Power up for POWER5™

IBM Systems Group

Power Up for POWER5™

Larry Amy pSeries Offering Manager June 16, 2004

IBM Confidential until announcement

This presentation is intended for the education of IBM and Business Partner sales personnel. It should not be distributed to customers. IBM Systems Group

Agenda

❧Introduction ❧POWER5™ processor technology ❧p5 systems ❧p5 virtualization technologies ❧p5 operating system environments

Introduction

This presentation is intended for the education of IBM and Business Partner sales personnel. It should not be distributed to customers. IBM Systems Group

Customer Challenges

CFO User Expectations Reduce Risk, Quick ROI New Applications Improve productivity More throughput Variable Costs Better Service Increase Business Value Improved Security Improve security High Availability

Manage More

With Less

Do it quickly

Be Flexible Technology Java, .NET Application Integration SAN / Tiered Storage Server Consolidation

HowHow to to rapidly rapidly deploy deploy and and maintain maintain applicationsapplications that that drive drive business business advantage! advantage!

Manage Increasing Amounts of Risk Increase Resource Utilization Lowest Total Cost of Ownership or Cost of Aggressive SLA’s including Operations Security and Disaster Recovery

BALANCING TOOLS !Enterprise RAS Systems and Components !Utility Computing--pricing /services !Resource Partitioning—finely grained !Autonomic Management and Provisioning

Make no compromises, accept no limitations

❧ IBM: The fastest growing UNIX® vendor in the industry* – Proven strategy – Gains in UNIX market share and revenue growth

❧ Technology innovation and solid roadmaps make IBM eServer pSeries the right choice for today and tomorrow

❧ Performance, RAS, flexibility, and The UNIX platform with momentum today technology are no longer trade-offs with price

*IDC Worldwide Quarterly Server Tracker for Q3 2002, Q4 2002, Q1 2003, Q2 2003 , Q3 2003 and Q4 2003

You’ll never look at computing the same way again

Jaw-dropping performance IBM eServer™ p5 520, 550, 570

Plus breakthrough virtualization innovations that will help you

" Lower costs " Improve flexibility " Improve manageability " Safeguard data " and be prepared for what tomorrow brings through a more simplified infrastructure

Announcing IBM p5 family

Next-generation IBM eServer p5 systems designed to deliver superior performance, flexibility and reliability

p5-520 p5-550 p5-570 Up to 2 processors Up to 4 processors Unique Building Block Up to 32GB memory Up to 64GB memory architecture for ‘pay as you grow’ 4 processors per block 12 IBM mainframe inspired Up to 128GB memory per block RAS features throughout

POWER5™

This presentation is intended for the education of IBM and Business Partner sales personnel. It should not be distributed to customers. IBM Systems Group POWER5™ Systems --- A New Standard POWER5 Systems --- A New Standard

❧ Second generation dual core chip ❧ Intelligent 2-way SMT ❧ Power management with no performance impact Shared L2

Mem CTL ❧ Small, dense chips

Distributed switch

276 million transistors 8 layers of metal 130 nm lithography (0.13 microns) 389 sq/mm die size

Modifications to POWER4™ to create POWER5™

POWER4 POWER5 P P P P P P P P

L3 L3 L2 L2 L3 Ctl L2 L2 Ctl L3

Fab Ctl Fab Ctl Fab Ctl Fab Ctl

L3 L3 Cntrl Cntrl Mem Ctl Mem Ctl Larger L2 & L3 Reduced Latencies Faster Memory Access L3 L3 64-way support

Mem Ctl Mem Ctl

Memory Memory Memory Memory

POWER4™ / POWER5™ differences

POWER4 Design POWER5 Design Benefit

2-way associative 4-way associative Improved L1 cache L1 Cache FIFO LRU performance

8-way associative 10-way associative Fewer L2 cache misses L2 cache 1.44MB 1.9MB Better performance

32MB 36MB Better cache performance L3 Cache 8-way associative 12-way associative 40% improvement 118 clock cycles ~80 clock cycles 4X improvement Memory Bandwidth 4GB / sec / chip ~16GB / sec / chip Faster memory access

