SlideShare a Scribd company logo

High Performance Computing: State of the Industry

I
IMEX Research

The document discusses the state of the high performance computing (HPC) industry. It covers key topics like HPC technologies, issues, trends and applications. Some of the main points made are: - HPC has seen rapid progress over the past few decades in terms of performance, price reduction and target audiences. Systems have become more affordable and accessible to smaller businesses and professionals. - Cluster computing has risen significantly, allowing organizations to build powerful systems from commodity hardware. Standardization has made HPC more mainstream. - Major market drivers include the availability of affordable industry-standard hardware and software, as well as cluster-ready applications accelerating adoption across industries like engineering, bioinformatics and finance. -

Technology
1 of 56
Downloaded 51 times
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
High Performance Computing State of the Industry Anil Vasudeva Principal Analyst & President anil@imexresearch.com IMEX RESEARCH.COM 408-268-0800 ©2000-2005 IMEX Research IMEX
• HPC - Markets Drivers & Industry Dynamics • State of HPC Technologies Clustering, Interconnects, Storage, Blade Servers • Industry Issues/Technology Hurdles Standards, Management, Fabrics, Thermals, IP Acceleration • Futures Trends - Computing & Grids • Holy Grails Volume Driven Economics, IP Evolution, TF/Desktop HPC in our Daily Lives (Academia > Hollywood > Wall Street) ©2000-2005 IMEX Research IMEX
Server Architectures: Competitive Technologies Price/Performance Performance: SMP vs. Clustering Serial vs.Parallel Computing Multinode Clusters Performance - tpc SMP w/8 way MPP w/4 way Serial Price 16 32 MF 8 N-way SMP 6 4 2 Clustering i Performance Number of Processors Scalable Perf. vs. Availability Scalable Performance MPP Clustering SMP FT UniP ©2000-2005 IMEX Research Availability IMEX
Market Segments by Applications OLTP Transaction 10 K Processing eCommerce Data DSS (RAID - 0, 3) (RAID - 1, 5, 6) 1K Warehousing (Latency) Visual DB IOPs 100 Scientific Computing HPC HPC Imaging TPC 10 Audio Streaming Streaming HPC Video 1 1 5 10 50 100 500 MB/sec *IOs per sesond for a required response time ( ms) ©2000-2005 IMEX Research IMEX
HPC – 2 decades of rapid progress 1991 1998 2005 2010 System Cray Y-MP C916 Sun HPC10000 Shuttle@newegg Blade 24x 333 MHz, Ultra 4x 2.6 GHz, x64 Architecture 16x Vector, 4 GB, Sparc II, 24 GB,S- Bit, 4 4x 10GHz, x64 bit, Bus Bus GB, GbE 16GB, 10GbE Win Server 2003 OS UNICOS Solaris 2.5.1 SP2 LongHorn/Linux Gflops ~10 ~10 ~10 ~100 Price $40 million $1 Million <$4000 ~$1000 Price Reduction 1 1/40 1/10,000 1/400,000 Government Small>Large Target Audience National Labs Large Enterprises Businesses Every Professional Bioinformatics, Applications Classified, Climate, Manufacturing, Materials Architects,Wall Physics Energy,Finance Sciences Street, Hollywood ©2000-2005 IMEX Research IMEX
The rapid rise of Clusters in HPC Rise of Cluster Computing 350 # Clustered Computers in Top 500.org 300 250 200 150 100 50 0 1999e 2000 2001 2002 2003 2004e ©2000-2005 IMEX Research IMEX
HPC Market Drivers • Availability of industry standard HW making HPC mainstream • HSV Servers @$200, Storage @25c/GB, Add-On GbE NICS at $10, GbE Switches at $2-5/port • Migration to Embedded ASICs on Blades • Rapid procurement, installation & System Integration simplifying acquisition • Cluster ready apps accelerating market growth • Engineering (LSTC, Ansys..) • Bioinformatics (Blast, Gaussian..) • Finance (Matlab, Excel…) • Oil & Gas (Eclipse, Promagic..) • Government (Uexplore, NOAA Hysplit..) ©2000-2005 IMEX Research IMEX
HPC Servers Rev - 15% of Total Server Market WW HPC Rev ($M) $9,000 $8,000 $7,000 $6,000 Departm ental Rev $M $5,000 $4,000 $3,000 Divisional $2,000 Enterprise $1,000 Capability $- 2003 2004 2005 2006 2007 2008 ©2000-2005 IMEX Research IMEX
HPC Market Leaders 2004 Market Leaders ©2000-2005 IMEX Research IMEX
Rise of Industry Standard Architecture in HPC X86 Architecture (Intel & AMD) using Linux and Microsoft based clusters dominate in High Perrformance Computing Architecture. TeraFlop HPC Using Off-the-Shelf Modules Intel servers with PCI Express, Infiniband HBAs & Switches, SATA/SAS HDD or AMD Opteron Servers, GbE HBA/Switches, SATA ©2000-2005 IMEX Research IMEX
Factors Affecting Performance • Applications Configurations affect Linpack significantly (3 metrics have the most affect) • Problem Size • Size of Blocks • Topology • Tightly Coupled MPI Applications • Very sensitive to network performance characteristics - internodal communications delay - OS Network Stack is a significant factor • Very sensitive to mismatched node performance - Random OS activities can add msec delays to usec type communication line delays ©2000-2005 IMEX Research IMEX
HPC Architecture & Challenges Architecture Goals & Criteria Goals • Maximize compute time HPC Communications Stack vs. messaging time • High constant cross- Entire Stack sectional throughput Needs Attention • Reduce or hide latency Data Application Metrics Upper Layers • Bandwidth TCP/IP (or Offload I/F) • Latency OS • System Interface CPU • SW Stack Memory Chipset Challenge is to move data from I/O interconnect Network Interface application to the network and back (w or w/o TOE) with minimum latency Data ©2000-2005 IMEX Research IMEX
Application Application Application Application Application Memory Memory Control Control Host Host Host Socket RDMA CPU RDMA Adapter TCP TCP CPU CPU IP IP Data Data Kernel Kernel Traditional Memory Memory Adapter Ethernet Ethernet RDMA Adapter RDMA Adapter Network ©2000-2005 IMEX Research IMEX
Interconnect Price/Performance Hierarchy High MB/s, Low Latency, Price/Port Quadrics Myrinet InfiniBand 10 GbE w/ RDMA, TOE 10 GbE w/ TOE 1 GbE Market Size Microsoft Chimney Software 2004-07 ©2000-2005 IMEX Research IMEX
State of Interconnect Fabrics EHERNET - LAN/SAN/MAN/WAN FE GbE ATM IBM/SP Latency/hop MPP SMP/NUMA Servernet Dolphin T3E Myrinet SGI Mellanox IB GbE w/TOE Quadrics FJ/Synfinity GbE w/RDMA 10GbE w/RDMA Bandwidth/link ©2000-2005 IMEX Research IMEX
Interconnects: Industry Progress ©2000-2005 IMEX Research IMEX
