AMD Opteron 6200 and 4200 Series Presentation

THE CORE OF THE CLOUD AMD Opteron™ 6200 and 4200 Series Processors

2 | AMD Opteron™ 6200/4200 Series Processors Launch Presentation | Under Embargo until 12:01 a.m. EST, November 14, 2011 |

THE UNDENIABLE BUSINESS AND I.T . UPLIFT

FROM CLOUD-BASED ECONOMICS

Mainframes

and Minis

UNIX and

PCs Virtualized

and Cloud

Distributed

Traditional IT Cloud deployments accelerate

gains in business agility and

efficiency while speeding the

rate of information exchange

Clouds

TIMELINE


SUCCESSFUL NEW SERVER PLATFORMS MUST BALANCE

WORKLOADS AND DATA CENTER REQUIRMENTS

DATABASE VIRTUALIZATION CLOUD HPC

Scalability for

Virtualization

Economics for

the Cloud

Performance for

data centers


Performance for data centers

Up to 84% higher performance*

Scalability for Virtualization

73% more memory bandwidth*

Economics for the Cloud

1/2 the power-per-core*

Requires 2/3 less floor space*

1/3 to 2/3 lower platform price*

A BETTER OPTION FOR

THE HYPEREFFICIENT,

VIRTUALIZED,

CLOUD-READY WORLD

*See backup processor and performance information on slides 22-23.

AVAILABLE TODAY:

SERVERS FEATURING THE ALL NEW AMD OPTERON PROCESSOR


AMD “BULLDOZER” TECHNOLOGY: BUILT TO WORK FOR SERVER TRENDS

World’s first 16 core x86 CPU is an innovative architecture like nothing else in the market

– Flex FP – the world’s only dynamic, flexible floating point complex

– Revamped integrated memory controller

– AMD Turbo CORE technology for additional performance

– TDP Power Cap

Driving huge benefits to customers

– Outstanding value and power efficiency

– Modular architecture for future innovation

Flexible Shared & Dedicated Resources approach helps maximize power efficiency and costs

– Shared components maximize performance per watt

– Dedicated components to help increase performance and scalability

Dedicated Components

Shared at the module level

Shared at the chip level


INTRODUCING THE NEW AMD OPTERON 6200 AND 4200 SERIES PROCESSORS

AMD Opteron™ 6200 Series Processor

(“Interlagos”)

Scalability Up to 4 sockets with up to 16 cores Up to 2 sockets with up to 8 cores

Memory 4 memory channels up to 1600 MHz memory 2 memory channels up to 1600 MHz memory

Frequency Up to 3.3 GHz base frequency & up to 3.7 GHz frequency using AMD Turbo CORE technology*

Cache

Cache: L1 - 16KB data per core + 64KB instruction

per module;

L2 - 1MB per core; L3 - 16MB per socket

Cache: L1 - 16KB data per core + 64KB instruction

per module; L2 - 1MB per core; L3 – 8MB per

socket

I/O Four x16 HyperTransport™ technology 3.0 links

@ up to 6.4GT/s per link

Three x16 HyperTransport™ technology 3.0 links

@ up to 6.4GT/s per link

Power 85W to 140 W TDP (Consistent with AMD Opteron 6100

Series)

35 to 95W TDP (Consistent with AMD Opteron 4100

Series)

AMD Opteron™ 4200 Series Processor

(“Valencia”)

The world’s first x86

16-core processor1

The world’s lowest x86

power-per-core2

1 Based on 16-core AMD Opteron 6200 Series processor compared to 6-core Intel Xeon 5600 Series and 10-core Intel Xeon E7 processors. 2 As of Nov 1, 2011, AMD Opteron™ processor Models 4200 EE have the lowest known power per core of any x86 server processor, at 35W TDP (35W /8

= 4.375W/core). Intel 's lowest power per core server processor, L5630, is 40W TDP (40W/4 = 10W/core). See

http://www.intel.com/Assets/PDF/prodbrief/323501.pdf. Previous record held by AMD Opteron processor Models 4100 EE at 35W TDP / 6 cores = 5.83

W/core.


WORLD’S MOST CONFIGURABLE CORES FOR AN AGILE CLOUD

*See backup processor and performance information on slides 24.

Processing Flexibility Platform & Virtualization

Flexibility

AMD Turbo CORE

AMD Core Select

Multiple Flex FP

processing modes

Virtual Cache Partitioning

New technical computing

instructions

Highest x86 core

count* for more VMs,

better server utilization

Scalable for rapid

growth/expansion

1P to 4P with same

platform and chipset

Supports 4 to 64 cores

Power Flexibility

TDP Power Cap

C6 state

6 power bands

Low & Ultra Low Voltage

memory support

Multi-generational platform consistency

Up to 46% lower power at idle*; as little as 4.375W per core

Up to 84% greater performance*


AMD TURBO CORE TECHNOLOGY

*Based on AMD Opteron™ 6200 Series processors with up to 300 MHz in P1 boost state and up to 1 GHz+ in P0 boost state over base P2 clock frequency.

