Решения и технологии Intel для высоко

Стратегия развития
архитектур и технологий
Intel для высокопроизводительных
вычислений
Николай Местер
Директор по развитию корпоративных
проектов
Ноябрь 2011
INTEL HPC - 2011
Legal Disclaimer
INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL® PRODUCTS. NO LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL OR
OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. EXCEPT AS PROVIDED IN INTEL’S TERMS AND CONDITIONS OF
SALE FOR SUCH PRODUCTS, INTEL ASSUMES NO LIABILITY WHATSOEVER, AND INTEL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO SALE
AND/OR USE OF INTEL® PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR
INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT. INTEL PRODUCTS ARE NOT INTENDED FOR USE IN MEDICAL, LIFE
SAVING, OR LIFE SUSTAINING APPLICATIONS.
• Intel may make changes to specifications and product descriptions at any time, without notice.
• All products, dates, and figures specified are preliminary based on current expectations, and are subject to change without notice.
• Intel, processors, chipsets, and desktop boards may contain design defects or errors known as errata, which may cause the product to deviate from published
specifications. Current characterized errata are available on request.
• Penryn, Nehalem, Westmere, Sandy Bridge, and other code names featured are used internally within Intel to identify products that are in development and not
yet publicly announced for release. Customers, licensees and other third parties are not authorized by Intel to use code names in advertising, promotion or
marketing of any product or services and any such use of Intel's internal code names is at the sole risk of the user
• 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.
• Intel, Xeon, Netburst, Core, VTune, and the Intel logo are trademarks of Intel Corporation in the United States and other countries.
*Other names and brands may be claimed as the property of others.
Copyright © 2010 Intel Corporation.
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
2
INTEL HPC - 2011
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.
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
3
Optimization Notice
INTEL HPC - 2011
Intel® compilers, associated libraries and associated development tools may include or utilize options that optimize for instruction sets that are
available in both Intel® and non-Intel microprocessors (for example SIMD instruction sets), but do not optimize equally for non-Intel
microprocessors. In addition, certain compiler options for Intel compilers, including some that are not specific to Intel micro-architecture, are
reserved for Intel microprocessors. For a detailed description of Intel compiler options, including the instruction sets and specific microprocessors
they implicate, please refer to the “Intel® Compiler User and Reference Guides” under “Compiler Options." Many library routines that are part of
Intel® compiler products are more highly optimized for Intel microprocessors than for other microprocessors. While the compilers and libraries in
Intel® compiler products offer optimizations for both Intel and Intel-compatible microprocessors, depending on the options you select, your code
and other factors, you likely will get extra performance on Intel microprocessors.
Intel® compilers, associated libraries and associated development tools may or may not optimize to the same degree for non-Intel microprocessors
for optimizations that are not unique to Intel microprocessors. These optimizations include Intel® Streaming SIMD Extensions 2 (Intel® SSE2),
Intel® Streaming SIMD Extensions 3 (Intel® SSE3), and Supplemental Streaming SIMD Extensions 3 (Intel® SSSE3) instruction sets and other
optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by
Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors.
While Intel believes our compilers and libraries are excellent choices to assist in obtaining the best performance on Intel® and non-Intel
microprocessors, Intel recommends that you evaluate other compilers and libraries to determine which best meet your requirements. We hope to
win your business by striving to offer the best performance of any compiler or library; please let us know if you find we do not.