InfiniBand Clustering • InfiniBand and Industry Standard Servers – Displacing Mainframe Computers in the Data Center with Three Standard Hardware Components – Powerful Low Cost Combo Out Performs SMP 12-Nodes (24-CPUs) Up to 72 GFlops of CPU Power; 24+ GB Memory Industry Standard Server InfiniBand Switch => InfiniBand HCA Enterprise Availability: ©2000-2005 IMEX Research With Fabric Failover IMEX
InfiniBand rears up its head • Low Cost IB 10G Silicon & Adapters – Very Low Cost 10G Host Adapters • Host Channel Adapter ~ $69 in volume • Switch Si at less than $30 a port in volume • Compelling price/performance • InfiniBand vs. 10GE Advantages in the Short Term – 15X – 20X price advantage vs 10GE – 3X – 10X latency advantage – 10X – 20X CPU utilization advantage – 3X-5X throughput advantage • Adoption by most majors • Products – InfiniBand to IP * – InfiniBand to FC – Advanced Management ©2000-2005 IMEX Research IMEX
Ethernet - TCP + iSCSI Offload Engines Typical 1GbE NIC File System File System Host File System SW Volume Manager Volume Manager Volume Manager SCSI Device Driver SCSI Device Driver SCSI Device Driver iSCSI iSCSI iSCSI Adapter TCP TCP TCP IP IP IP Ethernet Ethernet Ethernet Host CPU HBA Firmware HBA Hardware ©2000-2005 IMEX Research IMEX
Interconnects: Applicability Grids 100 Mbit 1 Gig 10 Gig 40 Gig Data Center Ethernet Ethernet Ethernet Ethernet LAN SAN Fibre Channel Storage RDMA NICs with iSCSI NAS Storage RDMA NICs Clusters Myrinet InfiniBand Blade Quadrics Chassis Network SMP PCI Express & Advanced Switching Backplanes SunFire Link, IBM Federation <1 Gigabit 1 Gigabit 10 Gigabit >10 Gigabit ©2000-2005 IMEX Research IMEX
InfiniBand Adopters • Major providers of InfiniBand Solutions • Enterprise Class Products – InfiniBand to IP – InfiniBand to FC – Advanced Mgmt • Platform support includes – Intel (IA32 + 64EMT, Itanium), AMD Opteron, Power PC, Sparc • Operating Systems – Linux, Windows, AIX, HPUX, Solaris, MAC OS X, VxWorks ©2000-2005 IMEX Research IMEX
HPC Interconnect – Leaders Myrinet for the highly latency sensitive applications and GbE for the majority applications Dominate the Interconnect for HPC. Infiniband is rearing to dominate at the Midrange latency applications ©2000-2005 IMEX Research IMEX
HPC Tools Availability ©2000-2005 IMEX Research IMEX
Data Center Environmentals Prep Temp F Speed fpm Sources: Fluent Airpak, Ansys CFD, ©2000-2005 IMEX Research IMEX
HPC embraces a broad spectrum of markets .. High Performance Computing World ©2000-2005 IMEX Research IMEX
3 Tier Computing Infrastructure Application End-Users  Uniform and ubiquitous physical Internet Connectivity – “Wire Once”  Any user to any server via Internet Tier –1  Any server to any server via the LAN Network Services:  Any sever to any storage via the SAN - Web Servers,  Dynamic Logical Binding of… - Firewalls, - Load balancers  server network identification LAN  server OS version  server application assignment Tier-2  application data volumes Application Services: - Application Servers  Resulting in  Application Mobility across servers  Data mobility across storage Tier-3 SAN Data Base Services - Database Servers - Storage Storage ©2000-2005 IMEX Research IMEX
E2E Internet & 3 Tier Computing HA, Secure Data Center FC SAN IP SAN NAS NAS IP Storage Network Web App Internet Web DB Edge Core Web App Optical Web DB Web App Web Web DB App Web Edge Applications Data Center Directory Security Policy Management Software OS Platform ©2000-2005 IMEX Research IMEX
Fabric based Clusters Architecture Ethernet, Network Fabric Wi-Fi. System Fabric Quadrics, Management Myrinet, SNMP, SCI Fabric IETF/ InfiniBand, CIM, Ethernet / IP, SMI-S, Ethernet IP SMASH. w/TOE, Ethernet IP w/TOE and RDMA. SCSI, Fiber Channel, ISCSI. ©2000-2005 IMEX Research IMEX
Fabric based Grid Applications Storage Systems ©2000-2005 IMEX Research IMEX
Scale Up (Consolidated SMP Servers) Server Platforms - Migrations Large SMP Clusters Parallel Sysplex High Availability/ - Scalable Nodes, Hi Perform.Clusters - Dynamic LPAR - Linux/AIX/Solaris Clusters - SW Provisioning/VMWare - HPC Clusters IBM: i890/p690/z990 DELL: PowerEdge 6650 SUN: SunFire 4800, 12K ORACLE: 9iRAC/PE6650 Blade Servers High Density Rack Mount - Highly Integrated Dense Form Factor - Rapid Deployment, Flexible Architect. RLX: ServerBlade 300i, 800i, HP: 40b, p FUJITSU: BX300, DELL:1655 IBM: HS20, Sun: Scale Out (Clustered Blade Servers) ©2000-2005 IMEX Research IMEX
Blades Servers - Infrastructure Management Modules Remote Mgmt+KVM over IP Midplane Networking: Gbit Ethernet Switches W/Connectors Storage: IP NAS or FC SAN To Blades & Back Switch Cooling N+1Fans/Cooling Modules Processor Blades Modules (6-24 typically) Power: N+1Power Supplies Blade Control Panel USB Ports, CD/Floppy I/F ~ GbE Switch supports Trunking/Port Aggregation Flow Control Gbit QoS Packet Prioritization Memory Ethernet SNMP/RMON Systems I/F Micro- DDR w IGMP/BOOTP/TFTP Monitor Processors ECC …… Module © 2003 Source: IMEX Research ©2000-2005 IMEX Research IMEX
Management – Blade Servers Ultra-dense HW • Density 20 % • Power Consumption 20% • Cable Management Management • Platforms integration:  Hardware  Software 80%  Network  Management • Scalability 80 % • Security • Manageability • Lowest TCO ©2000-2005 IMEX Research IMEX
Blade Servers Market Opportunity $12 WW Blade Servers $10 Market Opportunity (cum 2004-2008) Factory Rev $B $8 Revenues ~ $24 Billion Units ~ 9 Million $6 $4 $2 $- 2003 2004 2005 2006 2007 2008 Data: IMEX Research 2005 ©2000-2005 IMEX Research IMEX
Blade Servers: Vendor Positioning Index (As of 31 Mar 2004 - See IMEX Blade Servers Industry Report 2005 for new data) Strategy (Potential) IBM Fujitsu Siemens Hitachi HP Linux Intel © 2003-05 IMEX Research NEC Networx Dell Nexcom Tatung Sun Appro Egenera Penguin Rackable Sys HPC Syst Verari ©2000-2005 IMEX Research Delivery (Execution) IMEX
iSCSI on roadmap of storage vendors SAN iSCSI NAS (Fibre COST (IP) (IP) Channel) •iSCSI delivers low-cost Networked Storage –Higher availability,Increased usage High End –Centralized management,Storage functionality FC & iSCSI Mid-Tier FC & iSCSI Mid-Tier NAS & iSCSI Entry FC & iSCSI SERVICE LEVEL ©2000-2005 IMEX Research IMEX