Base frequency with

TDP headroom

All core boost activated

(up to 500MHz)

Max turbo activated

(up to 1GHz+, half cores)

All Core Boost

When there is TDP headroom in a

given workload, AMD Turbo CORE

technology is automatically activated

and can increase clock speeds by

300-500 MHz* across all cores.

Max Turbo Boost

When a lightly threaded workload sends half the

“Bulldozer” modules into C6 sleep state but also

requests max performance, AMD Turbo CORE

technology can increase clock speeds by up to

1 GHz+* across half the cores.

+


FLEX FP 256-BIT FPU AND NEW TECHNICAL COMPUTING INSTRUCTIONS

• A flexible floating point unit shared between 2 integer cores

• Simultaneously executes two 128-bit instructions or one 256-bit instruction

• Saves die space and conserves power for majority of non-FP applications

• Dedicated floating point scheduler, which minimizes latency for floating point applications

Instructions Applications/Use Cases

SSSE3, SSE4.1,

SSE4.2

(AMD and Intel)

• Video encoding and transcoding

• Biometrics algorithms

• Text-intensive applications

AESNI

PCLMULQDQ

(AMD and Intel)

• Application using AES encryption

• Secure network transactions

• Disk encryption (MSFT BitLocker)

• Database encryption (Oracle)

• Cloud security

AVX

(AMD and Intel)

Floating point intensive applications:

• Signal processing / Seismic

• Multimedia

• Scientific simulations

• Financial analytics

• 3D modeling

FMA4

(AMD Unique)

• Vector and matrix multiplications

• Polynomial evaluations

• Chemistry, physics, quantum mechanics and

digital signal processing

XOP

(AMD Unique)

• Numeric applications

• Multimedia applications

• Algorithms used for audio/radio


DESIGNED TO DRIVE DOWN POWER REQUIREMENTS

Low and Ultra Low Voltage Memory

1.35v DIMMs reduce voltage by 10%;

1.25v DIMMs reduce voltage by 16%¹

Reduces Idle CPU

Power By Up to 46%²

C6 power state

Shuts down clocks and power to idle cores

Intelligent Circuit

Design

All New Design

Minimizes the number of active transistors for lower power and better

performance

Enables More Power

Control for IT

TDP Power Cap

Flexibility to set power limits without capping

frequency

Up to 56% better power-per-core than Xeon³

¹Regular voltage=1.5v, low voltage=1.35v, ultra-low voltage=1.25v; ² ² Based on internal testing as of 8/2011: AMD Opteron™ processor model 6174 (12-core 2.2GHz) consumes 11.7W in

active idle C1E power state, while AMD Opteron™ processor model 6276 (16-core 2.3GHz) consumes only 6.4W in the active idle C1E power state with new C6 power gating employed.

System configuration: “Drachma” reference design kit, 32GB (8 x 4GB DDR3-1333) memory, 500GB SATA disk drive, Microsoft® Windows Server® 2008 x64 Enterprise Edition R2. SVR-

60; ³based on AMD Opteron 4200 Series processor with 8 cores at 35W TDP versus lowest wattage, highest core Intel Xeon processor with 6 cores at 60W TDP.

More Low Power

Memory Choices


TDP POWER CAP – THE ULTIMATE POWER AND UTILIZATION TOOL

Allows granular control of power

envelopes

Maximizes the density of servers

in the data center

Convert untapped space in the

rack into additional density

Customer-definable, up to 1W

increments

Many workloads will still operate

within the same performance

range despite the lower TDP

17 2U servers @

350W max power

5,950W total power

8U of wasted space

20 2U servers @

300W max power

6,000W total power

17% more servers

6000W Rack Example


0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1Q11 2Q11 3Q11

Xeon X5680

Xeon X5677

Xeon X5670

Xeon X5667

Xeon X5660

Xeon X5650

Xeon E5640

Xeon E5630

Xeon E5620

0

20

40

60

80

100

120

140

160

Price Performance

Xeon E5640 AMD Opteron 6276

BEATING THE MOST POPULAR XEON PROCESSORS BY 89%

AMD Opteron™ processor model 6276 beats Intel‟s most popular Xeon by 89%, at the same price

AMD delivers clearly superior performance where customers are actually spending their money

1 Source: Mercury Research, PC Processors and Chip Sets Updated Edition 3Q2011

²SPECint®_rate2006 Score= 480, 2 x AMD Opteron™ processors Model 6276 in Supermicro A+ Server 1022-URF server, 128GB (16 x 8GB DDR3-1333) memory, Red Hat Enterprise Linux 6.1, x86

Open64 4.2.5.2 Compiler Suite, based on AMD internal testing in Oct‟11. SPECint®_rate2006 Score = 253, 2 x Intel Xeon processors Model X5640 in Cisco UCS C210 M2 , 48GB (12 x 4GB DDR3-

1333) memory, SUSE Linux Enterprise Server 11 SP1, Intel C++ Compiler XE Version 12.0.1.116 Build 20101116, published Mar‟11 at http://www.spec.org/cpu2006/results/res2011q2/cpu2006-

