
Adapter
FC
Adapter
For
SAN
FC
Adapter
For
SAN
LAN
Adapter
LAN
Adapter
LAN
Adapter
FC
Adapter
For
SAN
FC
Adapter
For
SAN
Virtual I/O
Server 1
Virtual I/O
Server 2
Virtual
x16 x16 x16 x16
x8 x8
x8 x8

For enterprise class systems, I/O drawers for expanding I/O are supported. When these
drawers are used, all the slots connected to a PCIe Adapter for Expansion Drawer are bound
to a given processor. A similar concept for I/O redundancy can be used to maximize
availability of I/O access using two I/O drawers, one connected to each processor socket in
the system. A logical diagram with the PCIe slots and its processors can be seen in
Figure 3-10.
Figure 3-10 Logical diagram of processors and its associated PCIe slots
More information
For more information about PowerVM and Live Partition Mobility, see IBM PowerVM
Virtualization Introduction and Configuration, SG24-7940:
http://www.redbooks.ibm.com/abstracts/sg247940.html
3.5.6 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 available only with the Enterprise version of PowerVM.
PCIe Gen3 x16
PCIe Adapter for
Expansion Drawer
PCIe Gen3 x16
PCIe Adapter for
Expansion Drawer
PCIe Gen3 x16
PCIe Adapter for
Expansion Drawer
PCIe Gen3 x16
PCIe Adapter for
Expansion Drawer
System Node
PCIe Expansion Drawer
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
CPU
1
CPU
0
CPU
2
CPU
3
PCIe Expansion Drawer
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8
PCIeGen3x8

The physical memory of an IBM Power System can be assigned to multiple partitions in either
dedicated or shared mode. The system administrator can 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 used to increase memory usage 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 systems. For example,
AIX partitions can take advantage of Active Memory Expansion. Other operating systems
take advantage of Active Memory Sharing also.
For more information regarding Active Memory Sharing, see IBM PowerVM Virtualization
Active Memory Sharing, REDP-4470.
3.5.7 Active Memory Deduplication
In a virtualized environment, the systems might have a considerable amount of duplicated
information that is stored on RAM after each partition has its own operating system, and
some of them might even share the same kinds of applications. On heavily loaded systems,
this behavior might lead to a shortage of the available memory resources, forcing paging by
the Active Memory Sharing partition operating systems, the Active Memory Deduplication
pool, or both, which might decrease overall system performance.
Active Memory Deduplication allows the POWER Hypervisor to map dynamically identical
partition memory pages to a single physical memory page within a shared memory pool. This
way enables a better usage of the Active Memory Sharing shared memory pool, increasing
the system’s overall performance by avoiding paging. Deduplication can cause the hardware
to incur fewer cache misses, which also leads to improved performance.
Active Memory Deduplication depends on the Active Memory Sharing feature being available,
and it consumes CPU cycles that are donated by the Active Memory Sharing pool’s VIOS
partitions to identify deduplicated pages. The operating systems that are running on the
Active Memory Sharing partitions can “hint” to the POWER Hypervisor that some pages
(such as frequently referenced read-only code pages) are good for deduplication.
To perform deduplication, the hypervisor cannot compare every memory page in the Active
Memory Sharing 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, a look-up of its signature is done in 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 Active Memory Sharing pool.
From the LPAR perspective, the Active Memory Deduplication feature is not apparent. If an
LPAR attempts to modify a deduplicated page, the Power hypervisor grabs a free page from
the Active Memory Sharing pool, copies the duplicate page contents into the new page, and
maps the LPAR’s reference to the new page so the LPAR can modify its own unique page.

For more information regarding Active Memory Deduplication, see Power Systems Memory
Deduplication, REDP-4827.
3.5.8 Operating system support for PowerVM
Table 3-3 shows operating system support for virtualization features.
Table 3-3 Virtualization features supported by AIX, IBM i, and Linux
Feature AIX
6.1
TL9
SP1
AIX
7.1
TL03
SP1
IBM i
7.1
TR 9
IBM i
7.2
TR 1
RHEL
6.6
RHEL
7.1
SLES
11
SP3
SLES
12
Ubuntu
15.04
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 Yes Yes Yes Yes Yes Yes Yes Yes Yes
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 Yes Yes Yes Yes Yes Yes Yes Yes Yes
DLPAR processor add/remove Yes Yes Yes Yes Yes Yes Yes Yes Yes
DLPAR memory add Yes Yes Yes Yes Yes Yes Yes Yes Yes
DLPAR memory remove Yes Yes Yes Yes Yes Yes Yes Yes Yes
Micro-Partitioning Yes Yes Yes Yes Yes Yes Yes Yes Yes
Shared dedicated capacity Yes Yes Yes Yes Yes Yes Yes Yes Yes
Multiple Shared Processor Pools Yes Yes Yes Yes Yes Yes Yes Yes Yes
VIOS Yes Yes Yes Yes Yes Yes Yes Yes Yes
Integrated Virtualization
Manager
Yes Yes Yes Yes Yes Yes Yes Yes Yes
Suspend/resume and
hibernationa
a. At the time of writing, Suspend/Resume is not available. Check with your IBM System Services Representative
(SSR) for availability on POWER8 platforms.
Shared Storage Pools Yes Yes Yes Yes Yes Yes Yes Yes Yes
Thin provisioning Yes Yes Yes Yes Yes Yes Yes Yes Yes
Active Memory Sharingb
b. At the time of writing, Active Memory Sharing when used with Live Partition Mobility is not supported. Check with
your IBM SSR for availability on POWER8 platforms.
Active Memory Deduplication Yes Yes Yes Yes Yes Yes Yes Yes Yes
Live Partition Mobility Yes Yes Yes Yes Yes Yes Yes Yes Yes
Simultaneous multithreading
(SMT)
Active Memory Expansion Yes Yes No No No No No No No

For more information about specific features for Linux, see the following website:
http://pic.dhe.ibm.com/infocenter/lnxinfo/v3r0m0/index.jsp?topic=%2Fliaam%2Fsuppor
tedfeaturesforlinuxonpowersystemsservers.htm
3.5.9 Linux support
The IBM Linux Technology Center (LTC) contributes to the development of Linux by providing
support for IBM hardware in Linux distributions. In particular, the LTC has available tools and
code for the Linux communities to take advantage of the POWER8 technology and develop
POWER8 optimized software.
For more information about specific Linux distributions, see the following website:
http://pic.dhe.ibm.com/infocenter/lnxinfo/v3r0m0/index.jsp?topic=%2Fliaam%2Fliaamd
istros.htm
3.5.10 PowerVM simplification
Starting with HMC V8R8.2.0 PowerVM simplification was introduced. Figure 3-11 shows new
options starting from the HMC firmware V8R1 that are available for PowerVM simplification.
Selecting Manage PowerVM opens a new window, where a user can view and manage all
the aspects of a PowerVM configuration, such as SEA, virtual networks, and virtual storage
using the graphical interface only.
Figure 3-11 PowerVM new level tasks

Templates
Templates allow you to specify the configuration details for the system I/O, memory, storage,
network, processor, and other partition resources. A user can use the predefined or captured
templates that are available in the template library to deploy a system. Two types of templates
are available in the template library:
The System Template contains configuration details for system resources, such as system
I/O, memory, storage, processors, network, and physical I/O adapters. You can use these
system templates to deploy the system.
The Partition Template contains configuration details for the partition resources. A user
can configure logical partitions with the predefined templates or by creating a custom
templates.
Using these options, a user can deploy a system, select a System Template, and click
Deploy. After deploying a system, a user can choose a Partition Template from the library.
Performance
The Performance function opens the Performance and Capacity Monitoring window, as
shown in Figure 3-12 and Figure 3-13 on page 147.
Figure 3-12 HMC performance monitoring CPU Memory assignment (top of the window)

Figure 3-13 shows the bottom of the window where performance and capacity data are
presented in a graphical format.
Figure 3-13 HMC performance monitoring CPU Memory assignment (bottom of the window)

New GUI functions
As part of the enhancements introduced in this version, this new capability aids in the
management of PowerVM from a single UI. Figure 3-14 shows how all virtualization
components are now displayed in a single UI.
Figure 3-14 VIO servers, Virtual Network, and Virtual Storage are accessible from the UI

3.5.11 Single view of all Virtual I/O servers
The Manage Virtual I/O Servers function displays a list of VIOS that are configured in the
managed system. It also displays information about each VIOS configuration such as
allocated memory, allocated processing units, allocated virtual processors, and status. This
can be seen in Figure 3-15.
Figure 3-15 Virtual I/O server properties displayed in a single UI
3.5.12 Virtual Storage Management
HMC V8.8.1.0 allows you to manage and monitor storage devices in a PowerVM virtual
storage environment. Is possible to change the configuration of the virtual storage devices
that are allocated to each VIOS on the managed system. Also it lets you add a VIOS to a
shared storage pool cluster and manage all the shared storage pool clusters.

The new UI lets the user view the adapter configuration of the virtual storage devices that are
allocated to the VIOS. The adapter view provides a mapping of the adapters to the physical
storage device. By selecting a VIOS, you can manage the virtual storage devices that are
configured to a particular partition and select and view all the partitions with storage
provisioned by the VIOS. Figure 3-16 shows a single VIOS scenario managing multiple
Logical Partitions and its virtual SCSI devices.
Figure 3-16 Adapter assignment to each partition
3.5.13 Virtual Network Management
HMC V8.8.1.0.1 helps to manage PowerVM virtual networks through a user interface (UI).
This UI uses a defined set of concepts about networking technologies with specific
terminology introduced by IBM Power Architecture®.
As part of Virtual Network Management, new functions are added to perform actions on the
virtual network devices such as these:
Add Virtual Network wizard
Network Bridge Management
Link Aggregation Management
Power VM networking concepts review

Add Virtual Network wizard
Use the Add Virtual Network wizard button in the HMC to add an existing virtual network or
a new virtual network to the server.
The following tasks can be completed by using the Add Virtual Network wizard:
Create internal or bridged networks
Create tagged or untagged virtual networks
Create a virtual network on an existing or a new virtual switch
Create a load group or select an existing load group
Network Bridge Management
From a server that is managed by the HMC, it is possible to change the PowerVM virtual
network bridge properties, the changes allowed are:
Enable or disable network failover in the Failover field.
Enable or disable load balancing in the Load Balance field.
Change the primary Virtual I/O Server (VIOS) and the physical adapter location from the
table.
Enable Jumbo Frame on the network bridge for the virtual Ethernet adapter to
communicate to an external network.
Enable QoS on the network bridge to check the priority value of all tagged packets and
arrange those packets in the corresponding queue.
Link Aggregation Management
A link aggregation device can be added on the VIOS by using the Add Link Aggregation
device wizard. The same wizard can be used to change a Link Aggregation device’s
properties or remove a Link Aggregation device.
PowerVM networking concepts review
PowerVM includes extensive and powerful networking tools and technologies, which can be
used to enable more flexibility, better security, and enhanced usage of hardware resources.
Some of these terms and concepts are unique to the Power Architecture. Table 3-4
introduces the PowerVM virtual networking technologies.
Table 3-4 PowerVM Network technologies
PowerVM technology Definition
Virtual Network Enables interpartition communication without assigning a physical
network adapter to each partition. If the virtual network is bridged,
partitions can communicate with external networks. A virtual
network is identified by its name or VLAN ID and the associated
virtual switch.
Virtual Ethernet adapter Enables a client logical partition to send and receive network traffic
without a physical Ethernet adapter.
Virtual switch An in-memory, hypervisor implementation of a layer-2 switch.
Network bridge A software adapter that bridges physical and virtual networks to
enable communication. A network bridge can be configured for
failover or load sharing.
Link aggregation device A link aggregation (also known as Etherchannel) device is a
network port-aggregation technology that allows several Ethernet
adapters to be aggregated.