System Cost vs. High Availability z/390 $10 M Sysplex Avg. S/390 S/390 Syst z/390 MVS MVS IBM IBM Price Propl Propl $1 M Prop. Sun-Solaris Sun-Solaris HP-UX HP-UX IBM IBM IBM-AIX IBM-AIX $100 K Clustered NCR-SVR4 NCR-SVR4 Bull Bull Dell Dell UNIX AS400 Fujitsu Fujitsu HP HP IBM IBM NEC NEC Siemens Siemens Stratus Stratus UNIX Cluster Win Win $10K Linux Clustered Linux 0.1 1 10 100 Downtime Hrs./Syst/Yr 99.999% 99.99% 99.9% 99.0% System Availability ©2000-2005 IMEX Research IMEX
The iSCSI SAN Client Workstations Client requests data from App Server LAN Ethernet Header IP TCP DATA CRC Ethernet Switches Application & DB Servers (iSCSI Initiator) iSCSI-equipped App Server (“initiator”) requests data from an iSCSI storage (“target”) SAN Ethernet CRC IP TCP iSCSI SCSI DATA GigE Header Switches Storage Arrays (iSCSI Target) To Other LANs, Servers or IP SANs ©2000-2005 IMEX Research IMEX
Tiered Storage by Price/Performance Desktop PC Servers Large Enterprise Entry NAS Midline Storage SAN & NAS Workstations Mainstream NAS/SAN PC SAS SAS HBA/RAID FC RAID Chipset HBA Drive Interface SATA SAS SAS Drive Types 5400 RPM 10K RPM 7200 RPM 15K RPM SATA Drive SAS Drive FC Drive Source:Maxtor ©2000-2005 IMEX Research IMEX
Cost Driven Data Storage Markets 100% 90% SATA 80% 70% 60% SAS 50% 40% 30% FC 20% 10% P_SCSI 0% 2003 2004 2005 2006 2007 2008 ©2000-2005 IMEX Research IMEX
Grid Computing • Grid is a Catch-all Marketing Term • Means different things to different constituencies • Desktop Cycle-Stealing • Manahed HPC Clusters • Virtualization of Data Center Resources • Outsourcing to “Utility Data Centers” • Multiple, opposing requirements • Compute-intensive applications want to ship data tp idle processor resources • Data-Intensive applications want to ship computations to appropriate large data repositories • An evolving entity with many pulls & pushes • Economics, Feasibilities, management of hetero vs homogeneous HW ©2000-2005 IMEX Research IMEX
Grid Challenges: Hype vs. Reality • Cost of Local Grids (LAGs) – ala Blades cheaper than Wide Area Grids (WAGs) • Computing Cost $1000 > 1 cpu day (10 Terops) =$1 • 10 TB network transfer Cost = $1 • Internet BW cost $100/mbps/mo > 1GB network xfr cost ~$1 • Result: Local HPC Cluster is 10,000 cheaper than WAN Communication • MPI Style Apps work well in LAN Clusters but uneconomical in WAN • Data analysis best done by moving programs to data not data to programs • Small data, high compute applications woek well across internet • Internet is not the CPU back plane ©2000-2005 IMEX Research IMEX
Grids – Homogeneous vs Heterogeneous ©2000-2005 IMEX Research IMEX
Grid Services Architecture & Standards ©2000-2005 IMEX Research IMEX
Grid Computing – Across 3 Tiers Grid Control Storage Databases Application Servers ©2000-2005 IMEX Research IMEX
Utility Computing/Web Services Autonomics • Self Healing • Self Optimizing • Universal Identity • Single System Image • …. Internet Internet Network Switching Pool • HP - Utility Data Center Utility Controller Virtualization Load Balancer Pool • Sun - N1/SunONE/Orion Firewall Pool Server Virtualization • Microsoft - .Net Server Pool NAS Pool • IBM - OnDemand/Autonomics Storage Switching Pool Virtualization Storage Pool ©2000-2005 IMEX Research IMEX
Future Computing: Processes Driven Initiator Initiator Initiator Application Application Application Old model New model Hardware is the initiator Processes are the initiators HBA ID Initiator Allows applications to move storage wherever they are - storage no longer cares where they are Target ©2000-2005 IMEX Research IMEX
Application-Driven Storage Virtualization Application Application Application Moving from storage ...to storage associated associated with with application servers... processes Old Model New Model Server Hardware is the Allows applications to initiator. Storage move storage from the Drives are the Target virtualized pool, wherever they may be ©2000-2005 IMEX Research IMEX
Future: Embedded10Gb,TOE, iSCSI CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU TOE MC MC MC MC MC MC MC MC DMA DMA PCI Exp PCI Exp PCI Exp PCI Exp PCI Exp I/O Uses I/O Uses I/O 1GbE Bridge Bridge Syst Bridge Syst 10 GbE Mmry Mmry 1GbE TOE GbE TOE 1GbE GbE 1GbE PCI-X Ext TOE Mmry PCI PCI-X TOE PCI PCI-X with 10GbE 2GbE Ext.Mmry 2x1Gb TOE NIC TOE chipset 10 Gb TOE NIC Embedded 2003 2004 2005 2006 onwards ©2000-2005 IMEX Research IMEX
Future: IP Storage Management on a chip Host Services Integration Storage Provisioning File system monitoring Storage provisioning Win, LINUX, Solaris Layer IP SAN Management Management Management of MultiPath IO Supp Security IP SAN Console iSCSI HBAs and Failover (iSNS, CHAP, SRP) Management Layer Virtualization Mirroring Snapshot Fail-Over iSCSI Target Management Appliance LVM, Error Handling, SCSI Daemon, API Interoperability Service Layer HW Acceleration: TOE, iSCSI Offload, IPsec ©2000-2005 IMEX Research IMEX
Future Intelligent IP Infrastructure Present Data Center Infrastructue Future Infrastructue Clients LAN Ntwk Mgmt WS Application & Intelligent DB Servers Uniform IP-based Infrastructure Mgmt SAN Storage WS Mgmt WS Storage Sub Systems ©2000-2005 IMEX Research IMEX
Future - IP Based Infrastructure Servers Storage IP Networks Infrastructure Telecom Virtualization Provisioning Automation ©2000-2005 IMEX Research IMEX
HPC – From Academia to Wall St to Hollywood High Performance Entertainment High Performance Commercial Bioinformatics Decision-Support Entertainment Commercial BioInformatics Decision Audio/Video OnDemand Computing Computing Visualization Visualization Support Syst Audio/Video On Demand Systems Data: IMEX Research & Panasas 100+ Teraflops Rendering (Texture & Polygons) Data rate & capacity Throughput = 100 GB/s Throughput = 1.2 GB/s Throughput : DSL/Cable ©2000-2005 IMEX Research IMEX ©2004--05 IMEX Research.com
HPC: From Homeland Security to Bioinformatics ©2000-2005 IMEX Research IMEX
HPC From Academia to Hollywood ©2000-2005 IMEX Research IMEX
High Performance Computing State of the Industry Anil Vasudeva Principal Analyst & President anil@imexresearch.com IMEX RESEARCH.COM 408-268-0800 ©2000-2005 IMEX Research IMEX
State of Interconnect Fabrics Latency/hop 1 ms EHERNET - LAN/SAN/MAN/WAN FE GbE IB - CLUSTER ATM IBM/SP MPP 100 us Servernet Dolphin SMP/NUMA T3E Myrinet SGI GbE Quadrics Mellanox w/TOE 10 us FJ/Synfinity Synfinity-1 GbE 10GbE w/RDMA Scali w/RDMA 1us 100 Mb/sec 1Gb/sec 10 Gb/sec Bandwidth/link ©2000-2005 IMEX Research IMEX