20110329-15460.html

89%

Intel 2011 Xeon 5600 Series Shipments1

http://www.spec.org/cpu2006/results/res2011q2/cpu2006-20110329-15460.html







SAME PRICE, BETTER PERFORMANCE, WITH 2-3X THE BENEFIT¹

Intel AMD Advantage

Processor 2 Xeon E5640 2 AMD Opteron™ Series

Processors Model 6220 25% faster

Memory 16GB 32GB Twice the memory

RAID Controller None RAID 5 Controller Higher availability

Hard Drives 1 x 146GB 3 x 146GB Triple the storage

Price $4,736 $4,576 3% lower price

Helping businesses get the most for their budget

1 Pricing and specifications from www.hp.com as of 10/26/11. As of 10/26/11, pricing for the HP DL385 with AMD Opteron processor Model 6220 (1ku pricing $523) is

not yet available. The price of an HP DL385 with AMD Opteron processors Model 6134 (1ku pricing $523) was used as a proxy because pricing for AMD Opteron 6100

and 6200 Series processor-based servers is expected to be very similar. Server configurations are identical except on the specifications listed above and include

standard warranty, power supply and no OS. Difference in processor speeds is based on the top SPECint®_rate2006 results published by, or submitted by AMD to,

SPEC as of 10/26/2011 for 2P servers with the above identified processors and may differ from the performance of the servers priced and described above. For the

latest results, visit www.spec.org. See backup slide for configuration details.

http://www.spec.org/


2P 4P

Intel AMD

AMD OPTERON ADVANTAGES FOR VIRTUALIZATION

Customer Requirements:

Core density

High memory addressability

Large L3 cache

Cost efficiency

Microsoft Hyper-V

VMware vSphere

Microsoft Remote FX

VMware View

Xen, KVM

Estimated 72 – 77% platform

level savings on like to like

configurations¹

¹, ² See backup data on slide 25

At 1VM per core, AMD can host

672 VMs in a standard 42U rack

The same load would require

2.6 racks of Intel systems²,

3X the floor space

Greater Virtualization Density

Lower Cost per VM

Virtualization


PERFORMANCE THAT DRIVES THE CLOUD


Power efficiency

Core scalability

Node and thread density

High throughput

Windows® Azure

LAMP Stack

Java

Hadoop

OpenStack

Lowest x86 watts/core in the

industry: 5.3W for “Interlagos”

and 4.375W for “Valencia”¹

24-25% better performance in

web-focused workloads²

Core Density for Greater Throughput

Unmatched Power Efficiency

Web / Cloud

0

20

40

60

80

100

120

140

Spec Int SPEC JBB

Xeon 5690 AMD Opteron 6282

0

2

4

6

8

10

12

Xeon5600

AMDOpteron

6200

AMDOpteron

4200

¹, ² See backup data on slide 26


AMD’s Private Cloud

115K cores at work and counting

Supports thousands of engineers worldwide

40M+ tasks executed per month

90% utilization rate for both cores and systems

100% of engineering projects in the company

reside in the cloud

AMD IS POWERING THE CLOUD TODAY

High core count, scalability and

power efficiency are driving

clouds to adopt AMD

There are more than 2 million

AMD Opteron™ processors at

work today in the cloud


THE HPC LEADER

¹-3 See complete benchmark data on slides 27-29.


Scalable performance

Strong floating point performance

High memory throughput

More cores for highly threaded apps

Wide range of technical instructions

Linux OS

Open64

GCC

PGI Compilers

24-84% better performance at

significantly lower price¹

With almost twice the FLOPs

per sq. ft. with “Interlagos”, it

would take Intel almost 2 racks

to match AMD in density and

performance²

Greatest FLOPs per Sq. Foot

Superior Performance¹

HPC

-

50

100

150

200

SP

EC

FP

ST

RE

AM

LIN

PA

CK

LA

MM

PS

NA

MD

WR

F

Xeon 5670 AMD Opteron 6276

73GB/s memory

throughput3

73% more memory

bandwidth than Intel3

Maximum cores

per rack2

More FLOPs per sq. foot2

33% lower cost per core4


HPC: THE EARLY ADOPTER MARKET

HPC acts as leading indicator for new technology

Adopts and proves viability of new performance and

efficiency trends

As of June 2011, 1/3 of the world‟s top 100

Supercomputers are based on the AMD Opteron™

processor

AMD – the power behind the first system to break the

petaflop barrier - currently powers 4 petaflop-capable top

10 systems

Half a million "Bulldozer" cores have already shipped to

some of the most demanding customers in the world

U.S. Department of Energy

– Oak Ridge National Lab

– Sandia National Lab

– Los Alamos National Lab

– Pacific Northwest National Lab

National Oceanic & Atmospheric Administration

Department of Defense

The HECToR initiative in Europe

University of Stuttgart‟s High Performance Computing Center

Buddy Bland

Project Director, ORNL Leadership Computing Facility

“ORNL is deploying more than 25,000 of AMD‟s „Interlagos‟ processors

over the next few months as we upgrade Jaguar to the new Titan system.