For more information regarding PowerVM Simplification Enhancements, see IBM Power
Systems Hardware Management Console Version 8 Release 8.1.0 Enhancements,
SG24-8232.
3.6 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 HMC.
You can create a new system configuration, or you can create a system configuration that is
based on any of the following items:
Performance data from an existing system that the new system replaces
Performance estimates that anticipate future workloads that you must support
Sample systems that you can customize to fit your needs
Integration between the System Planning Tool and both the Workload Estimator and IBM
Performance Management 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.
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 have.
Using the SPT is an effective way of documenting and backing up key system settings and
partition definitions. With it, the user can create records of systems and export them to their
personal workstation or backup system of choice. These same backups can then be imported
back on to the same managed console when needed. This step 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 at the IBM System Planning Tool
website:
http://www.ibm.com/systems/support/tools/systemplanningtool/
Automatically deploy: Ask your IBM SSR or IBM 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 that is
the same as what is specified in your .sysplan file.

3.7 IBM Power Virtualization Center
IBM Power Virtualization Center (IBM PowerVC) is designed to simplify the management of
virtual resources in your Power Systems environment.
After the product code is loaded, the IBM PowerVC no-menus interface guides you through
three simple configuration steps to register physical hosts, storage providers, and network
resources, and start capturing and intelligently deploying your virtual machines (VMs), among
the other tasks shown in the following list:
Create virtual machines and then resize and attach volumes to them.
Import existing virtual machines and volumes so they can be managed by IBM PowerVC.
Monitor the usage of the resources that are in your environment.
Migrate virtual machines while they are running (hot migration).
Deploy images quickly to create virtual machines that meet the demands of your
ever-changing business needs.
IBM PowerVC is built on OpenStack. OpenStack is an open source software that controls
large pools of server, storage, and networking resources throughout a data center.
IBM PowerVC is available in two editions:
IBM Power Virtualization Center Express Edition
IBM Power Virtualization Center Standard Edition
Table 3-5 shows an overview of the key features that are included with IBM PowerVC
Editions.
Table 3-5 IBM PowerVC Editions key features overview.
For more information about IBM PowerVC, see IBM PowerVC Version 1.2 Introduction and
Configuration, SG24-8199.
IBM Power Virtualization Center Express
Edition
IBM Power Virtualization Center Standard
Edition
Supports IBM Power Systems hosts that are
managed by the Integrated Virtualization
Manager (IVM)
Supports storage area networks, local
storage, and a combination in the same
environment
Supports a single VIOS virtual machine on
each host
Supports IBM Power Systems hosts that are
managed by an HMC
Supports storage area networks
Supports multiple VIOSs virtual machines on
each host

3.8 IBM Power Virtualization Performance
IBM Power Virtualization Performance (IBM PowerVP) for Power Systems is a new product
that offers a performance view into an IBM PowerVM virtualized environment running on the
latest firmware of IBM Power Systems. It can show which virtual workloads are using specific
physical resources on an IBM Power Systems server.
IBM PowerVP helps reduce time and complexity to find and display performance bottlenecks
through a simple dashboard that shows the performance health of the system. It can help
simplify both prevention and troubleshooting and thus reduce the cost of performance
management.
It assists you in the following ways:
Shows workloads in real time, which highlights possible problems or bottlenecks
(overcommitted resources)
Helps better use virtualized IBM Power System servers by showing the distribution of
workloads
Can replay saved historical data
Helps with the resolution of performance-related issues
Helps to proactively address future issues that can affect performance
IBM PowerVP is integrated with the POWER Hypervisor and collects performance data
directly from PowerVM Hypervisor, which offers the most accurate performance information
about virtual machines running on IBM Power Systems. This performance information is
displayed on a real-time, continuous GUI dashboard and it is also available for historical
review.
Features of IBM PowerVP include these:
Real-time, continuous graphical monitor (dashboard) that delivers an easy-to-read display
showing the overall performance health of the Power Systems server.
Customizable performance thresholds that enable you to customize the dashboard to
match your monitoring requirements.
Historical statistics that enable you to go back in time and replay performance data
sequences to discover performance bottlenecks.
System-level performance views that show all LPARs (VMs) and how they are using real
system resources.
Virtual machine drill down, which gives you more performance details for each VM,
displaying detailed information about various resources, such as CPU, memory, and disk
activity.
Support for all virtual machine types, including AIX, IBM i, and Linux.
Background data collection, which enables performance data to be collected when the
GUI is not active.
IBM PowerVP even allows an administrator to drill down and view specific adapter, bus, or
CPU usage. An administrator can see the hardware adapters and how much workload is
placed on them. IBM PowerVP provides both an overall and detailed view of IBM Power
System server hardware, so it is easy to see how virtual machines are consuming resources.
For more information about this topic, see the following website:
http://www.ibm.com/systems/power/software/performance/

3.9 VIOS 2.2.3.51 features
The enterprise systems require the VIOS version 2.2.3.51 with APAR IV68443 and IV68444.
This release provides the same functionality as the previous release and is updated to
support the Power E870 and Power E880 and other new hardware.
VIOS 2.2.3.51 features include these:
Simplified Shared Ethernet Adapter Failover configuration setup
Shared Storage Pools enhancements.
With VIOS version 2.2.3.51, or later, the maximum number of virtual I/O slots that are
supported on the AIX, IBM i, and Linux partition is increased up to 32.
VIOS support:
– Supports IBM Power E870 and Power E880
– Supports 8 GB quad port PCIe generation 2 Fibre Channels adapter
– Supports PCIe Gen3 Fibre Channels adapter
There is support for Live Partition Mobility performance enhancements to better use the
10 Gb Ethernet Adapters that are used within the mobility process, and PowerVM server
evacuation function.
Shared Ethernet Adapter by default uses the largesend attribute.
The VIOS update packages can be downloaded from the IBM Fix Central website:
http://www-933.ibm.com/support/fixcentral
3.10 Dynamic Partition Remote Restart
Partition Remote Restart is a function designed to enhance availability of a partition on
another server when its original host server fails. This is a High Availability function of
PowerVM Enterprise Edition.
Starting from IBM Power Systems HMC V8.8.1.0, the requirement of enabling Remote
Restart of an LPAR only at creation time has been removed. Dynamic Partition Remote
Restart allows for the dynamic toggle of Remote Restart capability when an LPAR is
deactivated.

To verify that your managed system can support this capability, enter the following command
as shown in Example 3-1.
Example 3-1 Power VM remote restart capable
hscroot@slcb27a:~>lssyscfg -r sys -m Server1 -F capabilities
"active_lpar_mobility_capable,inactive_lpar_mobility_capable,os400_lpar_mobility_c
apable,active_lpar_share_idle_procs_capable,active_mem_dedup_capable,active_mem_ex
pansion_capable,hardware_active_mem_expansion_capable,active_mem_mirroring_hypervi
sor_capable,active_mem_sharing_capable,autorecovery_power_on_capable,bsr_capable,c
od_mem_capable,cod_proc_capable,custom_mac_addr_capable,dynamic_platform_optimizat
ion_capable,dynamic_platform_optimization_lpar_score_capable,electronic_err_report
ing_capable,firmware_power_saver_capable,hardware_power_saver_capable,hardware_dis
covery_capable,hardware_encryption_capable,hca_capable,huge_page_mem_capable,lpar_
affinity_group_capable,lpar_avail_priority_capable,lpar_proc_compat_mode_capable,l
par_remote_restart_capable,powervm_lpar_remote_restart_capable,lpar_suspend_capabl
e,os400_lpar_suspend_capable,micro_lpar_capable,os400_capable,5250_application_cap
able,os400_net_install_capable,os400_restricted_io_mode_capable,redundant_err_path
_reporting_capable,shared_eth_auto_control_channel_capable,shared_eth_failover_cap
able,sp_failover_capable,sriov_capable,vet_activation_capable,virtual_eth_disable_
capable,virtual_eth_dlpar_capable,virtual_eth_qos_capable,virtual_fc_capable,virtu
al_io_server_capable,virtual_switch_capable,vlan_stat_capable,vtpm_capable,vsi_on_
veth_capable,vsn_phase2_capable"
In the previous example, the highlighted text is indicating that the managed system is capable
of remotely restarting a partition.
From the HMC, select the Managed System Properties → Capabilities tab to display all of
the managed system capabilities as shown in Figure 3-17.
Figure 3-17 PowerVM Partition Remote Restart Capable

The capability is only displayed if the managed system supports it.
To activate a partition on a supported system to support Dynamic Partition Remote Restart,
enter the following command:
chsyscfg -r lpar -m <ManagedSystemName> -i
"name=<PartitionName>,remote_restart_capable=1"
To use the Remote Restart feature, the following conditions need to be met:
The managed system should support toggle partition remote capability.
The partition should be in the inactive state.
The partition type should be AIX, IBM i, or Linux.
The reserved storage device pool exists.
The partition should not own any of the below resources/settings:
– BSR
– Time Reference Partition
– Service Partition
– Opticonnect
– HSL
– Physical I/O
– HEA
– Error Reporting Partition
– Is part of EWLM
– Huge Page Allocation
– Owns Virtual Serial Adapters
– Belongs to I/O Fail Over Pool
– SR-IOV logical port
For further information. including the use of Partition Remote Restart, see the following
website:
http://www.ibm.com/support/knowledgecenter/POWER8/p8hat/p8hat_enadisremres.htm

