3. Legal Disclaimers: Performance
Performance tests and ratings are measured using specific computer systems and/or components and reflect the approximate
performance of Intel products as measured by those tests. Any difference in system hardware or software design or
configuration may affect actual performance. Buyers should consult other sources of information to evaluate the performance
of systems or components they are considering purchasing. For more information on performance tests and on the
performance of Intel products, Go to: http://www.intel.com/performance/resources/benchmark_limitations.htm.
Intel does not control or audit the design or implementation of third party benchmarks or Web sites referenced in this
document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance
benchmarks are reported and confirm whether the referenced benchmarks are accurate and reflect performance of systems
available for purchase.
Relative performance is calculated by assigning a baseline value of 1.0 to one benchmark result, and then dividing the actual
benchmark result for the baseline platform into each of the specific benchmark results of each of the other platforms, and
assigning them a relative performance number that correlates with the performance improvements reported.
SPEC, SPECint, SPECfp, SPECrate. SPECpower, SPECjAppServer, SPECjEnterprise, SPECjbb, SPECompM, SPECompL, and SPEC
MPI are trademarks of the Standard Performance Evaluation Corporation. See http://www.spec.org for more information.
TPC Benchmark is a trademark of the Transaction Processing Council. See http://www.tpc.org for more information.
SAP and SAP NetWeaver are the registered trademarks of SAP AG in Germany and in several other countries. See
http://www.sap.com/benchmark for more information.
INFORMATION IN THIS DOCUMENT IS PROVIDED “AS IS”. NO LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL OR OTHERWISE,
TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. INTEL ASSUMES NO LIABILITY WHATSOEVER
AND INTEL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO THIS INFORMATION INCLUDING LIABILITY OR
WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT,
COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT.
Performance tests and ratings are measured using specific computer systems and/or components and reflect the approximate
performance of Intel products as measured by those tests. Any difference in system hardware or software design or
configuration may affect actual performance. Buyers should consult other sources of information to evaluate the performance of
systems or components they are considering purchasing. For more information on performance tests and on the performance of
Intel products, reference www.intel.com/software/products.
Software and workloads used in performance tests may have been optimized for performance only on Intel
microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems,
components, software, operations and functions. Any change to any of those factors may cause the results to vary. You
should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including
the performance of that product when combined with other products.
3
4. 4
1. IDC “Server Workloads Forecast” 2009. 2.IDC “The Internet Reaches Late Adolescence” Dec 2009, extrapolation by Intel for 2015
2.Intel “Worldwide Device Estimates Year 2020 - Intel One Smart Network Work” forecast
3. Source: http://www.cisco.com/assets/cdc_content_elements/networking_solutions/service_provider/visual_networking_ip_traffic_chart.html
extrapolated to 2015
ネットユーザが
10億人以上に
ユーザーの増加1
150億台にのぼる
ネットワーク接続機器
デバイスの増加2
1,000 エクサバイト
を超えるインターネ
ット・トラフィック
データの増加3
ITに対する要求の急増は、性能、効率、コストの
最適な組み合わせを必要としています
より多くの処理をより効率的に
5. 5
同じ電力水準の前世代プロセッサーと比べて
最大 80%の性能向上1
インテル® アドバンスト・ベクトル・エクス
テンションにより演算処理時間を大幅に削減
インテル® ターボ・ブースト・テクノロジー
2.0 により、必要に応じて性能を向上
インテル® データダイレクト I/O に対応した
インテル® インテグレーテッド I/O により、
処理能力と帯域幅を拡大しながら
レイテンシーを削減2
最大4本のメモリーチャネル
DDR3-1600メモリーを
サポート
最大8 コア
最大20 MB の
キャッシュIntegrated
PCI Express* 3.0
1ソケットあたり最大
40レーンまでサポート
次世代データセンターの中核
インテル® Xeon® プロセッサー
E5-2600 製品ファミリー
1 Performance comparison using best submitted/published 2-socket server results on the SPECfp*_rate_base2006 benchmark as of 6 March 2012.