Our users are excited about the increase in performance over previous-

generation processors, and our sponsors are delighted with the power

savings that will make Titan one of the world‟s most powerful and efficient

research tools.”


Software and Ecosystem Partners

CUSTOMERS AND PARTNERS

Range of Infrastructure platforms

Systems and Infrastructure

More than 30 expected server platforms


THE NEW AMD OPTERON™ PROCESSOR:

THE CORE OF THE CLOUD

Designed for the inflection point around the

hyper-efficient, virtualized Cloud

Strong OEM and end-customer support out of

the gate

Superior performance, ranging from 24% to 84%

in key trending workloads

Increased virtualization scalability

As much as 56% lower power-per-core

Perfectly matched architecture for today‟s highly

threaded workloads including cloud, web,

virtualization, database and HPC


BACKUP


DATA: AVAILABLE TODAY: SERVERS FEATURING THE ALL NEW AMD OPTERON PROCESSOR

84% Better performance:

LINPACK (2P) AMD Opteron processor Model 6276 generates 84% more FLOPS than Intel Xeon processor Model X5670

– 239.1 FLOPS, AMD Opteron 6276, Trial 127, STEP B2G, Node 1, PROC. 32, PARAM WR01R2R4, N 83200, NB 100, P 4, Q 8, E .81, FC

gfortran 4.6, BLAS ACML 5.0.0.0, FCFLAGS -fomit-frame-pointer -O3 -funroll-loops -W -Wall -mavx -mfma4, NUMA hwloc, MPI OMPI 1.5.3,

DIMM MHz 1600, DIMM# 8, DIMM GB 8, Total Capacity 64, System SM H8DGT, OS SLES11 SP1, Kernel 2.6.32.28-fam15h-default, Notes

HPC mode

– 130.1 FLOPS, Intel Xeon X5670, Trial 5, STEP NA, Node 1, PROC. 1, PARAM WR01R2R4, N 51744, NB 112, P 2, Q 6, E .93, FC ifort 11.1,

BLAS MKL 10.3, FCFLAGS O3 -w -ansi-alias -i-static -openmp -nocompchk, NUMA –localalloc, MPI OMPI 1.5.1, DIMM MHz 1333, DIMM# 4,

DIMM GB 6, Total Capacity 24, System SM X8DTT, OS SLES11, Kernel 2.6.27, Notes Turbo

73% better scalability:

STREAM (2P) AMD Opteron processor Model 6276 has 73% higher memory bandwidth than Intel Xeon processor Model X5670

– 73, 2 x AMD Opteron™ processors Model 6276 in Supermicro H8DGT, 64GB (8 x 8GB DDR3-1600) memory, SuSE Linux® Enterprise Server

11 SP1 64-bit, x86 Open64 4.2.5-1 Compiler Suite

– 42, 2 x Intel Xeon processors Model X5670 in Supermicro X8DTT server, 24GB (6 x 4GB DDR3-1333) memory, SuSE Linux® Enterprise Server

11 SP1 64-bit, Intel Compiler v11.1.064


DATA: AVAILABLE TODAY: SERVERS FEATURING THE ALL NEW AMD OPTERON PROCESSOR

1/2 the power-per-core*

– As of Nov 1, 2011, AMD Opteron™ processor Models 4200 EE have the lowest known power per core of any x86 server processor, at 35W

TDP (35W/8 = 4.375W/core). Intel 's lowest power per core server processor, L5630, is 40W TDP (40W/4 = 10W/core). See

http://www.intel.com/Assets/PDF/prodbrief/323501.pdf. Previous record held by AMD Opteron processor Models 4100 EE at 35W TDP / 6

cores = 5.83 W/core.

Requires 2/3 less floor space*

– VMs/rack (2P and 4P) One rack of AMD Opteron 6200 Series-based servers can support 672 VMs (1 VM per core, 2U servers)

– This would take three racks and 56 2U Intel Xeon 5600 Series-based servers to do the same.

– Assumes 1 VM/core, AMD Opteron 6200 Series-based 2P 2U server has up to 32 cores, supports up to 32 VMs/server x 21 servers per rack,

which equals 672 VMs per server. Intel Xeon 5600 Series-based 2P 2U server has up to 12 cores, supports up to 12 VMs/server x 21 servers

per rack, which equals 252 VMs per server, Intel specs as of 11/4/11 at www.intc.com/pricelist.cfm.