More Related Content

PDF
Scandinavian Rail Development 2013 - Stig Kyster-Hansen
Russell Publishing
 
PDF
SUPPLY CHAIN OPTIMIZATIONL: JET FUEL SUPPLY CHAIN
Alkis Vazacopoulos
 
KEY
Interop Cloud Keynote with Terremark, VMWare, and Citrix
Alistair Croll
 
KEY
L0001 - The Terminology of the Eclipse Platform
Tonny Madsen
 
PDF
Using Content To Accelerate & Keep Leads Warm
G3 Communications
 
PDF
Otm con8923 pdf_8923_0002
jucaab
 
PDF
Service Density By Xelerated At Linley Seminar
Xelerated
 
PDF
Web Page Load Cycle Time Six Sigma Case Study
Steven Bonacorsi
 
Scandinavian Rail Development 2013 - Stig Kyster-Hansen
Russell Publishing
 
SUPPLY CHAIN OPTIMIZATIONL: JET FUEL SUPPLY CHAIN
Alkis Vazacopoulos
 
Interop Cloud Keynote with Terremark, VMWare, and Citrix
Alistair Croll
 
L0001 - The Terminology of the Eclipse Platform
Tonny Madsen
 
Using Content To Accelerate & Keep Leads Warm
G3 Communications
 
Otm con8923 pdf_8923_0002
jucaab
 
Service Density By Xelerated At Linley Seminar
Xelerated
 
Web Page Load Cycle Time Six Sigma Case Study
Steven Bonacorsi
 

What's hot (8)

PDF
Transnet’S Competitive Supplier Development Program (Csdp)
Railways and Harbours
 
PDF
exchange2010-Architecture
Selva G Kumar
 
PDF
Patel Work Skills
Monesh Patel
 
PDF
Maersk Line's Agile Journey LESS 2012
OzlemYuce
 
PDF
Virt Exchange2k7 Final Frontier V Mworld2007
Kong Yang
 
PPT
Enterprise Service Bus Part 1
Return on Intelligence
 
PPTX
Oracle da-15dec2011
Agora Group
 
PDF
Oracle Elastic Cloud
Freelance PR
 
Transnet’S Competitive Supplier Development Program (Csdp)
Railways and Harbours
 
exchange2010-Architecture
Selva G Kumar
 
Patel Work Skills
Monesh Patel
 
Maersk Line's Agile Journey LESS 2012
OzlemYuce
 
Virt Exchange2k7 Final Frontier V Mworld2007
Kong Yang
 
Enterprise Service Bus Part 1
Return on Intelligence
 
Oracle da-15dec2011
Agora Group
 
Oracle Elastic Cloud
Freelance PR
 
Ad

Viewers also liked (9)

PDF
슈퍼컴과 국가경쟁력
Lee Jysoo
 
PPTX
How HPC Transforms the Corporate Information Technology Ecosystem
inside-BigData.com
 
PDF
High Performance Computing using MPI
Ankit Mahato
 
PDF
HPC in R
Vyacheslav Arbuzov
 
PDF
High Performance Computing: an Introduction for the Society of Actuaries
Adam DeConinck
 
PDF
High–Performance Computing
BRAC University Computer Club
 
PDF
High performance computing - building blocks, production & perspective
Jason Shih
 
PDF
HPC Market Update from IDC
inside-BigData.com
 
PPT
Cluster Computing
BOSS Webtech
 
슈퍼컴과 국가경쟁력
Lee Jysoo
 
How HPC Transforms the Corporate Information Technology Ecosystem
inside-BigData.com
 
High Performance Computing using MPI
Ankit Mahato
 
HPC in R
Vyacheslav Arbuzov
 
High Performance Computing: an Introduction for the Society of Actuaries
Adam DeConinck
 
High–Performance Computing
BRAC University Computer Club
 
High performance computing - building blocks, production & perspective
Jason Shih
 
HPC Market Update from IDC
inside-BigData.com
 
Cluster Computing
BOSS Webtech
 
Ad

Similar to High Performance Computing: State of the Industry (20)

PPTX
Extending OPC-UA through Architecture Flexibility, Performance, and Scalability
Real-Time Innovations (RTI)
 
PDF
IT-as-a-Service: Cloud Computing and the Evolving Role of Enterprise IT
Bob Rhubart
 
PPTX
DAC 2012
FlexTiles Team
 
PDF
A series presentation
Sergey Marunich
 
PPTX
WSO2Con2011: Using WSO2 ESB with SAP ERP (Retail)
WSO2
 
PDF
10 fn s22
Scott Foster
 
PDF
10 fn s22
Scott Foster
 
PDF
MPLS-TP (MPLS Transport Profile)
Shivlu Jain
 
PDF
LTE Radio Layer 2 And Rrc Aspects
BP Tiwari
 
PDF
Rev 090004 Radio Layer 2 And Rrc Aspects
maddiv
 
PDF
10 fn s26
Scott Foster
 
PDF
10 fn s26
Scott Foster
 
PDF
Technology Disruption Brings New VAS Opportunities
Radisys Corporation
 
PDF
Varkon Semiconductor
Rajiv Parmar
 
PDF
OpenSPARC T1 Processor
DVClub
 
PDF
Simulation Directed Co-Design from Smartphones to Supercomputers
Eric Van Hensbergen
 
PDF
1
srimoorthi
 
PDF
3G & LTE Wireless Solutions
Continuous Computing
 
PPTX
eXtremeDB FE
hyeongchae lee
 
PDF
Imex Research Virtualization Executive Summary On Slideshare
M. R. Pamidi, Ph. D.
 