Simultaneous Better processor utilization No Yes Multi-Threading 40% System improvement

Better usage of processor Processor Addressing 1 processor 1/10 of processor resources

Chip Interconnect Type Distributed switch Enhanced dist. switch Better systems throughput Intra MCM data bus ½ Proc. speed Processor speed Better performance Inter MCM data bus ½ Proc. speed ½ Proc. speed 50% more transistors Size 412mm 389mm in the same space

Enhanced performance through simultaneous multi-threading (SMT)

No multi-threading Coarse/fine-grained Simultaneous Execution units in a multi-threading multi-threading processor are dedicated Execution units can be Execution units are to a single instruction made available to more than simultaneously available to stream one instruction stream during two instruction streams alternate cycles during the same cycle

CYCLES CYCLES CYCLES

Instruction Stream 1 Instruction Stream 2

IBM POWER technology roadmap for eServer Planned * PlannedPlanned * * Planned Planned * * 2001 2002-4 2004 2005-6 2006-7 POWER4 POWER4+ POWER5 POWER5+ POWER6 90 nm 65 nm 130 nm 130 nm 180 nm >> GHz >> GHz Ultra-high frequency cores 1.5+ 1.5+ > GHz > GHz Core Core GHz GHz Core Core L2 caches 1+ GHz 1+ GHz Shared L2 Core Core Advanced Core Core Shared L2 Distributed Switch System Features Shared L2 Distributed Switch Shared L2 Distributed Switch •Simultaneous multi-threading (SMT) Distributed Switch •Reduced size •Micro-partitioning •Lower power •Dynamic firmware updates •Larger L2 •Enhanced scalability, parallelism •Chip multi-processing •More LPARs (32) - Distributed switch •High throughput performance - Shared L2 •Enhanced cache/memory subsystem •Dynamic LPARs (16) * All statements regarding IBM's future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only. Autonomic Computing Enhancements

LogicalLogical (SMT) (SMT)

LPARLPAR LPARLPAR LPARLPAR VirtualVirtual

SharedShared

DedicatedDedicated

InactiveInactive (CoD) (CoD)

DeconfiguredDeconfigured

Physical (Installed)

This presentation is intended for the education of IBM and Business Partner sales personnel. It should not be distributed to customers. IBM Systems Group pSeries POWER5™ lineup with new virtualization capabilities in 3rd quarter 2004

POWER5 mid-range POWER5 4-way POWER5 2-way system

• Up to 2 processors • Up to 4 processors • Up to 16 processors • Up to 32GB memory • Up to 64GB memory • Up to 512GB memory • Micro-partitioning • Micro-partitioning • Micro-partitioning – Up to 20 partitions – Up to 40 partitions – Up to 160 partitions • Virtual LAN, Storage • Virtual LAN, Storage • Virtual LAN, Storage • On demand options • Building Blocks • SMP (not NUMA) • On demand options

POWER4 versus POWER5 Systems

POWER4 POWER5

p690

p5-570 p670 16-way

p5-570 12-way p655 p5-570 8-way

p650 p5-570 2- or 4-way

p630 p5-550 1-, 2- or 4-way

p5-520 p615 1- or 2-way

1 / 2 – way Architecture POWER5™

L3 Cache 36MB

Memory .5GB – 32GB

Packaging Deskside / 4U ( 19” rack )

DASD / Bays 8 DASD (Hot Plug)

I/O Expansion 6 slots (Hot Plug)

Integrated SCSI Dual Ultra 320

Internal RAID Optional

Integrated Ethernet Dual Ports 10/100/1000 Rack: 6.8"H x 17.4"W x 22.6" D Tower: 21.1"H x 7.9"W x 23.2"D Media Bays 3

RIO2 Drawers Yes / 4

Dynamic LPAR 20

Redundant Power Feature

Redundant Cooling Yes Linux 5.2 and/or 5.3

1 , 2, 4 – way Architecture POWER5™

L3 Cache 36MB

Memory 1GB – 64GB

Packaging Deskside / 4U ( 19” rack )

DASD / Bays 8 DASD (Hot Plug)