1/3 to 2/3 lower platform price*

– Top bin comparisons. Dell R710 with two top bin Intel Xeon processor Model X5690s is $7,103 . Since pricing for a Dell R715 with two top bin

AMD Opteron processor Model 6282 SE is not yet available, the similarly priced AMD Opteron processor Model 6140 was used and the server

yielded a price of $4564. That is a 36% price savings. Both servers were configured with 32GB RAM, 146GB 10K hdd, and 3yr base warranty

and large enterprise pricing is from www.dell.com as of 10/22/11. HP DL 580 with four top bin Intel Xeon processor Model E7-4870 is $29,336

at www.dell.com. Since pricing for an HP DL585 pricing with four top bin AMD Opteron Model 6828 SE is not yet available, the similarly priced

AMD Opteron processor Model 6140 was used and server yielded a price of $11,094. Both servers were configured with 64GB RAM, 72GB

15K hdd, and 3yr base warranty and large enterprise pricing is from www.hp.com as of 10/22/11. VMware vSphere pricing not included,

assuming both servers configured with versions 5.0 or 4.1u2, which are the same price for AMD- and Intel-based servers.

http://www.intc.com/pricelist.cfm


DATA: WORLD’S MOST CONFIGURABLE CORES FOR AN AGILE CLOUD

1 84% Better performance:

– LINPACK (2P) AMD Opteron processor Model 6276 generates 84% more FLOPS than Intel Xeon processor Model X5670

– 239.1 FLOPS, AMD Opteron 6276, Trial 127, STEP B2G, Node 1, PROC. 32, PARAM WR01R2R4, N 83200, NB 100, P 4, Q 8, E .81, FC gfortran 4.6, BLAS ACML 5.0.0.0,

FCFLAGS -fomit-frame-pointer -O3 -funroll-loops -W -Wall -mavx -mfma4, NUMA hwloc, MPI OMPI 1.5.3, DIMM MHz 1600, DIMM# 8, DIMM GB 8, Total Capacity 64, System SM

H8DGT, OS SLES11 SP1, Kernel 2.6.32.28-fam15h-default, Notes HPC mode

– 130.1 FLOPS, Intel Xeon X5670, Trial 5, STEP NA, Node 1, PROC. 1, PARAM WR01R2R4, N 51744, NB 112, P 2, Q 6, E .93, FC ifort 11.1, BLAS MKL 10.3, FCFLAGS O3 -w -

ansi-alias -i-static -openmp -nocompchk, NUMA –localalloc, MPI OMPI 1.5.1, DIMM MHz 1333, DIMM# 4, DIMM GB 6, Total Capacity 24, System SM X8DTT, OS SLES11, Kernel

2.6.27, Notes Turbo

² Based on internal testing as of 8/2011: AMD Opteron™ processor model 6174 (12-core 2.2GHz) consumes 11.7W in active idle C1E power state, while AMD Opteron™ processor

model 6276 (16-core 2.3GHz) consumes only 6.4W in the active idle C1E power state with new C6 power gating employed. System configuration: “Drachma” reference design kit, 32GB

(8 x 4GB DDR3-1333) memory, 500GB SATA disk drive, Microsoft® Windows Server® 2008 x64 Enterprise Edition R2. SVR-60

3 Based on 16-core AMD Opteron 6200 Series processor compared to 6-core Intel Xeon 5600 Series and 10-core Intel Xeon E7 processors.


DATA: “AMD OPTERON ADVANTAGES FOR VIRTUALIZATION”

1Server cost per VM (2P) AMD Opteron 6200-based server cost per VM is up to 72% less expensive

Assumes 1 VM/core, Dell R710 with two top bin Intel Xeon processor Model X5690 is $7,361, has 12 cores and 12VMs, $613/VM.

Since pricing for a Dell R715 with two top bin AMD Opteron processor Model 6282 SE is not yet available, the similarly priced AMD

Opteron processor Model 6140 was used and the server yielded a price of $5406, has 32 cores and 32VMs, $169/VM . Servers

were configured with 32GB RAM, 146GB 10K hdd, and 3yr base warranty and large enterprise pricing is from www.dell.com as of

10/27/11 and 10/28/11.

Server cost per VM (4P) AMD Opteron 6200-based server cost per VM is up to 77% less expensive

Assumes 1 VM/core, HP DL 580 G7 with four top bin Intel Xeon processor Model E7-4870 is $28,936, has 40 cores and 32VMs,

$723/VM. Since pricing for an HP DL585 with four top bin AMD Opteron processor Model 6282 SE is not yet available, the

similarly priced AMD Opteron processor Model 6140 was used and the server yielded a price of $10,698, has 64 cores and 64VMs,

$167/VM. Servers were configured with 64GB RAM, 72GB 15K hdd, and 3yr base warranty and large enterprise pricing is from

www.hp.com as of 11/4/11. VMware vSphere pricing not included, assuming both servers configured with versions 5.0 or 4.1u2,

which are the same price for AMD- and Intel-based servers.

²AMD Opteron 6200 Series-based 2P servers can support up to 32VMs each, assuming 1 VM per core. 21 servers fit in a rack x

32 VMs = 672 VMs per rack. Intel Xeon 5600 Series-based 2P servers can support up to 12VMs each, based on core counts

listed at www.intc.com/pricelist.cfm as of 10/24/11, which equates to 21 servers x 12 VMs = 252 VMs. # of vCPUs for 16 VMs =

max cores/16.


DATA: “PERFORMANCE THAT DRIVES THE CLOUD”

¹Intel Xeon processor Model L5609 is 10W/core (40W TDP/4 cores) as of 10/24/11, see http://www.intc.com/pricelist.cfm. AMD Opteron

6200 Series “HE” processors will be as low as 85W with 16 cores for 5.3 W/core. AMD Opteron 4200 Series “EE” processors will be as

low as 35W TDP with 8 cores for 4.375W/core.