Chapter 4. Reliability, availability, and
serviceability
This chapter provides information about IBM Power Systems reliability, availability, and
serviceability (RAS) design and features.
The elements of RAS can be described as follows:
Reliability Indicates how infrequently a defect or fault in a server occurs
Availability Indicates how infrequently the functioning of a system or application is
impacted by a fault or defect
Serviceability Indicates how well faults and their effects are communicated to system
managers and how efficiently and nondisruptively the faults are
repaired
4

4.1 Introduction
The POWER8 processor-based servers are available in two different classes:
Scale-out systems: For environments consisting of multiple systems working in concert. In
such environments, application availability is enhanced by the superior availability
characteristics of each system.
Enterprise systems: For environments requiring systems with increased availability. In
such environments, mission-critical applications can take full advantage of the scale-up
characteristics, increased performance, flexibility to upgrade, and enterprise availability
characteristics.
One key differentiator of the IBM POWER8 processor-based servers is that they leverage all
the advanced RAS characteristics of the POWER8 processor through the whole portfolio,
offering reliability and availability features that often are not seen in other scale-out servers.
Some of these features are improvements for POWER8 or features that were found
previously only in higher-end Power Systems.
The POWER8 processor modules support an enterprise level of reliability and availability. The
processor design has extensive error detection and fault isolation (ED/FI) capabilities to allow
for a precise analysis of faults, whether they are hard or soft. They use advanced technology,
including stacked latches and Silicon-on-Insulator (SOI) technology, to reduce susceptibility to
soft errors, and advanced design features within the processor for correction or try again after
soft error events. In addition, the design incorporates spare capacity that is integrated into
many elements to tolerate certain faults without requiring an outage or parts replacement.
Advanced availability techniques are used to mitigate the impact of other faults that are not
directly correctable in the hardware.
Features within the processor and throughout systems are incorporated to support design
verification. During the design and development process, subsystems go through rigorous
verification and integration testing processes by using these features. During system
manufacturing, systems go through a thorough testing process to help ensure high product
quality levels, again taking advantage of the designed ED/FI capabilities.
Fault isolation and recovery of the POWER8 processor and memory subsystems are
designed to use a dedicated service processor and are meant to be largely independent of
any operating system or application deployed.
The Power E870 and Power E880 are enterprise class systems and are designed to support
the highest levels of RAS. More details of the specific features exclusive to enterprise class
systems can be found in 4.4, “Enterprise systems availability details” on page 170.
4.1.1 RAS enhancements of POWER8 processor-based servers
Several features were included in the whole portfolio of the POWER8 processor-based
servers. Some of these features are improvements for POWER8 or features that were found
previously only in higher-end Power Systems, leveraging a higher RAS even for scale-out
equipment.
Here is a brief summary of these features:
Processor Enhancements Integration
POWER8 processor chips are implemented using 22 nm technology and integrated onto
SOI modules.

Chapter 4. Reliability, availability, and serviceability 161
The processor design now supports a spare data lane on each fabric bus, which is used to
communicate between processor modules. A spare data lane can be substituted for a
failing one dynamically during system operation.
A POWER8 processor module has improved performance compared to POWER7+,
including support of a maximum of twelve cores compared to a maximum of eight cores in
POWER7+. Doing more work with less hardware in a system provides greater reliability,
by concentrating the processing power and reducing the need for additional
communication fabrics and components.
The processor module integrates a new On Chip Controller (OCC). This OCC is used to
handle Power Management and Thermal Monitoring without the need for a separate
controller, which was required in POWER7+. In addition, the OCC can also be
programmed to run other RAS-related functions independent of any host processor.
The memory controller within the processor is redesigned. From a RAS standpoint, the
ability to use a replay buffer to recover from soft errors is added.
I/O Subsystem
The POWER8 processor now integrates PCIe controllers. PCIe slots that are directly
driven by PCIe controllers can be used to support I/O adapters directly in the systems or
be used to attach external I/O drawers. For greater I/O capacity, the POWER8
processor-based Power E870 and Power E880 servers also support a PCIe switch to
provide additional integrated I/O capacity.
Memory Subsystem
Custom DIMMs (CDIMMS) are used, which, in addition to the ability to correct a single
DRAM fault within an error-correcting code (ECC) word (and then an additional bit fault) to
avoid unplanned outages, also contain a spare DRAM module per port (per nine DRAMs
for x8 DIMMs), which can be used to avoid replacing memory.
The Power E870 and Power E880 systems have the option of mirroring the memory used
by the hypervisor. This reduces the risk of system outage linked to memory faults, as the
hypervisor memory is stored in two distinct memory CDIMMs.
Power Distribution and Temperature Monitoring
All systems make use of voltage convertors that transform the voltage level provided by
the power supply to the voltage level needed for the various components within the
system. The Power E870 and Power E880 systems contain two convertors for each
voltage level provided to any given processor or memory DIMM.
Convertors used for processor voltage levels are configured for redundancy so that when
one is detected as failing, it will be called out for repair while the system continues to run
with the redundant voltage convertor.
The convertors used for memory are configured with a form of sparing where when a
convertor fails, the system continues operation with another convertor without generating
a service event or the need to take any sort of outage for repair.
The processor module integrates a new On Chip Controller (OCC). This OCC is used to
handle Power Management and Thermal Monitoring without the need for a separate
controller, as was required in POWER7+. In addition, the OCC can also be programmed to
run other RAS-related functions independent of any host processor.
The E880 and E870 systems make use of triple redundant ambient temperature sensors.

4.2 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, and architecture-based
approach is designed to improve overall system reliability with each successive generation of
system offerings. Reliability can be improved in primarily three ways:
Reducing the number of components
Using higher reliability grade parts
Reducing the stress on the components
In the POWER8 systems, elements of all three are used to improve system reliability.
4.2.1 Designed for reliability
Systems that are 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 reduce the opportunity for system failures. The POWER8 chip has more
cores per processor module, and the I/O Hub Controller function is integrated in the
processor module, which generates a PCIe BUS directly from the Processor module. Parts
selection also plays a critical role in overall system reliability.

IBM uses stringent design criteria to select server grade components that are extensively
tested and qualified to meet and exceed a minimum design life of seven years. By selecting
higher reliability grade components, the frequency of all failures is lowered, and wear-out is
not expected within the operating system life. Component failure rates can be further
improved by burning in select components or running the system before shipping it to the
client. This period of high stress removes the weaker components with higher failure rates,
that is, it cuts off the front end of the traditional failure rate bathtub curve (see Figure 4-1).
Figure 4-1 Failure rate bathtub curve
4.2.2 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. Large
decreases in component reliability are directly correlated to relatively small increases in
temperature. All POWER processor-based systems are packaged to ensure adequate
cooling. Critical system components, such as the POWER8 processor chips, are positioned
on the system board so that they receive clear air flow during operation. POWER8 systems
use a premium fan with an extended life to further reduce overall system failure rate and
provide adequate cooling for the critical system components.
FailureRate(FIT)
Power-on Hours
~
~
Early life End-of-life
Average life
of a server
System
run-in
Wafer and chip test
In situ burn-in

4.3 Processor/Memory availability details
The more reliable a system or subsystem is, the more available it should be. Nevertheless,
considerable effort is made to design systems that can detect faults that do occur and take
steps to minimize or eliminate the outages that are associated with them. These design
capabilities extend availability beyond what can be obtained through the underlying reliability
of the hardware.
This design for availability begins with implementing an architecture for ED/FI.
First-Failure Data Capture (FFDC) is the capability of IBM hardware and microcode to
continuously monitor hardware functions. Within the processor and memory subsystem,
detailed monitoring is done by circuits within the hardware components themselves. Fault
information is gathered into fault isolation registers (FIRs) and reported to the appropriate
components for handling.
Processor and memory errors that are recoverable in nature are typically reported to the
dedicated service processor built into each system. The dedicated service processor then
works with the hardware to determine the course of action to be taken for each fault.
4.3.1 Correctable error introduction
Intermittent or soft errors are typically tolerated within the hardware design by using error
correction code or advanced techniques to try operations again after a fault.
Tolerating a correctable solid fault runs the risk that the fault aligns with a soft error and
causes an uncorrectable error situation. There is also the risk that a correctable error is
predictive of a fault that continues to worsen over time, resulting in an uncorrectable error
condition.
You can predictively deallocate a component to prevent correctable errors from aligning with
soft errors or other hardware faults and causing uncorrectable errors to avoid such situations.
However, unconfiguring components, such as processor cores or entire caches in memory,
can reduce the performance or capacity of a system. This in turn typically requires that the
failing hardware is replaced in the system. The resulting service action can also temporarily
impact system availability.
To avoid such situations in solid faults in POWER8, processors or memory might be
candidates for correction by using the “self-healing” features built into the hardware, such as
taking advantage of a spare DRAM module within a memory DIMM, a spare data lane on a
processor or memory bus, or spare capacity within a cache module.
When such self-healing is successful, the need to replace any hardware for a solid
correctable fault is avoided. The ability to predictively unconfigure a processor core is still
available for faults that cannot be repaired by self-healing techniques or because the sparing
or self-healing capacity is exhausted.