Extending OPC-UA through Architecture Flexibility, Performance, and Scalability
Real-Time Innovations (RTI)
 
IT-as-a-Service: Cloud Computing and the Evolving Role of Enterprise IT
Bob Rhubart
 
DAC 2012
FlexTiles Team
 
A series presentation
Sergey Marunich
 
WSO2Con2011: Using WSO2 ESB with SAP ERP (Retail)
WSO2
 
10 fn s22
Scott Foster
 
10 fn s22
Scott Foster
 
MPLS-TP (MPLS Transport Profile)
Shivlu Jain
 
LTE Radio Layer 2 And Rrc Aspects
BP Tiwari
 
Rev 090004 Radio Layer 2 And Rrc Aspects
maddiv
 
10 fn s26
Scott Foster
 
10 fn s26
Scott Foster
 
Technology Disruption Brings New VAS Opportunities
Radisys Corporation
 
Varkon Semiconductor
Rajiv Parmar
 
OpenSPARC T1 Processor
DVClub
 
Simulation Directed Co-Design from Smartphones to Supercomputers
Eric Van Hensbergen
 
1
srimoorthi
 
3G & LTE Wireless Solutions
Continuous Computing
 
eXtremeDB FE
hyeongchae lee
 
Imex Research Virtualization Executive Summary On Slideshare
M. R. Pamidi, Ph. D.
 

More from IMEX Research (15)

PDF
IMEXresearch software defined storage
IMEX Research
 
PDF
Architecting Next Generation Enterprise Network Storage
IMEX Research
 
PDF
Network Storage: State of the Industry
IMEX Research
 
PDF
Next-Gen Data Center Virtualization: Studies in Implementation
IMEX Research
 
PDF
Next Gen Data Center Implementing Network Storage with Server Blades, Cluster...
IMEX Research
 
PDF
Flash Usage Models for the Oracle Database
IMEX Research
 
PDF
The New Data Center: Consolidated, Converged, Virtualized & Automated
IMEX Research
 
PDF
Blade Servers & Virtualization State of the Industry
IMEX Research
 
PDF
Blade Servers & Virtualization: State of the Industry
IMEX Research
 
PDF
Next-Gen Data Center: Improving TCO & ROI in Data Centers Through Virtualizat...
IMEX Research
 
PDF
Next Generation Data Centers
IMEX Research
 
PDF
Key to Efficient Tiered Storage Infrastructure
IMEX Research
 
PDF
Solving the IO Bottleneck
IMEX Research
 
PDF
SSD: Ready for Enterprise and Cloud?
IMEX Research
 
PDF
Big Data Overview
IMEX Research
 
IMEXresearch software defined storage
IMEX Research
 
Architecting Next Generation Enterprise Network Storage
IMEX Research
 
Network Storage: State of the Industry
IMEX Research
 
Next-Gen Data Center Virtualization: Studies in Implementation
IMEX Research
 
Next Gen Data Center Implementing Network Storage with Server Blades, Cluster...
IMEX Research
 
Flash Usage Models for the Oracle Database
IMEX Research
 
The New Data Center: Consolidated, Converged, Virtualized & Automated
IMEX Research
 
Blade Servers & Virtualization State of the Industry
IMEX Research
 
Blade Servers & Virtualization: State of the Industry
IMEX Research
 
Next-Gen Data Center: Improving TCO & ROI in Data Centers Through Virtualizat...
IMEX Research
 
Next Generation Data Centers
IMEX Research
 
Key to Efficient Tiered Storage Infrastructure
IMEX Research
 
Solving the IO Bottleneck
IMEX Research
 
SSD: Ready for Enterprise and Cloud?
IMEX Research
 
Big Data Overview
IMEX Research
 

Recently uploaded (20)

PDF
Blue Purple Modern Animated Computer Science Presentation.pdf.pdf
Akar16
 
PDF
Electronic commerce courselecture one. Pdf
OmerMohamed64
 
PDF
Mobile App Security Testing_ A Comprehensive Guide.pdf
flufftailshop
 
PDF
Diabetes mellitus diagnosis method based random forest with bat algorithm
IAESIJAI
 
PDF
cuic standard and advanced reporting.pdf
SimoneTitaniumTomase
 
PPTX
Understanding_Digital_Forensics_Presentation.pptx
ImranKhan423233
 
PDF
Machine learning based COVID-19 study performance prediction
IAESIJAI
 
PDF
Shreyas Phanse Resume: Experienced Backend Engineer | Java • Spring Boot • Ka...
SHREYAS PHANSE
 
PDF
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...
Peter Tan
 
PDF
Dropbox Q2 2025 Financial Results & Investor Presentation
Dropbox
 
PDF
Approach and Philosophy of On baking technology
30anthuong17
 
PDF
How UI/UX Design Impacts User Retention in Mobile Apps.pdf
SynapseIndia
 
PDF
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
SOFTTECHHUB
 
PDF
NewMind AI Monthly Chronicles - July 2025
NewMind AI
 
PDF
KodekX | Application Modernization Development
KodekX
 
PPT
Teaching material agriculture food technology
LiaRayya
 
PDF
Build a system with the filesystem maintained by OSTree @ COSCUP 2025
Jian-Hong Pan
 
PDF
NewMind AI Weekly Chronicles - August'25 Week I
NewMind AI
 
PDF
Review of recent advances in non-invasive hemoglobin estimation
IAESIJAI
 
PDF
Empathic Computing: Creating Shared Understanding
Mark Billinghurst
 
Blue Purple Modern Animated Computer Science Presentation.pdf.pdf
Akar16
 
Electronic commerce courselecture one. Pdf
OmerMohamed64
 
Mobile App Security Testing_ A Comprehensive Guide.pdf
flufftailshop
 
Diabetes mellitus diagnosis method based random forest with bat algorithm
IAESIJAI
 
cuic standard and advanced reporting.pdf
SimoneTitaniumTomase
 
Understanding_Digital_Forensics_Presentation.pptx
ImranKhan423233
 
Machine learning based COVID-19 study performance prediction
IAESIJAI
 
Shreyas Phanse Resume: Experienced Backend Engineer | Java • Spring Boot • Ka...
SHREYAS PHANSE
 
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...
Peter Tan
 
Dropbox Q2 2025 Financial Results & Investor Presentation
Dropbox
 
Approach and Philosophy of On baking technology
30anthuong17
 
How UI/UX Design Impacts User Retention in Mobile Apps.pdf
SynapseIndia
 
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
SOFTTECHHUB
 
NewMind AI Monthly Chronicles - July 2025
NewMind AI
 
KodekX | Application Modernization Development
KodekX
 
Teaching material agriculture food technology
LiaRayya
 
Build a system with the filesystem maintained by OSTree @ COSCUP 2025
Jian-Hong Pan
 