²SPECint®_rate2006 Score= 526, 2 x AMD Opteron™ processors Model 6282 SE in Supermicro A+ Server 1022-URF server, 128GB (16

x 8GB DDR3-1333) memory, Red Hat Enterprise Linux 6.1, x86 Open64 4.2.5.2 Compiler Suite, based on AMD internal testing in Oct‟11.

SPECint®_rate2006 Score = 421, 2 x Intel Xeon processors Model X5690 in NEC Express5800/R120b-1 server, 96GB (12 x 8GB DDR3-

1333) memory, SUSE Linux Enterprise Server 11 SP1, Intel C++ Compiler XE v12.0.3.174, published Aug‟11 at

http://www.spec.org/cpu2006/results/res2011q3/cpu2006-20110801-17813.html.

SPECjbb2005 Score = 1211915 for AMD Opteron processor Model 6282 SE based on AMD internal testing in Oct‟11. SPECjbb2005

Score = 975257 for Intel Xeon processor Model X5690. Cisco UCS B200 M2 (Intel Xeon X5690, 3.46GHz) 48GB(12 x 4GB 2Rx4 PC3-

10600R, ECC, running 1333MHz) IBM Corporation IBM J9 JVM (build 2.4, JRE 1.6.0 IBM J9 2.4 Windows Server 2008 amd64-64

jvmwa6460sr6-20090923_42924 (JIT enabled, AOT enabled), published in Mar‟11 at

http://www.spec.org/osg/jbb2005/results/res2011q1/jbb2005-20110223-00958.html.

SPEC, SPECint, and SPECJBB are registered trademarks of the Standard Performance Evaluation Corporation. The results stated above

top reflect results published on http://www.spec.org/cpu2006/results/ as of 10/26/11


DATA: “THE HPC LEADER”

¹SPECfp

360, 2 x AMD Opteron™ processors Model 6276 in Supermicro A+ 1022-URFserver, 64GB (8 x 8GB DDR3-1600) memory, Red Hat Enterprise Linux 6.1 64-bit, x86 Open64 4.2.5.2

Compiler Suite

263, 2 x Intel Xeon processors Model X5670 in Cisco UCS B200 M2 server, 48GB (12 x 4GB DDR3-1333) memory, SUSE Linux® Enterprise Server 11 SP1 64-bit, Intel C++ Compiler

XE v12.0.1.116


STREAM

73, 2 x AMD Opteron™ processors Model 6276 in Supermicro H8DGT, 64GB (8 x 8GB DDR3-1600) memory, SuSE Linux® Enterprise Server 11 SP1 64-bit, x86 Open64 4.2.5-1

Compiler Suite

42, 2 x Intel Xeon processors Model X5670 in Supermicro X8DTT server, 24GB (6 x 4GB DDR3-1333) memory, SuSE Linux® Enterprise Server 11 SP1 64-bit, Intel Compiler v11.1.064

LINPACK

GFLOPS = # processors x processor clock speed x # cores per processor x # 32-bit operations per second

147.2, 1 x AMD Opteron™ processors Model 6276 x 16 cores x 2.3GHz x 4

70.32, 1 x Intel Xeon processors Model X5670 x 6 cores x 2.93GHz x 4

GROMACS

100s, Intel Xeon X5670, 1 node, 12 processes, 1 OMP, 1S, 1E, 40.67 GFLOPS, CC ICC 11.1, CCFLAGS-O3 -axS -no-prec-div, NUMA numactl, FFT fftw 3 MKL, PREC Single, MPI

OMPI 1.5.1, DIMM Spec 1333, DIMM Capacity(GB) 4, DIMM Count 6, Total Capacity (GB) 24, System SM X8DTT, OS SLES11 SP1, Kernel 2.6.32.12-0.7-default

84s, AMD Opteron 6276, 1 node, 32 processes, 1 OMP, 1S, 1E, 48.737 GFLOPS, GCC 4.60, CCFLAGS-O3 -O3 -funroll-loops, NUMA APPFILE, FFT 3.2.2, PREC Single, MPI OpenMPI

1.5.3, DIMM Spec 1600, DIMM Capacity(GB) 8, DIMM Count 4, Total Capacity (GB) 32, System SM H8DGT, OS SLES11 SP1, Kernel 2.6.32.12-0.7-default







DATA: “THE HPC LEADER” (CONT.)