I/O Expansion 5 / 4 slots (Hot Plug)

Integrated SCSI Dual Ultra 320

Internal RAID Optional

Integrated Ethernet Dual Ports 10/100/1000 Rack: 6.8"H x 17.4"W x 25.1" D Tower: 20.0"H x 6.8"W x 25.1"D Media Bays 3

RIO2 Drawers Yes / 8

Dynamic LPAR 40

Redundant Power Feature Linux Redundant Cooling Yes 5.2 and/or 5.3

4 – way Building Block Architecture POWER5™

L3 Cache 36MB

Memory* 2GB – 128GB (Base System)

Packaging 4U ( 19” rack )

DASD / Bays 6 DASD (Base system)

I/O Expansion 6 / 5 slots (Hot Plug)

Integrated SCSI Dual Ultra 320 (Base System)

Internal RAID Optional

Dual Ports 10/100/1000 Integrated Ethernet (Base System)

Rack: 6.8"H x 17.4"W x 28" D Media Bays 2 (Base System)

RIO2 Drawers Yes / 8 (Base System)

Dynamic LPAR 40 (Base System)

Redundant Power Yes Linux Redundant Cooling Yes 5.2 and/or 5.3

AIX / Linux POWER5 servers Rack mounted building blocks Building Block Systems ! 4U Building Block 16-way # 4 POWER5 processors, 2 DCMs ! System # 1-4 Units $Max 16 processors # SMP not NUMA 12-way # High Speed Interconnect Bus ! I/O # Integrated I/O 8-way # Optional I/O drawers ! Dynamic LPAR # Up to 160 partitions ! OS Support # AIX and Linux

Linux

This presentation is intended for the education of IBM and Business Partner sales personnel. It should not be distributed to customers. IBM Systems Group

IBM Virtualization Engine™

IBM Virtualization Engine™ Suite for Servers VE IBM Virtualization Engine VE Console Technologies Director Multiplatform (Integrated in Systems Hardware) Enterprise Workload Manager Power Hypervisor, VLANs, Systems Provisioning Virtual Storage, Shared Processor IBM Grid Toolbox Pool, Micro-partitioning

* Components are separately orderable as needed by client IBM Virtualization Engine™ Offerings for Storage* IBM TotalStorage SAN Volume Controller IBM TotalStorage SAN File System IBM TotalStorage Productivity Center

* Storage elements are part of IBM TotalStorage Open Software Family Components are separately orderable as needed by client

Lab Integrated & Tested

Powerful capabilities across the entire POWER5™ portfolio

New capabilities delivered with POWER5™

% Micro-partitioning % Virtual Storage % Virtual LAN support % Advanced Accounting / Chargeback % CoD Enhancements % Cross-LPAR Workload Management % Building block approach for midrange % Selective concurrent firmware upgrade

Virtualization Features

Micro-partitioning Available to all p5 systems Shared processor pool Requires AIX 5L 5.3 or

Virtual Storage appropriate Linux version

Virtual LAN Enables resource balancing

AIX 5L AIX 5LLinux Linux I/O Server

AIX 5L AIX 5L Linux Linux kernel kernel kernel kernel

Virtual I/O Virtual I/O Virtual I/O Virtual LAN Virtual I/O POWER Hypervisor™ Hardware Management LAN LAN Console IOA IOA IOA IOA IOA IOA (HMC) LAN

26 IBM Confidential until announced © 2004 IBM Corporation IBM Systems Group Field Skills & Education IBM Systems Group Micro-partitioning means systems resource allocation can be fine-tuned to changing business priorities

With Partitioning

Today Tomorrow* SMP partitioning: POWER5 SMP Without Partitioning One partition per CPU micro-partitioning: Multiple partitions At Full Capacity activated in each CPU

CPU 1 Dedicated, CPU 0 underutilized resources CPU 2 CPU 3

1 to 4 partitions Up to 40 in a 4-way SMP micro-partitions in a 4-way SMP

IBM furthers simplification through shared resources

Unique advanced virtualization features eliminate the need for separate Ethernet and storage devices for each application. – Virtual LAN enables high-speed, secure internal communications – Virtual Storage enables logical sharing of external devices in support of multiple applications