NewMind AI Weekly Chronicles - August'25 Week I
NewMind AI
 
Review of recent advances in non-invasive hemoglobin estimation
IAESIJAI
 
Empathic Computing: Creating Shared Understanding
Mark Billinghurst
 

High Performance Computing: State of the Industry

  • 1. High Performance Computing State of the Industry Anil Vasudeva Principal Analyst & President anil@imexresearch.com IMEX RESEARCH.COM 408-268-0800 ©2000-2005 IMEX Research IMEX
  • 2. • HPC - Markets Drivers & Industry Dynamics • State of HPC Technologies Clustering, Interconnects, Storage, Blade Servers • Industry Issues/Technology Hurdles Standards, Management, Fabrics, Thermals, IP Acceleration • Futures Trends - Computing & Grids • Holy Grails Volume Driven Economics, IP Evolution, TF/Desktop HPC in our Daily Lives (Academia > Hollywood > Wall Street) ©2000-2005 IMEX Research IMEX
  • 3. Server Architectures: Competitive Technologies Price/Performance Performance: SMP vs. Clustering Serial vs.Parallel Computing Multinode Clusters Performance - tpc SMP w/8 way MPP w/4 way Serial Price 16 32 MF 8 N-way SMP 6 4 2 Clustering i Performance Number of Processors Scalable Perf. vs. Availability Scalable Performance MPP Clustering SMP FT UniP ©2000-2005 IMEX Research Availability IMEX
  • 4. Market Segments by Applications OLTP Transaction 10 K Processing eCommerce Data DSS (RAID - 0, 3) (RAID - 1, 5, 6) 1K Warehousing (Latency) Visual DB IOPs 100 Scientific Computing HPC HPC Imaging TPC 10 Audio Streaming Streaming HPC Video 1 1 5 10 50 100 500 MB/sec *IOs per sesond for a required response time ( ms) ©2000-2005 IMEX Research IMEX
  • 5. HPC – 2 decades of rapid progress 1991 1998 2005 2010 System Cray Y-MP C916 Sun HPC10000 Shuttle@newegg Blade 24x 333 MHz, Ultra 4x 2.6 GHz, x64 Architecture 16x Vector, 4 GB, Sparc II, 24 GB,S- Bit, 4 4x 10GHz, x64 bit, Bus Bus GB, GbE 16GB, 10GbE Win Server 2003 OS UNICOS Solaris 2.5.1 SP2 LongHorn/Linux Gflops ~10 ~10 ~10 ~100 Price $40 million $1 Million <$4000 ~$1000 Price Reduction 1 1/40 1/10,000 1/400,000 Government Small>Large Target Audience National Labs Large Enterprises Businesses Every Professional Bioinformatics, Applications Classified, Climate, Manufacturing, Materials Architects,Wall Physics Energy,Finance Sciences Street, Hollywood ©2000-2005 IMEX Research IMEX
  • 6. The rapid rise of Clusters in HPC Rise of Cluster Computing 350 # Clustered Computers in Top 500.org 300 250 200 150 100 50 0 1999e 2000 2001 2002 2003 2004e ©2000-2005 IMEX Research IMEX
  • 7. HPC Market Drivers • Availability of industry standard HW making HPC mainstream • HSV Servers @$200, Storage @25c/GB, Add-On GbE NICS at $10, GbE Switches at $2-5/port • Migration to Embedded ASICs on Blades • Rapid procurement, installation & System Integration simplifying acquisition • Cluster ready apps accelerating market growth • Engineering (LSTC, Ansys..) • Bioinformatics (Blast, Gaussian..) • Finance (Matlab, Excel…) • Oil & Gas (Eclipse, Promagic..) • Government (Uexplore, NOAA Hysplit..) ©2000-2005 IMEX Research IMEX
  • 8. HPC Servers Rev - 15% of Total Server Market WW HPC Rev ($M) $9,000 $8,000 $7,000 $6,000 Departm ental Rev $M $5,000 $4,000 $3,000 Divisional $2,000 Enterprise $1,000 Capability $- 2003 2004 2005 2006 2007 2008 ©2000-2005 IMEX Research IMEX
  • 9. HPC Market Leaders 2004 Market Leaders ©2000-2005 IMEX Research IMEX
  • 10. Rise of Industry Standard Architecture in HPC X86 Architecture (Intel & AMD) using Linux and Microsoft based clusters dominate in High Perrformance Computing Architecture. TeraFlop HPC Using Off-the-Shelf Modules Intel servers with PCI Express, Infiniband HBAs & Switches, SATA/SAS HDD or AMD Opteron Servers, GbE HBA/Switches, SATA ©2000-2005 IMEX Research IMEX
  • 11. Factors Affecting Performance • Applications Configurations affect Linpack significantly (3 metrics have the most affect) • Problem Size • Size of Blocks • Topology • Tightly Coupled MPI Applications • Very sensitive to network performance characteristics - internodal communications delay - OS Network Stack is a significant factor • Very sensitive to mismatched node performance - Random OS activities can add msec delays to usec type communication line delays ©2000-2005 IMEX Research IMEX
  • 12. HPC Architecture & Challenges Architecture Goals & Criteria Goals • Maximize compute time HPC Communications Stack vs. messaging time • High constant cross- Entire Stack sectional throughput Needs Attention • Reduce or hide latency Data Application Metrics Upper Layers • Bandwidth TCP/IP (or Offload I/F) • Latency OS • System Interface CPU • SW Stack Memory Chipset Challenge is to move data from I/O interconnect Network Interface application to the network and back (w or w/o TOE) with minimum latency Data ©2000-2005 IMEX Research IMEX
  • 13. Application Application Application Application Application Memory Memory Control Control Host Host Host Socket RDMA CPU RDMA Adapter TCP TCP CPU CPU IP IP Data Data Kernel Kernel Traditional Memory Memory Adapter Ethernet Ethernet RDMA Adapter RDMA Adapter Network ©2000-2005 IMEX Research IMEX
  • 14. Interconnect Price/Performance Hierarchy High MB/s, Low Latency, Price/Port Quadrics Myrinet InfiniBand 10 GbE w/ RDMA, TOE 10 GbE w/ TOE 1 GbE Market Size Microsoft Chimney Software 2004-07 ©2000-2005 IMEX Research IMEX
  • 15. State of Interconnect Fabrics EHERNET - LAN/SAN/MAN/WAN FE GbE ATM IBM/SP Latency/hop MPP SMP/NUMA Servernet Dolphin T3E Myrinet SGI Mellanox IB GbE w/TOE Quadrics FJ/Synfinity GbE w/RDMA 10GbE w/RDMA Bandwidth/link ©2000-2005 IMEX Research IMEX
  • 17. InfiniBand Clustering • InfiniBand and Industry Standard Servers – Displacing Mainframe Computers in the Data Center with Three Standard Hardware Components – Powerful Low Cost Combo Out Performs SMP 12-Nodes (24-CPUs) Up to 72 GFlops of CPU Power; 24+ GB Memory Industry Standard Server InfiniBand Switch => InfiniBand HCA Enterprise Availability: ©2000-2005 IMEX Research With Fabric Failover IMEX