LAMMPS

471s, Intel Xeon X5670, Trial 24, Nodes 1, CXX Intel 11.1.064, CXXFLAGS -O2 -funroll-loops -fstrict-aliasin, MPI OMPI 1.5.1 + knem 0.9.5, DIMM Speed (MHz) 1333, DIMM Capacity

(GB) 4, DIMM Count 6, Total Capacity (GB) 24, System SM X8DTT, OS SLES11 SP1, Kernel 2.6.32.12-0.7-default, Notes Turbo ON, HT OFF, -mca btl_sm_use_knem 1

333s, AMD Opteron 6276, Trial 9, Nodes 1, CXX openCC 4.2.5.2-1, CXXFLAGS -O3 -OPT:Ofast -OPT:rsqrt=2 -march=bdver1 -mavx -mfma4, MPI OpenMPI 1.5.3, DIMM Speed (MHz)

1600, DIMM Capacity (GB) 8, DIMM Count 8, Total Capacity (GB) 64, System SM H8DGT, OS SLES11 SP1, Kernel 2.6.32.28-fam15h-default, Notes hpc

NAMD

.636 day/ns, Intel Xeon X5670, Trial 40, NAMD Version 2.7, Nodes 1, MB memory 274.328, CC Intel 11.1, CCFLAGS -ip -fno-rtti -O3 -xSSE4.2 -no-prec-div, MPI OMPI 1.5.1, DIMM

Speed (MHz) 1333, DIMM Capacity (GB) 4, DIMM Count 6, Total Capacity (GB) 24, System SM X8DTT, OS SLES11 SP1, Kernel 2.6.32.12-0.7-default, Notes default

.375 day/ns, AMD Opteron 6276, Trial 77, NAMD Version 2.8, Nodes 1, MB memory 260.277, CC opencc 4.2.5-2.1, CCFLAGS -O3 -m64 -march=bdver1 -mfma4 -mavx -

CG:compute_to=ON -OPT:Olimit=40000, MPI OpenMPI 1.5.3+knem 0.9.6, DIMM Speed (MHz) 1600, DIMM Capacity (GB) 8, DIMM Count 8, Total Capacity (GB) 64, System SM

H8DGT, OS SLES11 SP1, Kernel 2.6.32.28-fam15h-default, Notes default

WRF

224s, Intel Xeon X5670, Trial 124, STEP NA, NODE 1, FC Intel 11.1.064, NETCDF 4.1.1, FCFLAGS -w -O3 -ip -xSSE4.2 -fp-model fast=2 -no-prec-div -no-prec-sqr, NUMA NA, MPI

OMPI 1.5.1 + knem 0.9.5, DIMM # 6, DIMM GB 4, DIMM MHz 2 1333, Total Capacity 24, System SM X8DTT, OS SLES11 SP1, Kernel 2.6.32.12-0.7-default, Notes -mca

btl_sm_use_knem 1

180s, AMD Opteron 6276, Trial 436, STEP B2g, NODE 1, FC open64 4.2.5-1, NETCDF 4.1.2, FCFLAGS -O3 -HP -march=bdver1 -mavx -mfma4 -DpgiFortran -OPT:unroll_size=256 -

LNO:blocking=off -LANG:copyinout=o, NUMA APP FILE, MPI OMPI 1.5.3, DIMM # 8, DIMM GB 8, DIMM MHz 2 1600, Total Capacity 64, System SM H8DGT, OS SLES11 SP1, Kernel

2.6.32.28-fam15h-default, Notes NA

2 Intel can only process one 128-bit instruction at a time. See page 13 of the Intel Hot Chips presentation (http://www.hotchips.org/archives/hc18/3_Tues/HC18.S9/HC18.S9T4.pdf).

AMD Opteron™ 6200 Series has FMAC that can execute an FMA4 execution (a=b+c*d) in one cycle vs. 2 cycles that would be required for FMA3 or standard SSE floating point

calculation. Intel cannot multiply and add in one clock cycle (FMA4 and XOP instructions). See page 9 of the Intel Hot Chips presentation

(http://www.hotchips.org/archives/hc18/3_Tues/HC18.S9/HC18.S9T4.pdf).

²GFLOPS = # processors x processor clock speed x # cores per processor x # 32-bit operations per second. 2P server using AMD Opteron 6282 SE can generate up to 358 GFLOPs. 21

(2P) servers per rack x 358 = 7518 GFLOPs per rack. 2P server using Intel Xeon 5690 can generate up to 166 GLOPS per server. Therefore, it take over 2 racks (over 45 Intel-based

servers) to generate 7518 GFLOPs.

Information on Intel‟s specs as of 9/26/11 at http://www.intc.com/priceList.cfm. AMD specs based on pre-production values.


DATA: THE HPC LEADER

3 STREAM (2P) AMD Opteron processor Model 6276 has 73% higher memory bandwidth than Intel Xeon processor Model X5670

– 73, 2 x AMD Opteron™ processors Model 6276 in Supermicro H8DGT, 64GB (8 x 8GB DDR3-1600) memory, SuSE Linux® Enterprise Server

11 SP1 64-bit, x86 Open64 4.2.5-1 Compiler Suite

– 42, 2 x Intel Xeon processors Model X5670 in Supermicro X8DTT server, 24GB (6 x 4GB DDR3-1333) memory, SuSE Linux® Enterprise Server

11 SP1 64-bit, Intel Compiler v11.1.064

4 Comparison of 12-core AMD Opteron™ processor Moel 6234 expected price of $377 at launch with 4-

core Intel Xeon E5603 price of $188 according to www.intel.com as of 11/4/11.

