Технологии Intel в центрах обработки данных

Intel® Xeon® Processor E5-2600 v3 Производительность Игорь Сысоев

17 сентября 2014

Intel Confidential

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 may make changes to specifications and product descriptions at any time, without notice. Designers must not rely on the absence or characteristics of any features or instructions marked “reserved” or “undefined.”

Intel reserves these for future definition and shall have no responsibility whatsoever for conflicts or incompatibilities arising from future changes to them.

Intel processor numbers are not a measure of performance. Processor numbers differentiate features within each processor family, not across different processor families. See

http://www.intel.com/products/processor_number for details.

The products described in this document 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.

The code names presented in this document are only for use by Intel to identify products, technologies, or services in development, that have not been made commercially available to the public, i.e., announced, launched

or shipped. They are not "commercial" names for products or services and are not intended to function as trademarks.

Contact your local Intel sales office or your distributor to obtain the latest specifications and before placing your product order.

Copies of documents which have an order number and are referenced in this document, or other Intel literature may be obtained by calling 1-800-548-4725 or by visiting Intel's website at

http://www.intel.com/design/literature.htm.

No computer system can provide absolute security under all conditions. Intel® Trusted Execution Technology (Intel® TXT) requires a computer with Intel® Virtualization Technology, an Intel TXT-enabled processor, chipset,

BIOS, Authenticated Code Modules and an Intel TXT-compatible measured launched environment (MLE). Intel TXT also requires the system to contain a TPM v1.s. For more information, visit

http://www.intel.com/technology/security

Requires a system with Intel® Turbo Boost Technology. Intel Turbo Boost Technology and Intel Turbo Boost Technology 2.0 are only available on select Intel® processors. Consult your PC manufacturer. Performance

varies depending on hardware, software, and system configuration. For more information, visit http://www.intel.com/go/turbo

Intel® Virtualization Technology requires a computer system with an enabled Intel® processor, BIOS, and virtual machine monitor (VMM). Functionality, performance or other benefits will vary depending on hardware and

software configurations. Software applications may not be compatible with all operating systems. Consult your PC manufacturer. For more information, visit http://www.intel.com/go/virtualization

The original equipment manufacturer must provide TPM functionality, which requires a TPM-supported BIOS. TPM functionality must be initialized and may not be available in all countries.

The TCO or other cost reduction scenarios described in this document are intended to enable you to get a better understanding of how the purchase of a certain products, including Intel products, combined with a

number of situation-specific variables, might affect your future cost and savings. Circumstances will vary and there may be unaccounted-for costs related to the use and deployment of a given product. Nothing in this

document should be interpreted as either a promise of or contract for a given level of costs.

All products, computer systems, dates and figures specified are preliminary based on current expectations, and are subject to change without notice.

Intel® Xeon®, Intel® Atom™ , Intel® Xeon Phi™, Intel® Core™ i5, Intel® Trusted Execution Technology (Intel® TXT), Intel® Advanced Encryption Standard New Instructions (Intel® AES-NI), Intel® Advanced Vector Extensions (Intel®

AVX), Intel® Open Network Platform, Intel® Lustre*, Intel® SSD, Intel® Ethernet Controllers and the Intel logo are trademarks of Intel Corporation or in the US and other countries.

Copyright © 2013, Intel Corporation. All rights reserved.

* Other brands and names may be claimed as the property of others.