2 Source: Intel internal measurements of average time for an I/O device read to local system memory under idle conditions comparing Intel® Xeon®
processor E5-2600 product family (230 ns) vs. Intel® Xeon® processor 5500 series (340 ns). See notes in backup for configuration details
Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark
and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the
results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the
performance of that product when combined with other products.
For more information go to intel.com/performance
8. Higher
is better
8
前世代に比べて最大80%の性能向上
1
1.51
1.59 1.62
1.83 1.88
2.18
Intel® Xeon® Processor E5-2690 (8C, 2.9 GHz)
Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as
SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors
may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases,
including the performance of that product when combined with other products.
Linpack performance may vary based on thermal solution.
Source: Intel internal measurements and best published results as of 6 March 2012
For more information go to http://www.intel.com/performance
X5690 Baseline
(3.46GHz, 6C)
SPECvirt_sc*2011
SPECint*_rate
_base2006
SPECjbb*2005
SPECfp*_rate
_base2006
STREAM_MP Triad
Linpack
Intel® Xeon® Processor E5-2690 (8C, 2.9GHz, 135W) vs.
Intel® Xeon® Processor X5690 (6C, 3.46GHz, 130W)
9. チック・タック開発モデル:
Sustained Microprocessor Leadership
Intel® Core™
Microarchitecture
TOCK
New
Micro-
architecture
Merom
65nm
TICK
Penryn
New
Process
Technology
45nm
Intel® Microarchitecture
Codename Nehalem
TOCK
New
Micro-
architecture
Nehalem
45nm
TICK
Westmere
32nm
New
Process
Technology
Intel® Microarchitecture
Codename Sandy
Bridge
TOCK
Sandy
Bridge
32nm
New
Micro-
architecture
TICK
Ivy
Bridge
22nm
New
Process
Technology
Intel® Microarchitecture
Codename Haswell
TOCK
Haswell
22nm
New
Micro-
architecture
TICK
Future
14nm
New
Process
Technology
9
10. Sandy Bridgeコアのフロントエンドの改良
Totally new branch predictor
• Excellent Performance and
Power/Performance
Decoded Uop Cache
• Power down front end and use cached
results of decoding
• Improves performance by decreasing
front end latency
• Improve performance by ‘stitching’
together code across branches
• Sustain higher uop Bandwidth by “do
once” handling of X86 complexities
Lower Power
AND Higher Performance
32K
ICache
PreDecode
Instruction Queue
Decode
PreDecode
Instruction
Queue
Decode
uop Queue
Out of Order
Execution
uop Queue
Out of Order
Execution
Uop Cache
Hit Logic
Uop Cache
Hit Logic
Uop Cache
Data
Uop Cache
Data
Branch
Prediction
Branch
Prediction
Branch
Prediction
Branch
Prediction
32K Instruction
Cache
Sandy Bridge - Intel® Next Generation Microarchitecture10
11. Make the memory ports
symmetric
• Two 128-bit loads per cycle
• Second load port is fully
featured for misalignment
and store forwarding
• Increased store address BW
48 bytes per cycle internal
memory bandwidth
Load
Store
(address)
Store
(data)
32K Data Cache
256K
Mid-Level Cache
Store
Buffer
and
Forwarding
Scheduler 2 loads per cycle
increases performance for
many applications
Load
Store
(address)
Load
Store
(address)
32K Data Cache
8-byte banks
Store
Buffer
and
Forwarding
Sandy Bridge - Intel® Next Generation Microarchitecture
Sandy Bridgeコアのメモリクラスタの改良
11
12. 12
インテル® アドバンスト・ベクトル・エクステンション
• Intel® AVX extends all 16 XMM registers to 256bits
• Intel AVX works on either
• The whole 256-bits – for FP instructions
• The lower 128-bits (like existing SSE instructions)
• A drop-in replacement for all existing scalar/128-bit SSE instructions
• The upper part of the register is zeroed out
• Intel AVX targets high-performance
• 256-bit Multiply, Add and Shuffle engines
• 2nd load port
256 bits (upcoming)
YMM0
XMM0
128 bits (1999)
13. 「バスタブ」モデル:
Flow from the faucet…
(CPU power)
can exceed drain flow…
(thermal solution capability)
for a short time…
(depending on capacity, level and flow)
if the bathtub…
(energy or temperature)
is not full.