http://www.intel.com/


PRICING AND MODEL NUMBERS

Model Number Core Count Core Speed All-Core Turbo

Frequency

Max Turbo Core

Frequency TDP 1KU Pricing

6282 SE 16 2.6GHz 3.0GHz 3.3GHz 140W $1019

6276 16 2.3GHz 2.6GHz 3.2GHz 115W $788

6274 16 2.2GHz 2.5GHz 3.1GHz 115W $639

6272 16 2.1GHz 2.4GHz 3.0GHz 115W $523

6238 12 2.6GHz 2.9GHz 3.2GHz 115W $455

6234 12 2.4GHz 2.7GHz 3.0GHz 115W $377

6220 8 3.0GHz 3.3GHz 3.6GHz 115W $523

6212 8 2.6GHz 2.9GHz 3.2GHz 115W $266

6204 4 3.3GHz N/A N/A 115W $455

6262 HE 16 1.6GHz 2.1GHz 2.9GHz 85W $523

4284 8 3.0GHz 3.3GHz 3.7GHz 95W $316

4280 8 2.8GHz 3.1GHz 3.5GHz 95W $255

4238 6 3.3GHz 3.5GHz 3.7GHz 95W $255

4234 6 3.1GHz 3.3GHz 3.5GHz 95W $174

4226 6 2.7GHz 2.9GHz 3.1GHz 95W $125

4274 HE 8 2.5GHz 2.8GHz 3.5GHz 65W $377

4228 HE 6 2.8GHz 3.1GHz 3.6GHz 65W $255

4256 EE 8 1.6GHz 1.9GHz 2.8GHz 35W $377


DISCLAIMER AND ATTRIBUTION

DISCLAIMER

The information presented in this document is for informational purposes only and may contain technical inaccuracies, omissions and

typographical errors. The information contained herein is subject to change and may be rendered inaccurate for many reasons,

including but not limited to product and roadmap changes, component and motherboard version changes, new model and/or product

releases, product differences between differing manufacturers, software changes, BIOS flashes, firmware upgrades, or the like. AMD

assumes no obligation to update or otherwise correct or revise this information. However, AMD reserves the right to revise this

information and to make changes from time to time to the content hereof without obligation of AMD to notify any person of such

revisions or changes.

AMD MAKES NO REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE CONTENTS HEREOF AND ASSUMES NO

RESPONSIBILITY FOR ANY INACCURACIES, ERRORS OR OMISSIONS THAT MAY APPEAR IN THIS INFORMATION.

ATTRIBUTION

©2011 Advanced Micro Devices, Inc. All rights reserved. AMD, the AMD Arrow logo, AMD Opteron and combinations thereof are

trademarks of Advanced Micro Devices, Inc. Microsoft and Windows are registered trademarks of Microsoft Corporation in the U.S.

and/or other jurisdictions. Other names are for informational purposes only and may be trademarks of their respective owners.


GET MORE WITH AMD FOR EVERY IT DOLLAR SPENT

Intel-based server

$8,246

AMD Opteron™-based server

$7,915

Advantage

Processor 2 x Xeon X5690 2 x AMD Opteron 6282SE 26% faster processor

Memory 48GB 64GB 33% more memory

RAS upgrades None RAID 5 Controller

Redundant power supply

High availability

Hard Drives 1 x 300GB 3 x 300GB 3X the storage

Support 1yr 24x7 support

upgrade

Installation + 3yr 24x7 support

upgrade

3X the support

More hardware for the same IT budget Pricing and specifications from www.hp.com as of 10/26/11. As of 10/26/11, pricing for the HP DL385 with AMD Opteron processor Model 6282 SE (1ku pricing $1019) is not yet

available. The price of an HP DL385 with AMD Opteron processors Model 6172 (1ku pricing $989) was used as a proxy because pricing for AMD Opteron 6100 and 6200 Series

processor-based servers is expected to be very similar. . Server configurations are identical except on the specifications listed above and include no OS.. Difference in processor

speeds is based on the top SPECint®_rate2006 results published by, or submitted by AMD to, SPEC as of 10/26/2011 for 2P servers with the above identified processors and

may differ from the performance of the servers priced and described above. For the latest results, visit www.spec.org. See backup slide for configuration details.

http://www.spec.org/


BACKUP – SAME PRICE, 2-3X THE BENEFIT

526 using 2 x AMD Opteron™ processor Model 6282SE in Supermicro A+ 1022-URF server, 128GB (16 x

8GB DDR3-1333) memory, Red Hat Enterprise Linux Server release 6.1

419 using 2 x Intel Xeon X5690: http://www.spec.org/cpu2006/results/res2011q1/cpu2006-20110211-

14447.html

317 using 2 x AMD Opteron™ processor Model 6220 in Supermicro A+ 1022G-URF server, 64GB (8 x

8GB DDR3-1) memory, Red Hat Enterprise Linux Server release 6.1

253 using 2 x Intel Xeon E5640: http://www.spec.org/cpu2006/results/res2011q2/cpu2006-20110329-

15452.html











AMD Opteron 6200 and 4200 Series Presentation

Business

Transcript of AMD Opteron 6200 and 4200 Series Presentation