Notice revision #20101101
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
4
INTEL HPC - 2011
Intel для высокопроизводительных вычислений
Большие
кластеры
Тестирование,
оптимизация
Экспертиза
в области
приложений
Широкий
спектр
продуктов
Технологический
процесс
TeraScale
R&D
Платформы
Для
приложений
HPC
Лидирующие
Производительность,
энергоэффективность
Exa-Scale Labs
Блоки
Для
платформ
Many
Integrated
Core
архитектура
Долгосрочные инвестиции в HPC
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
5
Технологии Intel меняют сегмент HPC
INTEL HPC - 2011
Производительность, Энергоэффективность, Надежность, TCO
Процессоры
Xeon
®
Твердотельные диски
10GbE
Оптимизация
Уменьшая разрыв
MIC
Масштабируемая
производительность,
Энергоэффективность
Multi- и Many-Core
Между
производительности
Для интенсивного
1GbE and InfiniBand*,
ввода-вывода
с RDMA, Unified Networking
Платформеный подход к НРС
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
6
INTEL HPC - 2011
Intel в списке Top 500 в ноябре 2011
 Процессор Intel® Xeon® 5600 #1 в
списке процессоров (223 системы)
 Процессор Intel® Xeon® E5 - 10
систем включая #15 (LLNL)
 Анонсированы еще top10 систем :
Genci Curie, LRZ, IFERC
Архитектура Intel в списке Top 500
 85% новых систем в списке Top500
list
 93% of новых систем в Китае
 5 систем в top 10 (2,4,5,7,9)
(неизменно с июня 2011)
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
Источник: Top500.org, ноябрь 2011
7
INTEL HPC - 2011
Широкий набор процессорных решений для
Вычислительного континуума
Встроеные
Потребительская
электроника
Встроеные
Мобильные
Internet-системы
10X
Меньше
энергии
Лаптопы,
десктопы
Серверы,
Рабочие станции
Требования
Больше
Производ-ти
Критичные
системы
10X
Совместимо с крупнейшими программными разработками и экосистемой
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
8
Широкий набор процессорных решений
Встроеные
Потребительская
электроника
Встроеные
Мобильные
Internet-системы
10X
Меньше
энергии
Лаптопы,
десктопы
Серверы,
Рабочие станции
Требования
Больше
Производ-ти
INTEL HPC - 2011
Критичные
системы
10X
Совместимо с крупнейшими программными разработками и экосистемой
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
9
Повышая производительность и
энергоэффективность
Технология
22NM
Совместимость
TURBO-BOOST
Архитектура Core
AVX
Процессоры
MULTI/MANY-CORE
Замечание:
Не все ядра одинаковы!
Potential future options, subject to change without notice.
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
INTEL HPC - 2011
10
Закон Мура: Успешно претворяется Intel в жизнь
INTEL HPC - 2011
Производительность
транзистора
+20%
Каждая смена
технологического процесса
позволяет:
Энергия на
переключение
-30%
180 nm
1999
130 nm
2001
90 nm
2003
Производство
65 nm
2005
Разработка
45 nm
2007
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
Potential future options, no indication of
actual product or development, subject to change without notice.
32 nm
On
Track
Исследования
15nm
2009 22 nm
2011
11
INTEL HPC - 2011
Инновации тех.процесса определяют темп
развития индустрии
22nm
37%
Прирост
производительности
Low Voltage[+]
>50%
сокращение
энергопотребления при
сохранении
производительности]
Source: Intel
[+]Compared to Intel 32nm Technology
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
12
INTEL HPC - 2011
Технология Intel 22nm Tri-Gate
22nm
Planar
-0.2V
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
13
INTEL HPC - 2011
Технология Intel® Turbo Boost 2.0
Частота
“Dynamic”
Range @ Turbo
Frequency limits
Turbo bins
Base
Frequencies
Idle mode
Four-Core Turbo
Dual-Core Turbo
Single-Core Turbo
Умная и энергоргоэффективная производительность по
требованию
The number of Turbo bins shown is only for illustrative purposes and is not representative of the actual number of turbo bins available.
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
14
Масштабируя производительность
Производительность ядра
Будущее
16
“Sandy Bridge”
14
12
Intel® Xeon®
5600 Series
10
4
AESNI
Intel® Xeon®
5500 Series
8
6
AVX
SSE4.2
Intel® Xeon®
5200 Series
Intel® Xeon®
5400 Series
SSE4.1
Intel® Xeon® 5000 Series
2
Intel® Xeon®
0
2005
Source: Intel
Будущее
Potential future options, subject to change without notice.