4.3.2 Uncorrectable error introduction
An uncorrectable error can be defined as a fault that can cause incorrect instruction execution
within logic functions, or an uncorrectable error in data that is stored in caches, registers, or
other data structures. In less sophisticated designs, a detected uncorrectable error nearly
always results in the termination of an entire system. More advanced system designs in some
cases might be able to terminate just the application by using the hardware that failed. Such
designs might require that uncorrectable errors are detected by the hardware and reported to
software layers, and the software layers must then be responsible for determining how to
minimize the impact of faults.
The advanced RAS features that are built in to POWER8 processor-based systems handle
certain “uncorrectable” errors in ways that minimize the impact of the faults, even keeping an
entire system up and running after experiencing such a failure.
Depending on the fault, such recovery may use the virtualization capabilities of PowerVM in
such a way that the operating system or any applications that are running in the system are
not impacted or must participate in the recovery.
4.3.3 Processor Core/Cache correctable error handling
Layer 2 (L2) and Layer 3 (L3) caches and directories can correct single bit errors and detect
double bit errors (SEC/DED ECC). Soft errors that are detected in the level 1 caches are also
correctable by a try again operation that is handled by the hardware. Internal and external
processor “fabric” busses have SEC/DED ECC protection as well.
SEC/DED capabilities are also included in other data arrays that are not directly visible to
customers.
Beyond soft error correction, the intent of the POWER8 design is to manage a solid
correctable error in an L2 or L3 cache by using techniques to delete a cache line with a
persistent issue, or to repair a column of an L3 cache dynamically by using spare capability.
Information about column and row repair operations is stored persistently for processors, so
that more permanent repairs can be made during processor reinitialization (during system
reboot, or individual Core Power on Reset using the Power On Reset Engine.)
4.3.4 Processor Instruction Retry and other try again techniques
Within the processor core, soft error events might occur that interfere with the various
computation units. When such an event can be detected before a failing instruction is
completed, the processor hardware might be able to try the operation again by using the
advanced RAS feature that is known as Processor Instruction Retry.
Processor Instruction Retry allows the system to recover from soft faults that otherwise result
in an outage of applications or the entire server.
Try again techniques are used in other parts of the system as well. Faults that are detected on
the memory bus that connects processor memory controllers to DIMMs can be tried again. In
POWER8 systems, the memory controller is designed with a replay buffer that allows memory
transactions to be tried again after certain faults internal to the memory controller faults are
detected. This complements the try again abilities of the memory buffer module.

4.3.5 Alternative processor recovery and Partition Availability Priority
If Processor Instruction Retry for a fault within a core occurs multiple times without success,
the fault is considered to be a solid failure. In some instances, PowerVM can work with the
processor hardware to migrate a workload running on the failing processor to a spare or
alternative processor. This migration is accomplished by migrating the pertinent processor
core state from one core to another with the new core taking over at the instruction that failed
on the faulty core. Successful migration keeps the application running during the migration
without needing to terminate the failing application.
Successful migration requires that there is sufficient spare capacity that is available to reduce
the overall processing capacity within the system by one processor core. Typically, in highly
virtualized environments, the requirements of partitions can be reduced to accomplish this
task without any further impact to running applications.
In systems without sufficient reserve capacity, it might be necessary to terminate at least one
partition to free resources for the migration. In advance, PowerVM users can identify which
partitions have the highest priority and which do not. When you use this Partition Priority
feature of PowerVM, if a partition must be terminated for alternative processor recovery to
complete, the system can terminate lower priority partitions to keep the higher priority
partitions up and running, even when an unrecoverable error occurred on a core running the
highest priority workload.
Partition Availability Priority is assigned to partitions by using a weight value or integer rating.
The lowest priority partition is rated at 0 (zero) and the highest priority partition is rated at
255. The default value is set to 127 for standard partitions and 192 for Virtual I/O Server
(VIOS) partitions. Priorities can be modified through the Hardware Management Console
(HMC).
4.3.6 Core Contained Checkstops and other PowerVM error recovery
PowerVM can handle certain other hardware faults without terminating applications, such as
an error in certain data structures (faults in translation tables or lookaside buffers).
Other core hardware faults that alternative processor recovery or Processor Instruction Retry
cannot contain might be handled in PowerVM by a technique called Core Contained
Checkstops. This technique allows PowerVM to be signaled when such faults occur and
terminate code in use by the failing processor core (typically just a partition, although
potentially PowerVM itself if the failing instruction were in a critical area of PowerVM code).
Processor designs without Processor Instruction Retry typically must resort to such
techniques for all faults that can be contained to an instruction in a processor core.
4.3.7 Cache uncorrectable error handling
If a fault within a cache occurs that cannot be corrected with SEC/DED ECC, the faulty cache
element is unconfigured from the system. Typically, this is done by purging and deleting a
single cache line. Such purge and delete operations are contained within the hardware itself,
and prevent a faulty cache line from being reused and causing multiple errors.
During the cache purge operation, the data that is stored in the cache line is corrected where
possible. If correction is not possible, the associated cache line is marked with a special ECC
code that indicates that the cache line itself has bad data.

Nothing within the system terminates just because such an event is encountered. Rather, the
hardware monitors the usage of pages with marks. If such data is never used, hardware
replacement is requested, but nothing terminates as a result of the operation. Software layers
are not required to handle such faults.
Only when data is loaded to be processed by a processor core, or sent out to an I/O adapter,
is any further action needed. In such cases, if data is used as owned by a partition, then the
partition operating system might be responsible for terminating itself or just the program using
the marked page. If data is owned by the hypervisor, then the hypervisor might choose to
terminate, resulting in a system-wide outage.
However, the exposure to such events is minimized because cache-lines can be deleted,
which eliminates repetition of an uncorrectable fault that is in a particular cache-line.
4.3.8 Other processor chip functions
Within a processor chip, there are other functions besides just processor cores.
POWER8 processors have built-in accelerators that can be used as application resources to
handle such functions as random number generation. POWER8 also introduces a controller
for attaching cache-coherent adapters that are external to the processor module. The
POWER8 design contains a function to “freeze” the function that is associated with some of
these elements, without taking a system-wide checkstop. Depending on the code using these
features, a “freeze” event might be handled without an application or partition outage.
As indicated elsewhere, single bit errors, even solid faults, within internal or external
processor “fabric busses”, are corrected by the error correction code that is used. POWER8
processor-to-processor module fabric busses also use a spare data-lane so that a single
failure can be repaired without calling for the replacement of hardware.
4.3.9 Other fault error handling
Not all processor module faults can be corrected by these techniques. Therefore, a provision
is still made for some faults that cause a system-wide outage. In such a “platform” checkstop
event, the ED/FI capabilities that are built in to the hardware and dedicated service processor
work to isolate the root cause of the checkstop and unconfigure the faulty element were
possible so that the system can reboot with the failed component unconfigured from the
system.
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.
The auto-restart (reboot) option must be enabled from the Advanced System Management
Interface (ASMI) or from the Control (Operator) Panel.

4.3.10 Memory protection
POWER8 processor-based systems have a three-part memory subsystem design. This
design consists of two memory controllers in each processor module, which communicate to
buffer modules on memory DIMMS through memory channels and access the DRAM
memory modules on DIMMs, as shown in Figure 4-2.
Figure 4-2 Memory protection features
The memory buffer chip is made by the same 22 nm technology that is used to make the
POWER8 processor chip, and the memory buffer chip incorporates the same features in the
technology to avoid soft errors. It implements a try again for many internally detected faults.
This function complements a replay buffer in the memory controller in the processor, which
also handles internally detected soft errors.
The bus between a processor memory controller and a DIMM uses CRC error detection that
is coupled with the ability to try soft errors again. The bus features dynamic recalibration
capabilities plus a spare data lane that can be substituted for a failing bus lane through the
recalibration process.
The buffer module implements an integrated L4 cache using eDRAM technology (with soft
error hardening) and persistent error handling features.
The memory buffer on each DIMM has four ports for communicating with DRAM modules.
The 16 GB DIMM, for example, has one rank that is composed of four ports of x8 DRAM
modules, each port containing 10 DRAM modules.
Memory Controller
 Supports 128 Byte Cache Line
 Hardened “Stacked” Latches for Soft Error Protection
 And reply buffer to retry after soft internal faults
 Special Uncorrectable error handling for solid faults
Memory Bus
 CRC protection with recalibration and retry on error
 Spare Data lane can be dynamically substituted for
failed one
Memory Buffer
 Same technology as POWER8 Processor Chips
– Hardened “Stacked” Latches for Soft Error Protection
 Can retry after internal soft Errors
 L4 Cache implemented in eDRAM
– DED/SEC ECC Code
– Persistent correctable error handling
16 GB DIMM
 4 Ports of Memory
– 10 DRAMs x8 DRAM modules attached to each
port
– 8 Modules Needed For Data
– 1 Needed For Error Correction Coding
– 1 Additional Spare
 2 Ports are combined to form a 128 bit ECC
word
– 8 Reads fill a processor cache
 Second port can be used to fill a second
cache line
– (Much like having 2 DIMMs under one Memory
buffer but housed in the same physical DIMM)

For each such port, there are eight DRAM modules worth of data (64 bits) plus another
DRAM module’s worth of error correction and other such data. There is also a spare DRAM
module for each port that can be substituted for a failing port.
Two ports are combined into an ECC word and supply 128 bits of data. The ECC that is
deployed can correct the result of an entire DRAM module that is faulty. This is also known as
Chipkill correction. Then, it can correct at least an additional bit within the ECC word.
The additional spare DRAM modules are used so that when a DIMM experiences a Chipkill
event within the DRAM modules under a port, the spare DRAM module can be substituted for
a failing module, avoiding the need to replace the DIMM for a single Chipkill event.
Depending on how DRAM modules fail, it might be possible to tolerate up to four DRAM
modules failing on a single DIMM without needing to replace the DIMM, and then still correct
an additional DRAM module that is failing within the DIMM.
There are other DIMMs offered with these systems. A 32 GB DIMM has two ranks, where
each rank is similar to the 16 GB DIMM with DRAM modules on four ports, and each port has
ten x8 DRAM modules.
In addition, there is a 64 GB DIMM that is offered through x4 DRAM modules that are
organized in four ranks.
In addition to the protection that is provided by the ECC and sparing capabilities, the memory
subsystem also implements scrubbing of memory to identify and correct single bit soft-errors.
Hypervisors are informed of incidents of single-cell persistent (hard) faults for deallocation of
associated pages. However, because of the ECC and sparing capabilities that are used, such
memory page deallocation is not relied upon for repair of faulty hardware,
Finally, should an uncorrectable error in data be encountered, the memory that is impacted is
marked with a special uncorrectable error code and handled as described for cache
uncorrectable errors.
4.3.11 I/O subsystem availability and Enhanced Error Handling
Usage of multi-path I/O and VIOS for I/O adapters and RAID for storage devices should be
used to prevent application outages when I/O adapter faults occur.
To permit soft or intermittent faults to be recovered without failover to an alternative device or
I/O path, Power Systems hardware supports Enhanced Error Handling (EEH) for I/O
adapters and PCIe bus faults.
EEH allows EEH-aware device drivers to try again after certain non-fatal I/O events to avoid
failover, especially in cases where a soft error is encountered. EEH also allows device drivers
to terminate if there is an intermittent hard error or other unrecoverable errors, while
protecting against reliance on data that cannot be corrected. This action typically is done by
“freezing” access to the I/O subsystem with the fault. Freezing prevents data from flowing to
and from an I/O adapter and causes the hardware/firmware to respond with a defined error
signature whenever an attempt is made to access the device. If necessary, a special
uncorrectable error code may be used to mark a section of data as bad when the freeze is
first initiated.