(at limits)
Turbo 2.0 Implications:
CPU can safely operate >TDP for short periods
Potential for more time in Turbo mode, especially when system has been operating at low power
インテル® ターボ・ブースト・テクノロジー 2.0
パワーバジェットの概念
Thermal solution “drains”
energy from the system
CPU adds energy
to the system
Temperature
orEnergy
Thermal Budget
Thermal
Capacitance
13
パワーバジェット管理によって、ヘッドルームがある場合は
バースト的な電力消費が可能になり、最大限のパフォーマンスを発揮
14. インテル® ターボ・ブースト・テクノロジー 2.0
の動作例
Time
Power
Sleep or
Low power
“Next Gen
Turbo Boost”
“TDP”
C0
(Turbo)
After idle periods, the
system accumulates
“energy budget” and can
accommodate high
power/performance for a
few seconds
In Steady State conditions
the power stabilizes on TDP
Buildup thermal budget
during idle periods
Use accumulated
energy budget to
enhance user
experience
14
16. インテル® Xeon® プロセッサー E5 ファミリー
の電力効率
データセンターにおける最高の消費電力当たり性能を実現1
約 70%1
同じ消費電力でパフォーマンスが向上
X5690
(3.46GHz,
6C, 130W)
Scale memory, cache, &
I/O to match core needs
Fine grained control across
23 power domains
Improved efficiency
reduces operating
expenses
16
E5-2680
(2.7GHz,
8C, 130W)
性能に関するテストに使用されるソフトウェアとワークロードは、性能がインテル® マイクロプロセッサー用に最適化されていることがあります。SYSmark* や MobileMark* などの性能テストは、特定のコ
ンピューター・システム、コンポーネント、ソフトウェア、操作、機能に基づいて行ったものです。結果はこれらの要因によって異なります。製品の購入を検討される場合は、他の製品と組み合わせた場合
の本製品の性能など、ほかの情報や性能テストも参考にして、パフォーマンスを総合的に評価することをお勧めします。
1 出典:同じTDPでSPECfp*_rate_base2006ベンチマークを使用して行ったパフォーマンスの比較。ベースラインとなるスコア271は、http://www.spec.org/( 英語)に掲載された、インテル® コンパイ
ラー 12.1 を使用して行った2012 年1 月17 日現在の最良の公表値に基づく、前世代の2-way インテル® Xeon® プロセッサー X5690 の数値です。詳細については、http://www.spec.org/cpu2006/results/
res2012q1/ cpu2006-20111219-19195.html( 英語) を参照してください。新しいスコア466は、2 基のインテル® Xeon® プロセッサー E5-2680、ターボ有効、EIST 有効、ハイパースレッディング使用、
64 GB RAM、インテル® コンパイラー 12.1、THP 無効、Red Hat* Enterprise Linux* Server 6.1 を搭載したインテル® Rose City プラットフォームを使用してインテル社内で測定した推定値に基づきます。
SPECfp*_rate_base2006
271
466
17. Virtualization, in-memory databases and high performance
computing are driving need for more memory
LRDIMM enables up to 3x more capacity compared to RDIMM
Intel Xeon processor E5-2600 family based platforms w/
LRDIMMs have max capacity of 768 GB
メモリサポートの拡張 - LRDIMMの特徴
0
128
256
384
512
640
768
1066 1333
32 GB LRDIMM 32 GB RDIMM 16GB RDIMM
Memory
Controller
Data
Data
Memory
Controller
Data
Data
........
..........
Register
..........
........
..........
Buffer
..........
CMD/ADDR/
CLK
CMD/
ADDR/
CLK
Higher loading on data drivers Reduced loading on data drivers
RDIMM LRDIMM
17
19. 電力効率化のための4つの柱:
Knowledge and Control is Power!
Energy Efficient………
Performance:
• Intel® Turbo
Boost Technology
2.0
Data Center:
• Running Average
Power Limit
Software:
• Intel® Node
Manager 2.0
Platform:
• Improved Load
line
Monitor
Server Power
Consumption
Provision
Power Budget
to avoid
Performance
Loss
Improve
Density by
setting Power
Limits
Monitor for
Performance
Effects
19