Logical Micro

• Buy what you need • Simplify your environment • Respond more quickly to business dynamics

DNS Server Avg. Util. 10% File/Print Srver Avg. Util. 10% DB Server Avg. Util. 10% DataBaseDataBase ApplicationApplication WebWeb Email Server DNS Avg. Util. 10% 7 specialized File/Print servers to 1 File/Print Application Server Email Avg. Util. 20% Firewall Firewall Web Server Avg. Util. 10% Avg. Util. 10%

Virtualization is the key to effective on demand environments

Virtual servers can be created rapidly. Fast application deployment

Resources can be increased as growth Non-disruptive business demands. growth

Resources can be increased or Cost-effective approach to decreased dynamically. workload peaks

Resource utilization can be increased Improved efficiency and greatly, allocating resources to when and reduced IT costs where needed.

IBM eServer p5 HP Sun

Logical partitioning YES PA only No

Advanced virtualization technologies Micro-partitioning YES No No Virtual storage YES No No Virtual LAN YES No Limited Cross-partition workload management YES No No

CoD options Trial, On/Off YES Yes Yes Backup YES No No Reserve YES No No

Modular “pay as you grow” architecture YES No No

Run multiple operating systems YES Limited No

Single binary compatible architecture YES No No

This presentation is intended for the education of IBM and Business Partner sales personnel. It should not be distributed to customers. IBM Systems Group pSeries Software Flexibility

Linux on pSeries

Linux

❧ World fastest growing UNIX ❧ Open source momentum ❧ UNIX scalability, low TCO ❧ 64-bit platform advantages ❧ Enterprise class RAS, performance ❧ Growing performance & ❧ Bulletproof security scalability ❧ ❧ Flexible Workload Manager Reliability architecture ❧ ❧ Cluster management Large memory ❧ ❧ Dynamic LPAR Logical partitioning (LPAR) ❧ ❧ Keyed Capacity Upgrade on Demand Start small, expand easily ❧ ❧ Linux affinity Investment protection ❧ Linux affinity

AIX , Linux or both….

POWER5 HW support More solution choices 64-way SMP Leading scalability & performance SMT processor support Even more scalability Micro-partitioning Even more resource flexibility Up to 254 partitions Server consolidation File system size Shrink the file system size Virtual LAN Inter-partition communication Virtual Storage Storage support for 100’s of LPAR eWLM enabling Cross partition workload mgmt Interval-based accounting Provide accurate billing information NFSV4 (staged) Next generation network filesystem IP over FC Additional connectivity

"Through 2008, the server operating system shakeout will continue, and the server microprocessor shakeout will accelerate. As a result of these shakeouts, we are predicting the following: &There will only be three growth server operating systems (Windows, Linux and AIX) through 2008 (0.8 probability). &There will only be three growth server microprocessor architectures (IA-32, IPF and Power) through 2008 (0.8 probability). Linux will grow market share significantly, but primarily in two different market segments &massively parallel computing (supercomputing) &single-function commodity servers. "By YE08, AIX will grow market share approaching or passing Solaris as the No. 1 Unix operating system (0.8 probability). "The Itanium Processor Family (IPF) architecture will grow, but slowly. IPF server revenue will not cross the $1 billion revenue mark until 2005 (0.8 probability).

Source: Gartner Cannes Symposium November 2003

LinuxLinux App App Infrastructure Solutions ServersServers

Service Providers Directory & Security Services Transaction Web Servers Intranets Application Data and Extranets Web Servers Transaction Presentation Servers Data Server Servers

Storage

Internet Inter Partition Communication

Large-scale Server Solutions and Consolidations Deep Computing Solutions

Linux Images Linux Linux Virtual Virtual Servers Servers

Server farms OS OS

Linux Inter Partition Communication Images

Server farms

Single purpose Internet-related servers

POWER5™ RAS

%First Failure Data Capture Designed for worry-free %DDR Chipkill™ memory continuous computing %Bit-steering/redundant memory %Memory soft scrubbing %Redundant power, fans