In POWER8 processor-based systems, the external I/O hub and bridge adapters were
eliminated in favor of a topology that integrates PCIe Host Bridges into the processor module
itself. PCIe busses that are generated directly from a host bridge may drive individual I/O slots
or a PCIe switch. The integrated PCIe controller supports try again (end-point error recovery)
and freezing.
IBM device drivers under AIX are fully EEH-capable. For Linux under PowerVM, EEH support
extends to many frequently used devices. There might be various third-party PCI devices that
do not provide native EEH support.
4.4 Enterprise systems availability details
Besides all the standard RAS features described, Enterprise class systems allow for
increased RAS and availability by including several features and redundant components.
Here is a list of the main features exclusive to the Enterprise class systems:
Redundant Service Processor
The service processor is an essential component of a system, responsible for is the initial
power load (IPL), setup, monitoring, control and management. The control units, present
on enterprise class systems house two redundant service processors. In case of a failure
in either of the service processors, the second one allows for continued operation of the
system until a replacement is scheduled. Even a system with a single system node had
dual service processors in the system control unit.
Redundant System Clock Cards
Another component crucial to the system operations is the system clock cards. they are
responsible to providing synchronized clock signals for the whole system. The control
units, present on enterprise class systems house two redundant system clock cards. In
case of a failure in any of the clock cards, the second one allows for continued operation of
the system until a replacement is scheduled. Even a system with a single system node
had dual clock cards on the system control unit.
Dynamic Processor Sparing
Enterprise class systems are Capacity Upgrade on Demand capable. Processor sparing
helps minimize the impact to server performance caused by a failed processor. An inactive
processor is activated if a failing processor reaches a predetermined error threshold, thus
helping to maintain performance and improve system availability. Dynamic processor
sparing happens dynamically and automatically when using dynamic logical partitioning
(DLPAR) and the failing processor is detected prior to failure. Dynamic processor sparing
does not require the purchase of an activation code; it requires only that the system have
inactive CUoD processor cores available.
Dynamic Memory Sparing
Enterprise class systems are Capacity Upgrade on Demand capable. Dynamic memory
sparing helps minimize the impact to server performance caused by a failed memory
feature. Memory sparing occurs when on-demand inactive memory is automatically
activated by the system to temporarily replace failed memory until a service action can be
performed.

Active Memory Mirroring for Hypervisor
The hypervisor is the core part of the virtualization layer. Although minimal, its operational
data must reside in memory CDIMMs. In case of a failure of CDIMM the hypervisor could
become inoperative. The Active memory mirroring for hypervisor allows for the memory
blocks used by the hypervisor to be written in two distinct CDIMMs. If an uncorrectable
error is encountered during a read the data is retrieved from the mirrored pair and
operations continue normally.
4.5 Availability impacts of a solution architecture
Any given solution should not rely only on the hardware platform. Despite IBM Power
Systems being far superior RAS than other comparable systems, it is advisable to design a
redundant architecture surrounding the application in order to allow for easier maintenance
tasks and grater flexibility.
By working in a redundant architecture, some tasks that would require that a given application
would be brought offline, can now be execute with the application running, allowing for even
greater availability.
When determining a highly available architecture that fits your needs, the following topics are
worth considering:
Will I need to move my workloads off of an entire server during service or planned
outages?
If I use a clustering solution to move the workloads, how will the failover time affect my
services?
If I use a server evacuation solution to move the workloads, how long will it take to migrate
all the partitions with my current server configuration?
4.5.1 Clustering
IBM Power Systems running under PowerVM and IBM i, AIX and Linux support a spectrum of
clustering solutions. These solutions are designed to meet requirements not only for
application availability in regard to server outages, but also data center disaster management,
reliable data backups, and so forth. These offerings include distributed applications with IBM
DB2® PureScale, HA solutions using clustering technology with IBM PowerHA
SystemMirror®, and disaster management across geographies with PowerHA SystemMirror
Enterprise Edition.
For more information, see the following references:
PowerHA SystemMirror for IBM i Cookbook, SG24-7994:
Guide to IBM PowerHA SystemMirror for AIX Version 7.1.3, SG24-8167
IBM PowerHA SystemMirror for AIX Cookbook, SG24-7739

4.5.2 Virtual I/O redundancy configurations
Within each server, the partitions can be supported with a single VIOS. However, if a single
VIOS is used and that VIOS terminates for any reason (hardware or software caused), then
all the partitions using that VIOS will terminate.
Using Redundant VIOS servers would mitigate that risk. Maintaining redundancy of adapters
within each VIOS, in addition to having redundant VIOS, will avoid most faults that keep a
VIOS from running. Multiple paths to networks and SANs is therefore advised. Figure 4-3
shows a diagram of a partition accessing data from two distinct Virtual I/O Servers, each one
with multiple network and SAN adapters to provide connectivity.
Figure 4-3 Partition with dual redundant Virtual I/O Servers
Since each VIOS can largely be considered as an AIX based partition, each VIOS also needs
the ability to access a boot image, having paging space, and so forth under a root volume
group or rootvg. The rootvg can be accessed through a SAN, the same as the data that
partitions use. Alternatively, a VIOS can use storage locally attached to a server, either DASD
devices or SSD drives. For best availability, however accessed, the rootvgs should use
mirrored or RAID drives with redundant access to the devices.
4.5.3 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
and 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.
Live Partition Mobility provides systems management flexibility and improves system
availability through the following functions:
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 for maintenance because you can use it to work around scheduled
maintenance activities.
Virtual I/O
#1
Logical Partition
LANAdapter
LANAdapter
SANAdapter
SANAdapter
Virtual
LAN
Virtual
SAN
Virtual I/O
#2
LANAdapter
LANAdapter
SANAdapter
SANAdapter
Virtual
LAN
Virtual
SAN

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 that run on several small, under utilized
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 live partition
mobility, you can manage workloads with minimal downtime.
With the Server Evacuation feature, multiple migrations can occur based on the concurrency
setting of the HMC. Migrations are performed as sets, with the next set of migrations starting
when the previous set completes. Any upgrade or maintenance operations can be performed
after all the partitions are migrated and the source system is powered off.
You can migrate all the migration-capable AIX, IBM i and Linux partitions from the source
server to the destination server by running the following command from the HMC command
line:
migrlpar -o m -m source_server -t target_server --all
Hardware and operating system requirements for Live Partition Mobility
Live Partition Mobility is supported by default with enterprise systems, and it is supported in
compliance with all operating systems that are compatible with POWER8 technology.
The VIOS partition itself cannot be migrated.
For more information about Live Partition Mobility and how to implement it, see IBM PowerVM
Live Partition Mobility (Obsolete - See Abstract for Information), SG24-7460.
4.6 Serviceability
The purpose of serviceability is to repair the system while attempting to minimize or eliminate
service cost (within budget objectives) and maintaining application availability and high
customer satisfaction. Serviceability includes system installation, miscellaneous equipment
specification (MES) (system upgrades/downgrades), and system maintenance/repair.
Depending on the system and warranty contract, service might be performed by the
customer, an IBM System Services Representative (SSR), or an authorized warranty service
provider.
The serviceability features that are delivered in this system provide a highly efficient service
environment by incorporating the following attributes:
Design for SSR Set Up and Customer Installed Features (CIF).
Server Evacuation: This PowerVM function allows you to perform a server evacuation
operation. Server Evacuation is used to move all migration-capable LPARs from one
system to another if there are no active migrations in progress on the source or the target
servers.

Detection and Fault Isolation (ED/FI).
First Failure Data Capture (FFDC).
Guiding Light service indicator architecture is used to control a system of integrated LEDs
that lead the individual servicing the machine to the correct part as quickly as possible.
Service labels, service cards, and service diagrams available on the system and delivered
through the HMC.
Step-by-step service procedures available through the HMC.
This section provides an overview of how these attributes contribute to efficient service in the
progressive steps of error detection, analysis, reporting, notification, and repair found in all
POWER processor-based systems.
4.6.1 Detecting errors
The first and most crucial component of a solid serviceability strategy is the ability to detect
accurately and effectively 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 has no opportunity to evaluate and act if necessary.
POWER processor-based systems employ IBM z Systems® server-inspired error detection
mechanisms, extending from processor cores and memory to power supplies and hard disk
drives (HDDs).
4.6.2 Error checkers, fault isolation registers, and First-Failure Data Capture
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 that is coupled with Processor Instruction Retry and bus try again, to ECC
correction on caches and system buses.
Within the processor/memory subsystem error-checker, error-checker signals are captured
and stored in hardware 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 diagnostic tests
can be deterministic so that for every check station, the unique error domain for that checker
is defined and mapped to field-replaceable units (FRUs) that can be repaired when
necessary.
Integral to the Power Systems design is the concept of FFDC. FFDC is a technique that
involves sufficient error checking stations and co-ordination of faults so that faults are
detected and the root cause of the fault is isolated. FFDC also expects that necessary fault
information can be collected at the time of failure without needing 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 isolated at
the time of the failure without intervention by an IBM SSR. For all faults, good FFDC design
still makes failure information available to the IBM SSR, and this information can be used to
confirm the automatic diagnosis. More detailed information can be collected by an IBM SSR
for rare cases where the automatic diagnosis is not adequate for fault isolation.