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
INTEL HPC - 2011
Быстрее
Общая производ-ть
приложений
Улучшенная произв-ть потока
Улучшенная произв-ть на с
плавающей запятой и полосы
пропускания
Нерегулярный доступ к
данным
Сбалансированная
архитектура процессора и
системы
Простая разработка ПО и
поддержка
15
Intel® Advanced Vector Extension (AVX)
INTEL HPC - 2011
256-bit векторное расширение к SSE для F-интенсивных приложений
Свойства
Плюсы
Широкие вектора
До 2-х раз увеличение peak FLOPs
Increased from 128 bit to 256 bit
3 операнда,
не разрушающий синтах
Designed for efficiency and future extensibility
Меньше копий регистров, лучшее их
использование, параллелизм загрузок и
операций вычисления
128 bits (SSE)
XMM0
YMM0
256 bits (AVX)
Понятная стратегия перехода от SSE к AVX с Intel libraries, IPP и т.д.
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
http://www.intel.com/software/avx
16
INTEL HPC - 2011
Многоядерность.
подходы к решению.
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
17
INTEL HPC - 2011
Many Core и Multi-Core для HPC
Many Integrated Cores: 1-1.2GHz
Multi-core Intel Xeon: 2.2-3.5GHz
Die Size not to scale
 Каждое ядро меньше и менее
энергопотребляющее
 Лучше производительность потока
 Ниже производительность потока, но выше
общая производительность
 Multi-core предлагает прекрасную
производительность для широкаого спектра
приложений
 Лидирующее в индустрии соотношение
производительность на ватт для
последовательных и параллельных нагрузок
 Many core предлагает больший параллелизм
для компенсации меньшей частоты
 Общие программыне средства с Xeon ,
быстрая адоптация приложений и
оптимизация
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
 Больше ядро, больше кэш
18
INTEL HPC - 2011
Модель Tick-Tock
Tick
Tock
3D Tri-Gate
32nm
45nm
14nm
22nm
• • •
Intel® Core™
Nehalem
Sandy Bridge
Microarchitecture
Microarchitecture
Microarchitecture
Future
Microarchitecture
Будущее
SSE4.2/AESNI
AVX
Future Extensions
Potential future options, subject to change without notice.
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
19
INTEL HPC - 2011
Процессор Xeon® следующего поколения
Sandy Bridge “Tock”
Существенно лучшая производительность с большим
числом ядер, Intel® Hyper-threading и Turbo Technology
2x Flops / clock пиковая производительность с
инструкциями AVX
Доступны для построения Petascale решений
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
20
Intel® Xeon® Processor E5-2600
Gen3 x16
Gen3 x16
Gen3 x8
• Новая архитектура Sandy Bridge (on
32nm)
• Intel® Advanced Vector Extensions (AVX)
Higher BW
PCIe Gen3
• Integrated PCI Express*
I/O
QPI
QPI
QPI
C
C
• Новый тип соединения ядер - кольцо
C
• 4 канала DDR3 @ до 1600MHz
C
C
C
C
− До 512GB
(16 x 32GB QR RDIMMs)
− До 768GB
(24 x 32GB DR RDIMMs)
C
20MB
Cache
On-Die Interconnect
Significantly
Higher Cache BW
MC
DDR3
DDR3
DDR3
DDR3
Больше производительности, меньше энергопотребление
INTEL CONFIDENTIAL
Higher Socket
to Socket BW
4 channels
of DDR3
Лидирующа производительность
семейства Intel® Xeon® Processor E5
INTEL HPC - 2011
Максимально 152GF на сокет,
91% эффективности
172 пиковых GFLOPS / сокет –
2x ускорения используя Intel® AVX
**Первый в индустрии
интегрированный integrated
PCI Express* 3.0 – 2x пропускной
способности
Source: Nov’2011 Top500 listing: www.top500.org
#105 SNB (Intel Xeon E5: Sandy Bridge – EP, 8C 2.60GHz)
#20 ITL (AMD* Opteron* 6276, 16C 2.30GHz, Interlagos)
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. Configurations: Intel Internal measurements October 2011, See backup for configuration details. For more information go to
http://www.intel.com/performance . Any difference in system hardware or software design or configuration may affect actual performance. Copyright © 2010, Intel Corporation.