%Dynamic processor deallocation %ECC memory %Persistent memory deallocation %Hot-plug PCI slots, fans, power %Internal light path diagnostics %Hot-swappable disk bays %Concurrent firmware updates

IBM eServer p5 enables you to change the economics of your computing environment

1 Reduce the number of servers required to meet processing needs, minimizing overall costs through micro-partitioning

Eliminate redundant (and costly) adapters and storage 2 through advanced I/O and LAN technologies

3 Increase IT flexibility and therefore business flexibility by leveraging your choice of Capacity on Demand options

IBM-only ‘building block’ features in mid-range enable you to buy additional permanent processing power only when you 4 need it

Make no compromises and accept no limitations in an on demand world

Simplify your computing environment and… Safeguard your data and infrastructure to… Enable rapid business innovation

Without compromising the things you expect from IBM " Performance " Flexibility " Reliability " Security

Special Notices

This document was developed for IBM offerings in the United States as of the date of publication. IBM may not make these offerings available in other countries, and the information is subject to change without notice. Consult your local IBM business contact for information on the IBM offerings available in your area.

Information in this document concerning non-IBM products was obtained from the suppliers of these products or other public sources. Questions on the capabilities of non-IBM products should be addressed to the suppliers of those products. IBM may have patents or pending patent applications covering subject matter in this document. The furnishing of this document does not give you any license to these patents. Send license inquires, in writing, to IBM Director of Licensing, IBM Corporation, New Castle Drive, Armonk, NY 10504- 1785 USA.

All statements regarding IBM future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only. The information contained in this document has not been submitted to any formal IBM test and is provided "AS IS" with no warranties or guarantees either expressed or implied.

All examples cited or described in this document are presented as illustrations of the manner in which some IBM products can be used and the results that may be achieved. Actual environmental costs and performance characteristics will vary depending on individual client configurations and conditions.

IBM Global Financing offerings are provided through IBM Credit Corporation in the United States and other IBM subsidiaries and divisions worldwide to qualified commercial and government clients. Rates are based on a client's credit rating, financing terms, offering type, equipment type and options, and may vary by country. Other restrictions may apply. Rates and offerings are subject to change, extension or withdrawal without notice.

IBM is not responsible for printing errors in this document that result in pricing or information inaccuracies. All prices shown are IBM's United States suggested list prices and are subject to change without notice; reseller prices may vary. IBM hardware products are manufactured from new parts, or new and serviceable used parts. Regardless, our warranty terms apply. Many of the pSeries features described in this document are operating system dependent and may not be available on Linux. For more information, please check: http://www.ibm.com/servers/eserver/pseries/linux/whitepapers/linux_pseries.html.

Any performance data contained in this document was determined in a controlled environment. Actual results may vary significantly and are dependent on many factors including system hardware configuration and software design and configuration. Some measurements quoted in this document may have been made on development-level systems. There is no guarantee these measurements will be the same on generally-available systems. Some measurements quoted in this document may have been estimated through extrapolation. Users of this document should verify the applicable data for their specific environment. Revised February 6, 2004

Special Notices (Cont.)

The following terms are registered trademarks of International Business Machines Corporation in the United States and/or other countries: AIX, AIX/L, AIX/L(logo), alphaWorks, AS/400, Blue Gene, Blue Lightning, C Set++, CICS, CICS/6000, CT/2, DataHub, DataJoiner, DB2, DEEP BLUE, developerWorks, DFDSM, DirectTalk, DYNIX, DYNIX/ptx, e business(logo), e(logo)business, e(logo)server, Enterprise Storage Server, ESCON, FlashCopy, GDDM, IBM, IBM(logo), ibm.com, IBM TotalStorage Proven, IntelliStation, IQ-Link, LANStreamer, LoadLeveler, Lotus, Lotus Notes, Lotusphere, Magstar, MediaStreamer, Micro Channel, MQSeries, Net.Data, Netfinity, NetView, Network Station, Notes, NUMA-Q, Operating System/2, Operating System/400, OS/2, OS/390, OS/400, Parallel Sysplex, PartnerLink, PartnerWorld, POWERparallel, PowerPC, PowerPC(logo), Predictive Failure Analysis, pSeries, PTX, ptx/ADMIN, RISC System/6000, RS/6000, S/390, Scalable POWERparallel Systems, SecureWay, Sequent, ServerProven, SP1, SP2, SpaceBall, System/390, The Engines of e-business, THINK, ThinkPad, Tivoli, Tivoli(logo), Tivoli Management Environment, Tivoli Ready(logo), TME, TotalStorage, TURBOWAYS, VisualAge, WebSphere, xSeries, z/OS, zSeries.