4.6.3 Service processor
In POWER8 processor-based systems with a dedicated service processor, the dedicated
service processor is primarily responsible for fault analysis of processor/memory errors.
The service processor is a microprocessor that is powered separately from the main
instruction processing complex.
In addition to FFDC functions, the service processor performs many serviceability functions:
Several remote power control options
Reset and boot features
Environmental monitoring
The service processor interfaces with the OCC function, which monitors the server’s
built-in temperature sensors and sends instructions to the system fans to increase
rotational speed when the ambient temperature is above the normal operating range. By
using a designed 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 shut down a system in the following circumstances:
• The 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.
POWER Hypervisor (system firmware) and HMC connection surveillance.
The service processor monitors the operation of the firmware during the boot process, and
also monitors the hypervisor for termination. The hypervisor monitors the service
processor and can perform a reset and reload if it detects the loss of the service
processor. If the reset/reload operation does not correct the problem with the service
processor, the hypervisor notifies the operating system, and then the operating system
can then take appropriate action, including calling for service. The FSP also monitors the
connection to the HMC and can report loss of connectivity to the operating system
partitions for system administrator notification.
Uncorrectable error recovery
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.
The auto-restart (reboot) option must be enabled from the ASMI or from the Control
(Operator) Panel.
Concurrent access to the service processors menus of the ASMI
This access allows nondisruptive abilities to change system default parameters,
interrogate service processor progress and error logs, and set and reset service indicators

(Light Path for low-end servers), and access all service processor functions without having
to power down the system to the standby state. The administrator or IBM SSR dynamically
can access the menus from any web browser-enabled console that is attached to the
Ethernet service network, concurrently with normal system operation. Some options, such
as changing the hypervisor type, do not take effect until the next boot.
Management of the interfaces for connecting uninterruptible power source systems to the
POWER processor-based systems and performing timed power-on (TPO) sequences.
4.6.4 Diagnosing
General diagnostic objectives are to detect and identify problems so that they can be resolved
quickly. The IBM diagnostic strategy includes the following elements:
Provide a common error code format that is equivalent to a system reference code,
system reference number, checkpoint, or firmware error code.
Provide fault detection and problem isolation procedures. Support a remote connection
ability that is used by the IBM Remote Support Center or IBM Designated Service.
Provide interactive intelligence within the diagnostic tests with detailed online failure
information while connected to IBM back-end system.
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.
Because of the FFDC technology that is designed in to IBM servers, re-creating diagnostic
tests for failures or requiring user intervention is not necessary. Solid and intermittent errors
are designed to be correctly detected and isolated at the time that the failure occurs. Runtime
and boot time diagnostic tests fall into this category.
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 diagnostic tests that are supplemented with system firmware
initialization and configuration of I/O hardware, followed by OS-initiated software test routines.
To minimize boot time, the system determines which of the diagnostic tests are required to be
started to ensure correct operation, which is based on the way that the system was powered
off, or on the boot-time selection menu.
Host Boot IPL
In POWER8, the initialization process during IPL changed. The Flexible Service Processor
(FSP) is no longer the only instance that initializes and runs the boot process. With POWER8,
the FSP initializes the boot processes, but on the POWER8 processor itself, one part of the
firmware is running and performing the Central Electronics Complex chip initialization. A new
component that is called the PNOR chip stores the Host Boot firmware and the Self Boot
Engine (SBE) is an internal part of the POWER8 chip itself and is used to boot the chip.
With this Host Boot initialization, new progress codes are available. An example of an FSP
progress code is C1009003. During the Host Boot IPL, progress codes, such as CC009344,
appear.
If there is a failure during the Host Boot process, a new Host Boot System Dump is collected
and stored. This type of memory dump includes Host Boot memory and is off-loaded to the
HMC when it is available.

Run time
All Power Systems servers can monitor critical system components during run time, and they
can take corrective actions when recoverable faults occur. The IBM hardware error-check
architecture can report non-critical errors in the Central Electronics Complex 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 were 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 by using logic that
is 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 the following items:
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, diagnostic tests are best performed by operating system-specific drivers,
most notably adapters or I/O devices that are owned directly by a logical partition. In these
cases, the operating system device driver often works 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. In non-HMC managed servers, the OS can start the Call Home
application to report the service event to IBM. For optional HMC managed servers, the event
is reported to the HMC, which can initiate the Call Home request to IBM. I/O devices can also
include specific exercisers that can be started by the diagnostic facilities for problem
recreation (if required by service procedures).
4.6.5 Reporting
In the unlikely event that a system hardware or environmentally induced failure is diagnosed,
IBM Power Systems servers report the error through various 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.
Using the Call Home infrastructure, the system automatically can send an alert through a
phone line to a pager, or call for service if there is a critical system failure. A hardware fault
also illuminates the amber system fault LED, which is on the system unit, to alert the user of
an internal hardware problem.
On POWER8 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 can notify the system administrator that it isolated a likely cause of
the system problem. The service processor event log also records unrecoverable checkstop
conditions, forwards them to the SFP application, and notifies the system administrator.
After the information is logged in the SFP application, if the system is correctly 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 also

contains the client contact information as defined in the IBM Electronic Service Agent (ESA)
guided setup wizard. With the new HMC V8R8.1.0 a Serviceable Event Manager is available
to block problems from being automatically transferred to IBM. For more information about
this topic, see “Service Event Manager” on page 193 for more details.
Error logging and analysis
When the root cause of an error is identified by a fault isolation component, an error log entry
is created with basic data, such as the following examples:
An error code that uniquely describes 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 that contains information about the effect that the repair has 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, a call is placed with service and support, and a notification is sent to the contact
that is defined in the ESA-guided setup wizard.
Remote support
The Remote Management and Control (RMC) subsystem is delivered as part of the base
operating system, including the operating system that runs on the HMC. RMC provides a
secure transport mechanism across the LAN interface between the operating system and the
optional HMC and is used by the operating system diagnostic application for transmitting error
information. It performs several other functions, but they are not used for the service
infrastructure.
Service Focal Point application for partitioned systems
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 be reported only once, regardless
of how many logical partitions experience the potential effect of the error. The SFP application
on the management console or in the Integrated Virtualization Manager (IVM) is responsible
for aggregating duplicate error reports, and ensures that all errors are recorded for review and
management. The SFP application provides other service-related functions, such as
controlling service indicators, setting up Call Home, and providing guided maintenance.
When a local or globally reported service request is made to the operating system, the
operating system diagnostic subsystem uses the RMC subsystem to relay error information to
the optional HMC. For global events (platform unrecoverable errors, for example), the service
processor also forwards error notification of these events to the HMC, providing a redundant
error-reporting path in case there are errors in the RMC subsystem 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 or from the service processor. It then looks at all active
service event requests within a predefined timespan, analyzes the failure to ascertain the root
cause and, if enabled, initiates a Call Home for service. This methodology ensures that all
platform errors are reported through at least one functional path, ultimately resulting in a
single notification for a single problem. Similar service functionality is provided through the

SFP application on the IVM for providing service functions and interfaces on non-HMC
partitioned servers.
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 that is collected depends on the invocation
method, but includes information such as 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 either to assist the service
support organization with preparing a service action plan for the IBM SSR or for additional
analysis.
System dump handling
In certain circumstances, an error might require a memory dump to be automatically or
manually created. In this event, the memory dump may be off-loaded to the optional HMC.
Specific management console information is included as part of the information that optionally
can be sent to IBM Support for analysis. If additional information that relates to the memory
dump is required, or if viewing the memory dump remotely becomes necessary, the
management console memory dump record notifies the IBM Support center regarding on
which managements console the memory dump is located. If no management console is
present, the memory dump might be either on the FSP or in the operating system, depending
on the type of memory dump that was initiated and whether the operating system is
operational.
4.6.6 Notifying
After a Power Systems server detects, diagnoses, and reports 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
client notification might 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 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. These events include the following examples:
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 because they indicate that something happened
that requires client awareness if the client wants to take further action. These events can be
reported to IBM at the discretion of the client.
Call Home
Call Home refers to an automatic or manual call from a customer location to an IBM Support
structure with error log data, server status, or other service-related information. The Call

Home feature starts the service organization so that the appropriate service action can begin.
Call Home can be done through HMC or most non-HMC managed systems.
Although configuring a Call Home function is optional, clients are encouraged to implement
this feature to obtain service enhancements, such as reduced problem determination and
faster and potentially more accurate transmission of error information. In general, using the
Call Home feature can result in increased system availability. The ESA application can be
configured for automated Call Home. For more information, see 4.7.4, “Electronic Services
and Electronic Service Agent” on page 191.
Vital product data 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 IBM SSRs, enabling the IBM SSRs to assist in keeping the firmware and software current
on the server.
IBM Service and Support 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 that are taken by the IBM SSR, is recorded for problem
management by the support and development organizations. The problem is then tracked
and monitored until the system fault is repaired.
4.6.7 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)
Terracotta-colored touch points indicate that a component (FRU or CRU) can be
concurrently maintained.
Blue-colored touch points delineate components that may not be concurrently maintained
(they might require that the system is 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 require basic tools such as flathead screw drivers, to service the hardware
components.
Positive retention
Positive retention mechanisms help 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 risk become loose during
shipping or installation, which prevents 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 function is for scale-out systems that can be repaired by clients. In the
Light Path LED implementation, when a fault condition is detected on the POWER8
processor-based system, an amber FRU fault LED is illuminated (turned on solid), which is
then rolled up to the system fault LED. The Light Path system pinpoints the exact part by
lighting the amber FRU fault LED that is associated with the part that must be replaced.
The service person can clearly identify components for replacement by using specific
component level identify LEDs, and can also guide the IBM SSR directly to the component by
signaling (flashing) the FRU component identify LED, and rolling up to the blue enclosure
Locate LED.
After the repair, the LEDs shut off automatically when the problem is fixed. The Light Path
LEDs are only visible while system is in standby power. There are two gold caps
implemented. The gold cap is used to illuminate the amber LEDs once power is removed from
the system. One is inside the drawer, to identify DIMMs, processors and VRMs and one is in
the RAID assembly.
IBM KnowledgeCenter
IBM Knowledge Center provides you with a single information center where you can access
product documentation for IBM systems hardware, operating systems, and server software.
The latest version of the documentation is accessible through the Internet; however, a
CD-ROM based version is also available.
The purpose of KnowledgeCenter, in addition to providing client related product information,
is to provide softcopy information to diagnose and fix any problems that might occur with the
system. Because the information is electronically maintained, changes due to updates or
addition of new capabilities can be used by service representatives immediately.
The IBM KnowledgeCenter can be found online at:
http://www.ibm.com/support/knowledgecenter/
Service labels
Service providers use these labels to assist with maintenance actions. Service labels are in
various formats and positions, and are intended to transmit readily available information to the
service representative during the repair process.
Several of these service labels and their purposes are described in the following list:
Location diagrams are strategically positioned on the system hardware and relate
information about 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, LEDs,
and power supplies.
Remove or replace procedure labels contain procedures that are often found on a cover of
the system or in other locations that are accessible to the service representative. 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 an LCD display (two rows by
16 elements) 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 an IBM SSR or client to change various boot-time options
and for other limited service functions.
Concurrent maintenance
The IBM POWER8 processor-based systems are designed with the understanding that
certain components have higher intrinsic failure rates than others. These components can
include fans, power supplies, and physical storage devices. Other devices, such as I/O
adapters, can begin to wear from repeated plugging and unplugging. For these reasons,
these devices are designed to be concurrently maintainable when properly configured.
Concurrent maintenance is facilitated by the redundant design for the power supplies, fans,
and physical storage.
In addition to the previously mentioned components, the operator panel can be replaced
concurrently by using service functions of the ASMI menu.
Repair and verify services
Repair and verify (R&V) services are automated service procedures that are used to guide a
service provider, step-by-step, through the process of repairing a system and verifying that
the problem was repaired. The steps are customized in the appropriate sequence for the
particular repair for the specific system being serviced. The following scenarios are covered
by R&V services:
Replacing a defective FRU or a 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
R&V procedures can be used by both end-user engineers and IBM SSR 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 R&V 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.
Clients can subscribe through the subscription services on the IBM Support Portal to obtain
notifications about the latest updates that are available for service-related documentation.
4.7 Manageability
Several functions and tools help manageability so you can efficiently and effectively manage
your system.