Copyright © 2011Intel Corporation. All rights reserved. 1 8 GT/s and 128b/130b encoding in PCIe* 3.0 specification is estimated to double the interconnect bandwidth over the PCIe* 2.0 specification
*Other brands and names are the property of their respective owners
22
Улучшенная производительность приложений в 1.7x
Семейство Intel® Xeon® Processor E5
Relative Geometric Mean Scores by segment. Actual performance will vary by workload. Higher is better
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
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.
Configurations: Intel Internal measurements October 2011, See backup for configuration details. For more
information go to http://www.intel.com/performance . Any difference in system hardware or software design or
configuration may affect actual performance. Copyright © 2010, Intel Corporation.
INTEL HPC - 2011
23
INTEL HPC - 2011
Многоядерность.
Many Integrated Core процессоры
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
24
INTEL HPC - 2011
Повышая производительность процессора
Many-Core для высокопараллельных приложений
FLOPS/Процессор
1.E+15 Peta
MIC
1.E+14
Many-Core
1.E+13
Tera-Scale R&D
1.E+12 Tera
1.E+11
1.E+10
1.E+09 Giga
1.E+06
Intel® Core™ mArch
Pentium® III Architecture
1.E+08
1.E+07
Multi-Core
Pentium® 4 Architecture
486
386
Pentium® II Architecture
Pentium® Architecture
Future options subject to change without notice. Source: Intel
Время
estimated
All dates, product descriptions, availability, and plans are forecasts and subject to change without notice.
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
25
INTEL HPC - 2011
Архитектура Intel® Many Integrated (MIC)
Первая индустриальная многоцелевая Many-Core архитектура
Knights Ferry Software Development Platform
Обработка высоко паралельных нагрузок
Утилизация маленьких IA ядер и множества потоков
Программируемость и Масштабируемость
Преимущества стандарной модели IA, модели памяти, средств
разработки
MIC
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
Промышленная система доступная в ближайшем будущем
Анонс Knights Corner на 22nm с >50 ядрами и лидирующей
производительностью
26
Many Integrated Core
• Платформа для разработки ПО
• Доступна для разработки
• 50+ ядер, 1.2 GHz
• 4 threads / core
• GDDR5
Платформа для разработки под Intel® MIC architecture
Intel® MIC архитектура:
VECTOR
IA CORE
VECTOR
IA CORE
…
VECTOR
IA CORE
VECTOR
IA CORE
INTERPROCESSOR NETWORK
COHERENT
CACHE
COHERENT
CACHE
COHERENT
CACHE
COHERENT
CACHE
…
…
COHERENT
CACHE
COHERENT
CACHE
COHERENT
CACHE
COHERENT
CACHE
INTERPROCESSOR NETWORK
VECTOR
IA CORE
VECTOR
IA CORE
…
VECTOR
IA CORE
VECTOR
IA CORE
MEMORY and I/O INTERFACES
FIXED FUNCTION LOGIC
Архитектура сопроцессора Intel
Много ядер и много потоков
Стандарты IA программирования и модели памяти
Программирование под Intel® MIC
Single Source
Compilers
and Runtimes
Intel® Xeon ®
processor
Intel® Xeon®
processor family
Intel® MIC
architecture
co-processor
Общие с Intel® Xeon®
• Языки
• C, C++, Fortran компиляторы
• Intel developer tools
и библиотеки
• Техника программирования и
оптимизации
• Поддержка экосистемы
Исключает необходимость двух типов программирования
INTEL HPC - 2011
Texas Advanced Computing Center
Работает с октября 2011
Целевые коды в Биологии, астрофизике,
Computational Fluid Dynamics (CFD)
TACC: Experimental Cluster
8 Intel® Xeon® nodes / 8 Intel® MIC nodes
Планируется январь 2013 – “Stampede”
10 petaflop Intel® Xeon/Intel ® MIC
US National Cyber Infrastructure Super Computer
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
30
INTEL HPC - 2011
“Stampede”
•
•
•
•
•
•
•
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
(posted 22-Sep-2011)
10 PFLOPS пиковой производительности
272 TB памяти
14 PB СХД
Запуск в 2013
Intel® Xeon® E5 (Sandy Bridge-EP) процессоры
Intel® MIC (Knights Corner) co-процессоры
FDR InfiniBand (56Gb/s) кластерная фабрика
31
INTEL HPC - 2011
Сметая барьеры
Пересекая TeraFlops реальной производительности
ASCI Red: 1TF
Knights Corner: 1TF
1997 Первая система 1 TF Sustained
2011 первый чип 1 TF Sustained
9298 Pentium II Xeon
22nm техпроцесс
OS: Cougar
OS: Linux
72 стоек
1 PCI express slot
Source and Photo: http://en.wikipedia.org/wiki/ASCI_Red
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
32
INTEL HPC - 2011
4 линейки продуктов
Efficient
Performance
Essential
Performance
Advanced
Performance
Distributed
Performance
C/C++ developers targeting
Intel® Atom™ Processors
based Embedded or Mobile
devices
C/C++ developers
Microsoft Visual Studio*
Take advantage of
multicore
C/C++ and Fortran
developers
Windows* and Linux*
High performance, cross
platform apps
C/C++ and Fortran
developers
on Windows* and Linux*
High performance MPI
clusters
Copyright © 2011Intel Corporation. All rights reserved.