The following terms are trademarks of International Business Machines Corporation in the United States and/or other countries: AIX/L(logo), AIX 5L, AIX PVMe, AS/400e, BladeCenter, Chipkill, Cloudscape, DB2 OLAP Server, DB2 Universal Database, DFDSM, DFSORT, Domino, e-business(logo), e-business on demand, eServer, GigaProcessor, HACMP, HACMP/6000, i5/OS, IBMLink, IBM Virtualization Engine, IMS, Intelligent Miner, iSeries, NUMACenter, POWER, POWER Hypervisor, Power Architecture, Power Everywhere, POWER Hypervisor, PowerPC Architecture, PowerPC 603, PowerPC 603e, PowerPC 604, PowerPC 750, POWER2, POWER2 Architecture, POWER3, POWER4, POWER4+, POWER5, POWER5+, POWER6, Redbooks, Sequent (logo), SequentLINK, Server Advantage, ServeRAID, Service Director, SmoothStart, SP, S/390 Parallel Enterprise Server, ThinkVision, Tivoli Enterprise, TME 10, TotalStorage Proven, Ultramedia, VideoCharger, Visualization Data Explorer, X-Architecture, z/Architecture.

A full list of U.S. trademarks owned by IBM may be found at: http://www.ibm.com/legal/copytrade.shtml.

UNIX is a registered trademark in the United States and other countries licensed exclusively through The Open Group.

Microsoft, Windows, Windows NT and the Windows logo are registered trademarks of Microsoft Corporation in the United States and/or other countries.

Intel, Itanium and Pentium are registered trademarks and Intel Xeon and MMX are trademarks of Intel Corporation in the United States and/or other countries.

AMD Opteron is a trademark of Advanced Micro Devices, Inc.

Java and all Java-based trademarks and logos are trademarks of Sun Microsystems, Inc. in the United States and/or other countries.

Other company, product and service names may be trademarks or service marks of others. Revised May 3, 2004

Notes on Benchmarks and Values

The benchmarks and values shown herein were derived using particular, well configured, development-level computer systems. Unless otherwise indicated for a system, the values were derived using external cache, if external cache is supported on the system. Buyers should consult other sources of information to evaluate the performance of systems they are considering buying and should consider conducting application oriented testing. For additional information about the benchmarks, values and systems tested, contact your local IBM office or IBM authorized reseller or access the following on the Web:

TPC http://www.tpc.orgLinpack http://www.netlib.no/netlib/benchmark/performance.ps Pro/E http://www.proe.com SPEC http://www.spec.org GPC http://www.spec.org/gpc NotesBench Mail http://www.notesbench.org VolanoMark http://www.volano.com STREAM http://www.cs.virginia.edu/stream/

Unless otherwise indicated for a system, the performance benchmarks were conducted using AIX V4.3 or AIX 5L. IBM C Set++ for AIX and IBM XL FORTRAN for AIX with optimization were the compilers used in the benchmark tests. The preprocessors used in some benchmark tests include KAP 3.2 for FORTRAN and KAP/C 1.4.2 from Kuck & Associates and VAST-2 v4.01X8 from Pacific-Sierra Research. The preprocessors were purchased separately from these vendors. Other software packages like IBM ESSL for AIX and MASS for AIX were also used in some benchmarks.

The following SPEC and Linpack benchmarks reflect microprocessor, memory architecture, and compiler performance of the tested system (XX is either 95 or 2000):

–SPECintXX - SPEC component-level benchmark that measures integer performance. Result is the geometric mean of eight tests comprising the CINTXX benchmark suite. All of these are written in the C language. SPECint_baseXX is the result of the same tests as CINTXX with a maximum of four compiler flags that must be used in all eight tests.