4.7.1 Service user interfaces
The service interface allows support personnel or the client to communicate with the service
support applications in a server by 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 that are 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. Here are the primary service interfaces:
Light Path (See “Light Path” on page 181 and “Service labels” on page 181.)
Service processor and ASMI
Operator panel
Operating system service menu
SFP on the HMC
SFP Lite on IVM
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 the following 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 ASMI Secure Sockets
Layer (SSL) network connections. The service processor can view and manage the
machine-wide settings by using the ASMI, and enables complete system and partition
management from the HMC.
The service processor uses two Ethernet ports that run at 1 Gbps speed. Consider 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. 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 (HMC1 port) is configured as 169.254.2.147.
– Service processor eth1 (HMC2 port) is configured as 169.254.3.147.
Analyzing a system that does not boot: The FSP can analyze a system that does not
boot. Reference codes and detailed data is available in the ASMI and are transferred to the
HMC.

The following functions are available through the service processor:
Call Home
ASMI
Error information (error code, part number, and location codes) menu
View of guarded components
Limited repair procedures
Generate dump
LED Management menu
Remote view of ASMI menus
Firmware update through a USB key
Advanced System Management Interface
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 VPD. 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 available only 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 unwanted IP addresses or ranges.
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:
Use a 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 ASMI from a management console, complete the following steps:
a. Open Systems Management from the navigation pane.
b. From the work window, select one of the managed systems.
c. From the System Management tasks list, click Operations → Launch Advanced
System Management (ASM).
Use a web browser.
At the time of writing, supported web browsers are Microsoft Internet Explorer (Version
10.0.9200.16439), Mozilla Firefox ESR (Version 24), and Chrome (Version 30). Later
versions of these browsers might work, but are not officially supported. The JavaScript
language and cookies must be enabled and TLS 1.2 might need to 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 an SSL web
connection to the service processor. To establish an SSL connection, open your browser
by using the following address:
https://<ip_address_of_service_processor>
Note: To make the connection through Internet Explorer, click Tools Internet Options.
Clear the Use TLS 1.0 check box, and click OK.

Use 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.
Command-line start of the ASMI
Either on the HMC itself or when properly configured on a remote system, it is possible to
start ASMI web interface from the HMC command line. Open a terminal window on the
HMC or access the HMC with a terminal emulation and run the following command:
asmmenu --ip <ip address>
On the HMC itself, a browser window opens automatically with the ASMI window and,
when configured properly, a browser window opens on a remote system when issued from
there.
The operator panel
The service processor provides an interface to the operator panel, which is used to display
system status and diagnostic information. 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
Here are several of the operator panel features:
A 2 x 16 character LCD display
Reset, enter, power On/Off, increment, and decrement buttons
Amber System Information/Attention, and a green Power LED
Blue Enclosure Identify LED
Altitude sensor
USB Port
Speaker/Beeper
The following functions are available through the operator panel:
Error information
Generate dump
View machine type, model, and serial number
Limited set of repair functions
Operating system service menu
The system diagnostic tests consist of IBM i service tools, stand-alone diagnostic tests that
are loaded from the DVD drive, and online diagnostic tests (available in AIX).
Online diagnostic tests, 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 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.
The modes are as follows:
Service mode
This mode requires a service mode boot of the system and enables the checking of
system devices and features. Service mode provides the most complete self-check of the
system resources. All system resources, except the SCSI adapter and the disk drives that
are used for paging, can be tested.

Concurrent mode
This 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 a diagnostic application before testing can be
done.
Maintenance mode
This 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 started. Maintenance mode requires that all activity on the operating
system is stopped. Run shutdown -m 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.
IBM i and its associated machine code provide dedicated service tools (DSTs) as part of the
IBM i licensed machine code (Licensed Internal Code) and System Service Tools (SSTs) as
part of IBM i. DSTs can be run in dedicated mode (no operating system is loaded). DSTs and
diagnostic tests are a superset of those available under SSTs.
The IBM i End Subsystem (ENDSBS *ALL) command can shut down all IBM and customer
applications subsystems except for 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 keeps all applications running, by using
the IBM i Start Service Tools (STRSST) command (when signed onto IBM i with the
appropriately secured user ID).
With DSTs and SSTs, you can look at various logs, run various diagnostic tests, or take
several kinds of system memory dumps or other options.
Depending on the operating system, the following service-level functions are what you
typically see when you use 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 memory dump manager
Hardware service manager
Call Home/Customer Notification
Error information menu
LED management menu
Alternative method: When you order a Power System, a DVD-ROM or DVD-RAM might
be an option. An alternative method for maintaining and servicing the system must be
available if you do not order the DVD-ROM or DVD-RAM.

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 streamline this process.
Each logical partition reports errors that it detects and forwards the event to the 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 memory dumps.
4.7.2 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 its associated I/O adapters.
Firmware entitlement
With the new HMC Version V8R8.1.0.0 and Power Systems servers, the firmware installations
are restricted to entitled servers. The customer must be registered with IBM and entitled with
a service contract. During the initial machine warranty period, the access key is already
installed in the machine by manufacturing. The key is valid for the regular warranty period plus
some additional time. The Power Systems Firmware is relocated from the public repository to
the access control repository. The I/O firmware remains on the public repository, but the
server must be entitled for installation. When the lslic command is run to display the
firmware levels, a new value, update_access_key_exp_date, is added. The HMC GUI and the
ASMI menu show the Update access key expiration date.
When the system is no longer entitled, the firmware updates fail. Some new System
Reference Code (SRC) packages are available:
E302FA06: Acquisition entitlement check failed
E302FA08: Installation entitlement check failed
Any firmware release that was made available during the entitled time frame can still be
installed. For example, if the entitlement period ends on 31 December 2014, and a new
firmware release is release before the end of that entitlement period, then it can still be
installed. If that firmware is downloaded after 31 December 2014, but it was made available
before the end of the entitlement period, it can still be installed. Any newer release requires a
new update access key.

You can find an update access key at the IBM CoD Home website:
http://www-912.ibm.com/pod/pod
To access the IBM entitled Software Support page for further details, go to the following
website:
http://www.ibm.com/servers/eserver/ess
Firmware updates
System firmware is delivered as a release level or a service pack. Release levels support the
general availability (GA) of new functions 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 introduce new functions. A service pack is an update to an existing release level.
The management console is used for system firmware updates. By using the management
console, you can take advantage of the CFM option when concurrent service packs are
available. CFM is the IBM Power Systems Firmware updates that can be partially or wholly
concurrent or nondisruptive. With the introduction of CFM, IBM is increasing its clients’
opportunity to stay on a given release level for longer periods. Clients that want maximum
stability can defer until there is a compelling reason to upgrade, such as the following
reasons:
A release level is approaching its end-of-service date (that is, it has been available for
about a year, and soon service will not be supported).
They want to move a system to a more standardized release level when there are multiple
systems in an environment with similar hardware.
A new release has a new function that is needed in the environment.
A scheduled maintenance action causes a platform reboot, which provides an opportunity
to also upgrade to a new firmware release.
The updating and upgrading of system firmware depends on several factors, including the
current firmware that is 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, found at the following website:
http://www.ibm.com/support/fixcentral/
You might also want to review the preferred practice white papers that are found at the
following website:
http://www14.software.ibm.com/webapp/set2/sas/f/best/home.html
Firmware update steps
The system firmware consists of service processor microcode, Open Firmware microcode,
and Systems Power Control Network (SPCN) microcode.
The firmware and microcode can be downloaded and installed either from the HMC, or from a
running partition.
Note: The update access key expiration date requires a valid entitlement of the system to
perform firmware updates.

Power Systems has a permanent firmware boot side (A side) and a temporary firmware boot
side (B side). New levels of firmware must be installed first on the temporary side to test the
update’s compatibility with existing applications. When the new level of firmware is approved,
it can be copied to the permanent side.
For access to the initial websites that address this capability, see the POWER8 section on the
IBM Support Portal:
https://www.ibm.com/support/entry/portal/product/power/
For POWER8 based Power Systems, select the POWER8 systems link.
Within this section, search for Firmware and HMC updates to find the resources for keeping
your system’s firmware current.
If there is an HMC to manage the server, the HMC interface can be used to view the levels of
server firmware and power subsystem firmware that are installed and that 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 is 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.
Figure 4-4 shows the different levels in the HMC.
Figure 4-4 HMC System Firmware window
IBM provides the CFM function on the Power E870 and Power E880 models. This function
supports applying nondisruptive system firmware service packs to the system concurrently
(without requiring a reboot operation to activate changes).
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. Deferred fixes, if any, are identified in the Firmware Update Descriptions table of the
firmware document. For deferred fixes within a service pack, only the fixes in the service pack
that 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:
01SV810_030_030 = POWER8 Entry Systems Firmware for 8286-41A and 8286-42A
PPNNSSS_FFF_DDD
PP Package identifier 01 -
NN Platform and class SV Low end
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 the currently installed and the new firmware differ.
The service pack level (FFF) and the last disruptive service pack level (DDD) are equal in
the new firmware.
Otherwise, an installation is concurrent if the service pack level (FFF) of the new firmware is
higher than the service pack level that is installed on the system and the conditions for
disruptive installation are not met.
4.7.3 Concurrent firmware maintenance improvements
Since POWER6, firmware service packs are concurrently applied and take effect immediately.
Occasionally, a service pack is shipped where most of the features can be concurrently
applied, but because changes to some server functions (for example, changing initialization
values for chip controls) cannot occur during operation, a patch in this area required a system
reboot for activation.
With the Power-On Reset Engine (PORE), the firmware can now dynamically power off
processor components, change the registers, and reinitialize while the system is running,
without discernible impact to any applications running on a processor. This potentially allows
concurrent firmware changes in POWER8, which in earlier designs required a reboot to take
effect.
Activating new firmware functions requires installation of a firmware release level. This
process is disruptive to server operations and requires a scheduled outage and full server
reboot.
4.7.4 Electronic Services and Electronic Service Agent
IBM 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 that is 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
Electronic Service Agent
Electronic Services news page
The Electronic Services news page is a single Internet entry point that replaces the multiple
entry points that traditionally are used to access IBM Internet services and support. With the
news page, you can gain easier access to IBM resources for assistance in resolving technical
problems.
Electronic Service Agent
The ESA is software that is on your server. It monitors events and transmits system inventory
information to IBM on a periodic, client-defined timetable. The ESA 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 SSRs when they help answer your
questions or diagnose problems. Installation and use of ESA for problem reporting enables
IBM to provide better support and service for your IBM server.