*Other brands and names are the property of their respective owners
33
Спасибо!
Nikolay.Mester@intel.com
+7 495 641-4599
Intel® Xeon® Processor E5-2600
Features and Benefits
I/O
QPI
C
C
Higher Socket
to Socket BW
Significantly
Higher Cache BW
MC
DDR3
DDR3
DDR3
Expect Greater Performance & Lower Power
INTEL CONFIDENTIAL – NDA REQUIRED
QPI
C
20MB
Cache
Memory Controller
36
QPI
C
DDR3
• New Sandy Bridge micro-architecture on 32nm
Gen3 x16
Gen3 x16
process for greater performance-per-watt
Gen3 x8
• Intel® Advanced Vector Extensions (AVX)
accelerate floating point intensive applications
Higher BW
PCIe Gen3
C
• Intel® Turbo Boost 2.0 technology delivers more
turbo upside potential
C
• 40 lanes of integrated PCI Express* 3.0 for better
I/O latency and bandwidth
C
• High bandwidth, low latency, bi-directional ring
interconnect allows faster access to the 20MB
C
multi-banked last level cache
• Intel® Hyper-Threading technology enables up to On-Die Interconnect
16 computational threads
• Integrated memory controller with 4 channels of
DDR3 and 46b physical addressing facilitate
The “uncore” encompasses the I/O, QPI,
Last Level Cache (L3) and the Integrated
greater memory capacity
4 channels
of DDR3
A Family of Processors
to Meet Your HPC Needs
Powerful, Intelligent, Adaptable
EP offers up to 67% more performance over previous
generation Intel® Xeon® 5500 series
Single Programming Model
EX offers up to 39% greater performance over
previous generation Intel® Xeon® 7500 series
MIC Extreme performance for highly parallel
workloads.
The smarter way to scale and save energy;
delivering intelligent performance now
•
•
See legal disclaimer in backup
Results have been estimated based on internal Intel analysis and are provided for informational purposes only. Any difference in system hardware or software design or
configuration may affect actual performance. For more complete information about performance and benchmark results, visit www.intel.com/benchmarks
SNB Core: New Features
AVX
Instructions
Double Load BW
for L1 Data Cache
L1 Data Cache
Execution
Units
Deeper Buffer
and
Improved
Scheduling
Out-of-Order
Scheduling &
Retirement
Data
Poisoning
1G
Page TLB
L2 Cache
& Interrupt
Servicing
Memory Ordering
& Execution
Instruction
Decode &
Microcode
Paging
Branch Prediction
Improved
Branch
Prediction
Instruction Fetch
& L1 Cache
Higher Decode
Throughput
Improved
Instructions Fusing
Intel Confidential
Intel and the Intel logo are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. Other names and brands
may be claimed as the property of others. All products, dates, and figures are preliminary and are subject to change without any notice. Copyright © 2011, Intel Corporation.
38