  • 18. InfiniBand rears up its head • Low Cost IB 10G Silicon & Adapters – Very Low Cost 10G Host Adapters • Host Channel Adapter ~ $69 in volume • Switch Si at less than $30 a port in volume • Compelling price/performance • InfiniBand vs. 10GE Advantages in the Short Term – 15X – 20X price advantage vs 10GE – 3X – 10X latency advantage – 10X – 20X CPU utilization advantage – 3X-5X throughput advantage • Adoption by most majors • Products – InfiniBand to IP * – InfiniBand to FC – Advanced Management ©2000-2005 IMEX Research IMEX
  • 19. Ethernet - TCP + iSCSI Offload Engines Typical 1GbE NIC File System File System Host File System SW Volume Manager Volume Manager Volume Manager SCSI Device Driver SCSI Device Driver SCSI Device Driver iSCSI iSCSI iSCSI Adapter TCP TCP TCP IP IP IP Ethernet Ethernet Ethernet Host CPU HBA Firmware HBA Hardware ©2000-2005 IMEX Research IMEX
  • 20. Interconnects: Applicability Grids 100 Mbit 1 Gig 10 Gig 40 Gig Data Center Ethernet Ethernet Ethernet Ethernet LAN SAN Fibre Channel Storage RDMA NICs with iSCSI NAS Storage RDMA NICs Clusters Myrinet InfiniBand Blade Quadrics Chassis Network SMP PCI Express & Advanced Switching Backplanes SunFire Link, IBM Federation <1 Gigabit 1 Gigabit 10 Gigabit >10 Gigabit ©2000-2005 IMEX Research IMEX
  • 21. InfiniBand Adopters • Major providers of InfiniBand Solutions • Enterprise Class Products – InfiniBand to IP – InfiniBand to FC – Advanced Mgmt • Platform support includes – Intel (IA32 + 64EMT, Itanium), AMD Opteron, Power PC, Sparc • Operating Systems – Linux, Windows, AIX, HPUX, Solaris, MAC OS X, VxWorks ©2000-2005 IMEX Research IMEX
  • 22. HPC Interconnect – Leaders Myrinet for the highly latency sensitive applications and GbE for the majority applications Dominate the Interconnect for HPC. Infiniband is rearing to dominate at the Midrange latency applications ©2000-2005 IMEX Research IMEX
  • 24. Data Center Environmentals Prep Temp F Speed fpm Sources: Fluent Airpak, Ansys CFD, ©2000-2005 IMEX Research IMEX
  • 25. HPC embraces a broad spectrum of markets .. High Performance Computing World ©2000-2005 IMEX Research IMEX
  • 26. 3 Tier Computing Infrastructure Application End-Users  Uniform and ubiquitous physical Internet Connectivity – “Wire Once”  Any user to any server via Internet Tier –1  Any server to any server via the LAN Network Services:  Any sever to any storage via the SAN - Web Servers,  Dynamic Logical Binding of… - Firewalls, - Load balancers  server network identification LAN  server OS version  server application assignment Tier-2  application data volumes Application Services: - Application Servers  Resulting in  Application Mobility across servers  Data mobility across storage Tier-3 SAN Data Base Services - Database Servers - Storage Storage ©2000-2005 IMEX Research IMEX
  • 27. E2E Internet & 3 Tier Computing HA, Secure Data Center FC SAN IP SAN NAS NAS IP Storage Network Web App Internet Web DB Edge Core Web App Optical Web DB Web App Web Web DB App Web Edge Applications Data Center Directory Security Policy Management Software OS Platform ©2000-2005 IMEX Research IMEX
  • 28. Fabric based Clusters Architecture Ethernet, Network Fabric Wi-Fi. System Fabric Quadrics, Management Myrinet, SNMP, SCI Fabric IETF/ InfiniBand, CIM, Ethernet / IP, SMI-S, Ethernet IP SMASH. w/TOE, Ethernet IP w/TOE and RDMA. SCSI, Fiber Channel, ISCSI. ©2000-2005 IMEX Research IMEX
  • 29. Fabric based Grid Applications Storage Systems ©2000-2005 IMEX Research IMEX
  • 30. Scale Up (Consolidated SMP Servers) Server Platforms - Migrations Large SMP Clusters Parallel Sysplex High Availability/ - Scalable Nodes, Hi Perform.Clusters - Dynamic LPAR - Linux/AIX/Solaris Clusters - SW Provisioning/VMWare - HPC Clusters IBM: i890/p690/z990 DELL: PowerEdge 6650 SUN: SunFire 4800, 12K ORACLE: 9iRAC/PE6650 Blade Servers High Density Rack Mount - Highly Integrated Dense Form Factor - Rapid Deployment, Flexible Architect. RLX: ServerBlade 300i, 800i, HP: 40b, p FUJITSU: BX300, DELL:1655 IBM: HS20, Sun: Scale Out (Clustered Blade Servers) ©2000-2005 IMEX Research IMEX
  • 31. Blades Servers - Infrastructure Management Modules Remote Mgmt+KVM over IP Midplane Networking: Gbit Ethernet Switches W/Connectors Storage: IP NAS or FC SAN To Blades & Back Switch Cooling N+1Fans/Cooling Modules Processor Blades Modules (6-24 typically) Power: N+1Power Supplies Blade Control Panel USB Ports, CD/Floppy I/F ~ GbE Switch supports Trunking/Port Aggregation Flow Control Gbit QoS Packet Prioritization Memory Ethernet SNMP/RMON Systems I/F Micro- DDR w IGMP/BOOTP/TFTP Monitor Processors ECC …… Module © 2003 Source: IMEX Research ©2000-2005 IMEX Research IMEX
  • 32. Management – Blade Servers Ultra-dense HW • Density 20 % • Power Consumption 20% • Cable Management Management • Platforms integration:  Hardware  Software 80%  Network  Management • Scalability 80 % • Security • Manageability • Lowest TCO ©2000-2005 IMEX Research IMEX
  • 33. Blade Servers Market Opportunity $12 WW Blade Servers $10 Market Opportunity (cum 2004-2008) Factory Rev $B $8 Revenues ~ $24 Billion Units ~ 9 Million $6 $4 $2 $- 2003 2004 2005 2006 2007 2008 Data: IMEX Research 2005 ©2000-2005 IMEX Research IMEX
  • 34. Blade Servers: Vendor Positioning Index (As of 31 Mar 2004 - See IMEX Blade Servers Industry Report 2005 for new data) Strategy (Potential) IBM Fujitsu Siemens Hitachi HP Linux Intel © 2003-05 IMEX Research NEC Networx Dell Nexcom Tatung Sun Appro Egenera Penguin Rackable Sys HPC Syst Verari ©2000-2005 IMEX Research Delivery (Execution) IMEX
  • 35. iSCSI on roadmap of storage vendors SAN iSCSI NAS (Fibre COST (IP) (IP) Channel) •iSCSI delivers low-cost Networked Storage –Higher availability,Increased usage High End –Centralized management,Storage functionality FC & iSCSI Mid-Tier FC & iSCSI Mid-Tier NAS & iSCSI Entry FC & iSCSI SERVICE LEVEL ©2000-2005 IMEX Research IMEX
  • 36. System Cost vs. High Availability z/390 $10 M Sysplex Avg. S/390 S/390 Syst z/390 MVS MVS IBM IBM Price Propl Propl $1 M Prop. Sun-Solaris Sun-Solaris HP-UX HP-UX IBM IBM IBM-AIX IBM-AIX $100 K Clustered NCR-SVR4 NCR-SVR4 Bull Bull Dell Dell UNIX AS400 Fujitsu Fujitsu HP HP IBM IBM NEC NEC Siemens Siemens Stratus Stratus UNIX Cluster Win Win $10K Linux Clustered Linux 0.1 1 10 100 Downtime Hrs./Syst/Yr 99.999% 99.99% 99.9% 99.0% System Availability ©2000-2005 IMEX Research IMEX