To learn how Electronic Services can work for you, see the following website (an IBM ID is
required):
http://www.ibm.com/support/electronicsupport
Here are some of the benefits of Electronic Services:
Increased uptime
The ESA tool enhances the warranty or maintenance agreement by providing faster
hardware error reporting and uploading system information to IBM Support. This can
translate to less time that is wasted monitoring the symptoms, diagnosing the error, and
manually calling IBM Support to open a problem record.
Its 24x7 monitoring and reporting mean no more dependence on human intervention or
off-hours customer personnel when errors are encountered in the middle of the night.
Security
The ESA tool is designed to be secure in monitoring, reporting, and storing the data at
IBM. The ESA tool securely transmits either through the Internet (HTTPS or VPN) or
modem, and can be configured to communicate securely through gateways to provide
customers a single point of exit from their site.
Communication is one way. Activating ESA does not enable IBM to call into a customer's
system. System inventory information is stored in a secure database, which is protected
behind IBM firewalls. It is viewable only by the customer and IBM. The customer's
business applications or business data is never transmitted to IBM.
More accurate reporting
Because system information and error logs are automatically uploaded to the IBM Support
center with the service request, customers are not required to find and send system
information, decreasing the risk of misreported or misdiagnosed errors.
When inside IBM, problem error data is run through a data knowledge management
system and knowledge articles are appended to the problem record.
Customized support
By using the IBM ID that you enter during activation, you can view system and support
information by selecting My Systems at the Electronic Support website:
My Systems provides valuable reports of installed hardware and software, using
information that is collected from the systems by ESA. Reports are available for any
system that is associated with the customers IBM ID. Premium Search combines the
function of search and the value of ESA information, providing advanced search of the
technical support knowledge base. Using Premium Search and the ESA information that
was collected from your system, your clients can see search results that apply specifically
to their systems.
For more information about how to use the power of IBM Electronic Services, contact your
IBM SSR, or see the following website:

Service Event Manager
The Service Event Manager allows the user to decide which of the Serviceable Events are
called home with the ESA. It is possible to lock certain events. Some customers might not
allow data to be transferred outside their company. After the SEM is enabled, the analysis of
the possible problems might take longer.
The SEM can be enabled by running the following command:
chhmc -c sem -s enable
You can disable SEM mode and specify what state in which to leave the Call Home feature
by running the following commands:
chhmc -c sem -s disable --callhome disable
chhmc -c sem -s disable --callhome enable
You can do the basic configuration of the SEM from the HMC GUI. After you select the
Service Event Manager, as shown in Figure 4-5, you must add the HMC console.
Figure 4-5 HMC selection for Service Event Manager

In the next window, you can configure the HMC that is used to manage the Serviceable
Events and proceed with further configuration steps, as shown in Figure 4-6.
Figure 4-6 Initial SEM window
Here are detailed descriptions of the different configurable options:
Registered Management Consoles
“Total consoles” lists the number of consoles that are registered. Select Manage
Consoles to manage the list of RMCs.
Event Criteria
Select the filters for filtering the list of serviceable events that are shown. After the
selections are made, click Refresh to refresh the list based on the filter values.
Approval state
Select the value for approval state to filter the list.
Status
Select the value for the status to filter the list.
Originating HMC
Select a single registered console or All consoles to filter the list.
Serviceable Events
The Serviceable Events table shows the list of events based on the filters that are
selected. To refresh the list, click Refresh.
The following menu options are available when you select an event in the table:
View Details...
Shows the details of this event.
View Files...
Shows the files that are associated with this event.

Approve Call Home
Approves the Call Home of this event. This option is available only if the event is not
approved already.
The Help / Learn more function can be used to get more information about the other available
windows for the Serviceable Event Manager.
4.8 Selected POWER8 RAS capabilities by operating system
Table 4-2 provides a list of the Power Systems RAS capabilities by operating system. The
HMC is an optional feature on scale-out Power Systems servers.
Table 4-2 Selected RAS features by operating system
RAS feature AIX
V7.1 TL3 SP3
V6.1 TL9 SP3
IBM i
V7R1M0 TR8
V7R2M0
Linux
RHEL6.5
RHEL7.1
SLES11SP3
SLES12
Ubuntu 15.04
Processor
FFDC for fault detection/error isolation X X X
Dynamic Processor Deallocation X X Xa
Dynamic Processor Sparing using capacity from spare pool X X Xa
Core Error Recovery
Alternative processor recovery X X Xa
Partition Core Contained Checkstop X X Xa
I/O subsystem
PCI Express bus enhanced error detection X X X
PCI Express bus enhanced error recovery X X Xb
PCI Express card hot-swap X X Xa
Memory availability
Memory Page Deallocation X X X
Special Uncorrectable Error Handling X X X
Fault detection and isolation
Storage Protection Keys X Not used by
OS
Not used by
OS
Error log analysis X X Xb
Serviceability
Boot-time progress indicators X X X
Firmware error codes X X X

Operating system error codes X X Xb
Inventory collection X X X
Environmental and power warnings X X X
Hot-swap DASD / media X X X
Dual disk controllers / Split backplane X X X
EED collection X X X
SP “Call Home” on non-HMC configurations X X Xa
IO adapter/device stand-alone diagnostic tests with PowerVM X X X
SP mutual surveillance with POWER Hypervisor X X X
Dynamic firmware update with HMC X X X
Service Agent Call Home Application X X Xa
Service Indicator LED support X X X
System dump for memory, POWER Hypervisor, and SP X X X
Information center / IBM Systems Support Site service
publications
X X X
System Support Site education X X X
Operating system error reporting to HMC SFP application X X X
RMC secure error transmission subsystem X X X
Healthcheck scheduled operations with HMC X X X
Operator panel (real or virtual) X X X
Concurrent Operator Panel Maintenance X X X
Redundant HMCs X X X
Automated server recovery/restart X X X
High availability clustering support X X X
Repair and Verify Guided Maintenance with HMC X X X
PowerVM Live Partition / Live Application Mobility With PowerVM
Enterprise Edition
X Xc
X
EPOW
EPOW errors handling X X Xa
a. Supported in POWER Hypervisor, but not supported in a PowerKVM environment
b. Supported in POWER Hypervisor, with limited support in a PowerKVM environment
c. For POWER8 systems, IBM i requires IBM i 7.1 TR9 and IBM i 7.2 TR1.
RAS feature AIX
V7.1 TL3 SP3
V6.1 TL9 SP3
IBM i
V7R1M0 TR8
V7R2M0
Linux
RHEL6.5
RHEL7.1
SLES11SP3
SLES12
Ubuntu 15.04

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 Power Systems HMC Implementation and Usage Guide, SG24-7491
IBM Power Systems S812L and S822L Technical Overview and Introduction, REDP-5098
IBM Power System S822 Technical Overview and Introduction, REDP-5102
IBM Power Systems S814 and S824 Technical Overview and Introduction, REDP-5097
IBM Power System E870 and E880 Technical Overview and Introduction, REDP-5137
IBM Power Systems SR-IOV: Technical Overview and Introduction, REDP-5065
IBM PowerVM Best Practices, SG24-8062
IBM PowerVM Enhancements What is New in 2013, SG24-8198
IBM PowerVM Virtualization Introduction and Configuration, SG24-7940
IBM PowerVM Virtualization Managing and Monitoring, SG24-7590
Performance Optimization and Tuning Techniques for IBM Processors, including IBM
POWER8, SG24-8171
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
Online resources
These websites are also relevant as further information sources:
Active Memory Expansion: Overview and Usage Guide documentation:
http://public.dhe.ibm.com/common/ssi/ecm/en/pow03037usen/POW03037USEN.PDF
IBM EnergyScale for POWER8 Processor-Based Systems white paper:
http://public.dhe.ibm.com/common/ssi/ecm/po/en/pow03125usen/POW03125USEN.PDF
IBM Power Facts and Features - IBM Power Systems, IBM PureFlex System, and Power
Blades:
http://www.ibm.com/systems/power/hardware/reports/factsfeatures.html
IBM Power System E870 server specifications:
http://www.ibm.com/systems/power/hardware/e870/specs.html

IBM Power System E880 server specifications:
http://www.ibm.com/systems/power/hardware/e880/specs.html
Help from IBM
IBM Support and downloads
ibm.com/support
IBM Global Services
ibm.com/services

®
REDP-5137-00
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™
New modular
architecture for
increased reliability
Enterprise POWER8
processor-based
servers
Exceptional memory
and I/O bandwidth
This IBM Redpaper publication is a comprehensive guide covering the
IBM Power System E870 (9119-MME) and IBM Power System E880
(9119-MHE) servers that support IBM AIX, IBM i, and Linux operating
systems. The objective of this paper is to introduce the major
innovative Power E870 and Power E880 offerings and their relevant
functions:
The new IBM POWER8 processor, available at frequencies of 4.024
GHz, 4.190 GHz, and 4.350 GHz.
Significantly strengthened cores and larger caches
Two integrated memory controllers with improved latency and
bandwidth
Integrated I/O subsystem and hot-pluggable PCIe Gen3 I/O slots
Improved reliability, serviceability, and availability (RAS) functions
IBM EnergyScale technology that provides features such as power
trending, power-saving, capping of power, and thermal
measurement
This publication is for professionals who want to acquire a better
understanding of IBM Power Systems products.
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 E870 and Power E880 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.
Back cover

Ibm power systems e870 and e880 technical overview and introduction

More Related Content

What's hot (20)

Similar to Ibm power systems e870 and e880 technical overview and introduction (20)

More from Diego Alberto Tamayo (20)

Recently uploaded (20)

Ibm power systems e870 and e880 technical overview and introduction