–SPECint_rateXX - Geometric average of the eight SPEC rates from the SPEC integer tests (CINTXX). SPECint_base_rateXX is the result of the same tests as CINTXX with a maximum of four compiler flags that must be used in all eight tests.

–SPECfpXX - SPEC component-level benchmark that measures floating-point performance. Result is the geometric mean of ten tests, all written in FORTRAN, included in the CFPXX benchmark suite. SPECfp_baseXX is the result of the same tests as CFPXX with a maximum of four compiler flags that must be used in all ten tests.

–SPECfp_rateXX - Geometric average of the ten SPEC rates from SPEC floating-point tests (CFPXX). SPECfp_base_rateXX is the result of the same tests as CFPXX with a maximum of four compiler flags that must be used in all ten tests.

–SPECweb96 - Maximum number of Hypertext Transfer Protocol (HTTP) operations per second achieved on the SPECweb96 benchmark without significant degradation of response time. The Web server software is ZEUS v.1.1 from Zeus Technology Ltd.

–SPECweb99 - Number of conforming, simultaneous connections the Web server can support using a predefined workload. The SPECweb99 test harness emulates clients sending the HTTP requests in the workload over slow Internet connections to the Web server. The Web server software is Zeus from Zeus Technology Ltd.

–SPECweb99_SSL - Number of conforming, simultaneous SSL encryption/decryption connections the Web server can support using a predefined workload. The Web server software is Zeus from Zeus Technology Ltd.

–SPEC OMP2001 - Measures performance based on OpenMP applications.

–SPECsfs97_R1 - Measures speed and request-handling capabilities of NFS (network file server) computers. Revised May 28, 2003

Notes on Benchmarks and Values (Cont.)

–SPECjAppServer200X (where X is 1 or 2) - Measures the performance of Java Enterprise Application Servers using a subset of J2EE APIs in a complete end-to-end Web application.

The Linpack benchmark measures floating-point performance of a system. –Linpack DP (Double Precision) - n=100 is the array size. The results are measured in megaflops (MFLOPS). –Linpack SP (Single Precision) - n=100 is the array size. The results are measured in MFLOPS. –Linpack TPP (Toward Peak Performance) - n=1,000 is the array size. The results are measured in MFLOPS. –Linpack HPC (Highly Parallel Computing) - solves the largest system of linear equations possible. The results are measured in GFLOPS.

STREAM is a simple synthetic benchmark program that measures sustainable memory bandwidth (in MB/s) and the corresponding computation rate for simple vector kernels. Both standard and tuned results may be reported. http://www.cc.virginia.edu/stream/

VolanoMark is a 100% pure Java server benchmark that creates long-lasting network client connections in groups of 20 and measures how long it takes for the clients to take turns broadcasting their messages to the group. The benchmark reports a score as the average number of messages transferred by the server per second.

–The following Transaction Processing Performance Council (TPC) benchmarks reflect the performance of the microprocessor, memory subsystem, disk subsystem, and some portions of the network: –tpmC - TPC Benchmark C throughput measured as the average number of transactions processed per minute during a valid TPC-C configuration run of at least twenty minutes. –$/tpmC - TPC Benchmark C price/performance ratio reflects the estimated five year total cost of ownership for system hardware, software, and maintenance and is determined by dividing such estimated total cost by the tpmC for the system. –QppH is the power metric of TPC-H and is based on a geometric mean of the 17 TPC-H queries, the insert test, and the delete test. It measures the ability of the system to give a single user the best possible response time by harnessing all available resources. QppH is scaled based on database size from 30GB to 10TB. –QthH is the throughput metric of TPC-H and is a classical throughput measurement characterizing the ability of the system to support a multiuser workload in a balanced way. A number of query users is chosen, each of which must execute the full set of 17 queries in a different order. In the background, there is an update stream running a series of insert/delete operations. QthH is scaled based on the database size from 30GB to 10TB. –$/QphH is the price/performance metric for the TPC-H benchmark where QphH is the geometric mean of QppH and QthH. The price is the five-year cost of ownership for the tested configuration and includes maintenance and software support.