  • 37. The iSCSI SAN Client Workstations Client requests data from App Server LAN Ethernet Header IP TCP DATA CRC Ethernet Switches Application & DB Servers (iSCSI Initiator) iSCSI-equipped App Server (“initiator”) requests data from an iSCSI storage (“target”) SAN Ethernet CRC IP TCP iSCSI SCSI DATA GigE Header Switches Storage Arrays (iSCSI Target) To Other LANs, Servers or IP SANs ©2000-2005 IMEX Research IMEX
  • 38. Tiered Storage by Price/Performance Desktop PC Servers Large Enterprise Entry NAS Midline Storage SAN & NAS Workstations Mainstream NAS/SAN PC SAS SAS HBA/RAID FC RAID Chipset HBA Drive Interface SATA SAS SAS Drive Types 5400 RPM 10K RPM 7200 RPM 15K RPM SATA Drive SAS Drive FC Drive Source:Maxtor ©2000-2005 IMEX Research IMEX
  • 39. Cost Driven Data Storage Markets 100% 90% SATA 80% 70% 60% SAS 50% 40% 30% FC 20% 10% P_SCSI 0% 2003 2004 2005 2006 2007 2008 ©2000-2005 IMEX Research IMEX
  • 40. Grid Computing • Grid is a Catch-all Marketing Term • Means different things to different constituencies • Desktop Cycle-Stealing • Manahed HPC Clusters • Virtualization of Data Center Resources • Outsourcing to “Utility Data Centers” • Multiple, opposing requirements • Compute-intensive applications want to ship data tp idle processor resources • Data-Intensive applications want to ship computations to appropriate large data repositories • An evolving entity with many pulls & pushes • Economics, Feasibilities, management of hetero vs homogeneous HW ©2000-2005 IMEX Research IMEX
  • 41. Grid Challenges: Hype vs. Reality • Cost of Local Grids (LAGs) – ala Blades cheaper than Wide Area Grids (WAGs) • Computing Cost $1000 > 1 cpu day (10 Terops) =$1 • 10 TB network transfer Cost = $1 • Internet BW cost $100/mbps/mo > 1GB network xfr cost ~$1 • Result: Local HPC Cluster is 10,000 cheaper than WAN Communication • MPI Style Apps work well in LAN Clusters but uneconomical in WAN • Data analysis best done by moving programs to data not data to programs • Small data, high compute applications woek well across internet • Internet is not the CPU back plane ©2000-2005 IMEX Research IMEX
  • 42. Grids – Homogeneous vs Heterogeneous ©2000-2005 IMEX Research IMEX
  • 43. Grid Services Architecture & Standards ©2000-2005 IMEX Research IMEX
  • 44. Grid Computing – Across 3 Tiers Grid Control Storage Databases Application Servers ©2000-2005 IMEX Research IMEX
  • 45. Utility Computing/Web Services Autonomics • Self Healing • Self Optimizing • Universal Identity • Single System Image • …. Internet Internet Network Switching Pool • HP - Utility Data Center Utility Controller Virtualization Load Balancer Pool • Sun - N1/SunONE/Orion Firewall Pool Server Virtualization • Microsoft - .Net Server Pool NAS Pool • IBM - OnDemand/Autonomics Storage Switching Pool Virtualization Storage Pool ©2000-2005 IMEX Research IMEX
  • 46. Future Computing: Processes Driven Initiator Initiator Initiator Application Application Application Old model New model Hardware is the initiator Processes are the initiators HBA ID Initiator Allows applications to move storage wherever they are - storage no longer cares where they are Target ©2000-2005 IMEX Research IMEX
  • 47. Application-Driven Storage Virtualization Application Application Application Moving from storage ...to storage associated associated with with application servers... processes Old Model New Model Server Hardware is the Allows applications to initiator. Storage move storage from the Drives are the Target virtualized pool, wherever they may be ©2000-2005 IMEX Research IMEX
  • 48. Future: Embedded10Gb,TOE, iSCSI CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU CPU TOE MC MC MC MC MC MC MC MC DMA DMA PCI Exp PCI Exp PCI Exp PCI Exp PCI Exp I/O Uses I/O Uses I/O 1GbE Bridge Bridge Syst Bridge Syst 10 GbE Mmry Mmry 1GbE TOE GbE TOE 1GbE GbE 1GbE PCI-X Ext TOE Mmry PCI PCI-X TOE PCI PCI-X with 10GbE 2GbE Ext.Mmry 2x1Gb TOE NIC TOE chipset 10 Gb TOE NIC Embedded 2003 2004 2005 2006 onwards ©2000-2005 IMEX Research IMEX
  • 49. Future: IP Storage Management on a chip Host Services Integration Storage Provisioning File system monitoring Storage provisioning Win, LINUX, Solaris Layer IP SAN Management Management Management of MultiPath IO Supp Security IP SAN Console iSCSI HBAs and Failover (iSNS, CHAP, SRP) Management Layer Virtualization Mirroring Snapshot Fail-Over iSCSI Target Management Appliance LVM, Error Handling, SCSI Daemon, API Interoperability Service Layer HW Acceleration: TOE, iSCSI Offload, IPsec ©2000-2005 IMEX Research IMEX
  • 50. Future Intelligent IP Infrastructure Present Data Center Infrastructue Future Infrastructue Clients LAN Ntwk Mgmt WS Application & Intelligent DB Servers Uniform IP-based Infrastructure Mgmt SAN Storage WS Mgmt WS Storage Sub Systems ©2000-2005 IMEX Research IMEX
  • 51. Future - IP Based Infrastructure Servers Storage IP Networks Infrastructure Telecom Virtualization Provisioning Automation ©2000-2005 IMEX Research IMEX
  • 52. HPC – From Academia to Wall St to Hollywood High Performance Entertainment High Performance Commercial Bioinformatics Decision-Support Entertainment Commercial BioInformatics Decision Audio/Video OnDemand Computing Computing Visualization Visualization Support Syst Audio/Video On Demand Systems Data: IMEX Research & Panasas 100+ Teraflops Rendering (Texture & Polygons) Data rate & capacity Throughput = 100 GB/s Throughput = 1.2 GB/s Throughput : DSL/Cable ©2000-2005 IMEX Research IMEX ©2004--05 IMEX Research.com
  • 53. HPC: From Homeland Security to Bioinformatics ©2000-2005 IMEX Research IMEX
  • 54. HPC From Academia to Hollywood ©2000-2005 IMEX Research IMEX
  • 55. High Performance Computing State of the Industry Anil Vasudeva Principal Analyst & President anil@imexresearch.com IMEX RESEARCH.COM 408-268-0800 ©2000-2005 IMEX Research IMEX
  • 56. State of Interconnect Fabrics Latency/hop 1 ms EHERNET - LAN/SAN/MAN/WAN FE GbE IB - CLUSTER ATM IBM/SP MPP 100 us Servernet Dolphin SMP/NUMA T3E Myrinet SGI GbE Quadrics Mellanox w/TOE 10 us FJ/Synfinity Synfinity-1 GbE 10GbE w/RDMA Scali w/RDMA 1us 100 Mb/sec 1Gb/sec 10 Gb/sec Bandwidth/link ©2000-2005 IMEX Research IMEX