Revised February 6, 2004

Notes on Benchmarks and Values (Cont.)

The following graphics benchmarks reflect the performance of the microprocessor, memory subsystem, and graphics adapter: –SPECxpc results - Xmark93 is the weighted geometric mean of 447 tests executed in the x11perf suite and is an indicator of 2D graphics performance in an X environment. Larger values indicate better performance. –SPECplb results (graPHIGS) - PLBwire93 and PLBsurf93 are geometric means of literal and optimized Picture Level Benchmark (PLB) tests for 3D wireframe and 3D surface tests, respectively. Larger values indicate better performance. –SPECopc results - Viewperf 7 (3dsmax-01, DRV-08, DX-07, Light-05, ProE-01, UGS-01) and Viewperf 6.1.2 (AWadvs-04, DRV-07, DX-06, Light-04, medMCAD-01, ProCDRS-03) are weighted geometric means of individual viewset metrics. Larger values indicate better performance.

The following graphics benchmarks reflect the performance of the microprocessor, memory subsystem, graphics adapter and disk subsystem. –SPECapc Pro/Engineer 2000i2 results - PROE2000I2_2000370 was developed by the SPECapc committee to measure UNIX and Windows workstations in a comparable real-world environment. Larger numbers indicate better performance.

The NotesBench Mail workload simulates users reading and sending mail. A simulated user will execute a prescribed set of functions 4 times per hour and will generate mail traffic about every 90 minutes. Performance metrics are: –NotesMark - transactions/minute (TPM). –NotesBench users - number of client (user) sessions being simulated by the NotesBench workload. –$/NotesMark - ratio of total system cost divided by the NotesMark (TPM) achieved on the Mail workload. –$/User - ratio of total system cost divided by the number of client sessions successfully simulated for the NotesBench Mail workload measured. Total system cost is the price of the server under test to the client, including hardware, operating system, and Domino Server licenses.

Application Benchmarks –SAP - Benchmark overview information: http:// www.sap-ag.de/solutions/technology/bench.htm; Benchmark White Paper September, 2000; http://www.sap-ag.de/solutions/technology/pdf/50020428.pdf. –PeopleSoft - To get information on PeopleSoft benchmarks, contact PeopleSoft directly or the PeopleSoft/IBM International Competency Center in San Mateo, CA. –Oracle Applications - Benchmark overview information: http://www.oracle.com/apps_benchmark/ –Baan - The Baan benchmark demonstrates the scalability of Baan ERP solutions. The test results provide the number of Baan Reference Users (BRUs) that can be supported on a specific system. BRU is a single on-line user or a batch unit workload. These metrics are consistent with those used internally by both IBM and Baan to size systems. To get more information on Baan benchmarks, go to http://www.ssaglobal.com. –J.D. Edwards Applications - Product overview information at http://www.jdedwards.com.

Revised December 9, 2003

Notes on Performance Estimates

rPerf

– rPerf (Relative Performance) is an estimate of commercial processing performance relative to other pSeries systems. It is derived from an IBM analytical model which uses characteristics from IBM internal workloads, TPC and SPEC benchmarks. The rPerf model is not intended to represent any specific public benchmark results and should not be reasonably used in that way. The model simulates some of the system operations such as CPU, cache and maximum memory available. However, the model does not simulate disk or network I/O operations.

– rPerf estimates are calculated based on systems with maximum memory and the latest levels of AIX 5L and other pertinent software at the time of system announcement. Actual performance will vary based on configuration details. The IBM ~ pSeries 640 is the baseline reference system and has a value of 1.0. Although rPerf may be used to compare estimated IBM UNIX commercial processing performance, actual system performance may vary and is dependent upon many factors including system hardware configuration and software design and configuration.

– All performance estimates are provided "AS IS" and no warranties or guarantees are expressed or implied by IBM. Buyers should consult other sources of information, including system benchmarks, to evaluate the performance of a system they are considering buying. For additional information about rPerf, contact your local IBM office or IBM authorized reseller.

Revised April 1, 2004