Legal Disclaimer

2

http://www.intel.com/products/processor_number

http://www.intel.com/design/literature.htm

http://www.intel.com/technology/security

http://www.intel.com/go/turbo

http://www.intel.com/go/virtualization

3

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 http://www.intel.com/performance Intel does not control or audit third-party benchmark data or the web sites referenced in this document. You should visit the referenced web site and confirm whether referenced data are accurate. *Other names and brands may be claimed as the property of others. 1. Twenty-seven performance world records based on two-socket configurations. Source as of September 8, 2014. Full details available at: http://www.intel.com/content/www/us/en/benchmarks/server/xeon-e5-2600-v3/xeon-e5-2600-v3-summary.html 2. Comparison based on SPECpower_ssj2008 results published as of Sept 8 2014. New configuration: Sugon I620-G20 platform with two Intel Xeon Processor E5-2699 v3, 8x16GB DDR4-2133 DR-RDIMM, Microsoft Windows Server 2008 R2, IBM J9 VM, 10,599 overall ssj_ops/watt. Source, results submitted as of Sept 8 2014. 3. Source as of September 8, 2014. New configuration: Hewlett-Packard Company HP ProLiant ML350 Gen9 platform with two Intel® Xeon® Processor E5-2699 v3, Oracle Java Standard Edition 8 update 11, 190,674 SPECjbb2013-MultiJVM max-jOPS, 47,139 SPECjbb2013-MultiJVM critical-jOPS. Source. Baseline: Cisco Systems Cisco UCS C240 M3 platform with two Intel® Xeon® Processor E5-2697 v2, Oracle Java Standard Edition 7 update 45, 63,079 SPECjbb2013-MultiJVM max-jOPS , 23,797 SPECjbb2013-MultiJVM critical-jOPS. Source. 4. Comparisons based by Intel on publicly available specification sheet data &/or white papers

Delivering Leadership Performance 27 New World Records1!

Including World’s Most Energy Efficient Server2

Up to 3x performance improvement for efficient, high performance building blocks3

Up to 6x throughput

with latest SSD4

Up to 40GbE to improve

network virtualization

PUBLIC

http://www.intel.com/performance

http://www.spec.org/

http://h20195.www2.hp.com/V2/GetDocument.aspx?docname=4AA5-4748ENW&cc=us&lc=en

http://www.spec.org/jbb2013/results/res2014q1/jbb2013-20140121-00050.html

Key Differences Between E5-2600 v2 & E5-2600 v3 Xeon E5-2600 v2 “Ivy Bridge-EP”

Xeon E5-2600 v3 “Haswell-EP”

Core Count Up to 12 Cores Up to 18 Cores

Frequency TDP & Turbo Frequencies TDP & Turbo Freq

AVX & AVX Turbo Freq

AVX Support Intel® AVX

8 DP Flops/Clock Intel® AVX 2

16 DP Flops/Clock

Memory Type 4xDDR3 channels

RDIMM, UDIMM, LRDIMM 4xDDR4 channels RDIMM, LRDIMM

Memory Frequency (MHz)

1866 (1DPC), 1600, 1333, 1033 RDIMM: 2133 (1DPC), 1866 (2DPC), 1600

LRDIMM: 2133 (1&2DPC), 1600

QPI Speed Up to 8.0 GT/s Up to 9.6 GT/s

TDP

Up to 130W Server, 150W Workstation

Up to 145W Server, 160W Workstation Increase due to Integrated Voltage

Regulator

Power Management

Same P-states for all cores Same core & uncore frequency

Per-core P-states Independent uncore frequency scaling

Energy Efficient Turbo

4

PUBLIC

Advanced

25-30 MB LLC cache -2.5MB/core 9.6 GT/s QPI DDR4-2133 Intel® Hyper-Threading Intel® Turbo boost

Standard

Basic

Segmented Optimized

Frequency Optimized

≥2.5MB/core LLC cache – See SKU 9.6 GT/s QPI DDR4-2133 Intel® Hyper-Threading Intel® Turbo boost 2U cooling

Low Power 2.5 MB/Core LLC Cache E5-2650L v3: 9.6 GT/s QPI E5-2630L v3: 8.0 GT/s QPI E5-2650L v3: DDR4-2133 E5-2630L v3: DDR4-1866 Intel® Hyper-Threading Intel® Turbo boost

12C 120W 2.3 GHz

10C 105W 2.3 GHz

12C 120W 2.5 GHz

8C 85W 2.4 GHz

12C 65W 1.8 GHz

8C 55W 1.8 GHz

6C 85W 1.6 GHz

Intel® Xeon® Processor E5-2600 v3 Product Family SKU Stack

Workstation Only

12C 135W 2.6 GHz

10C 105W 2.6 GHz

10C 2S 160W 3.1 GHz

35-45 MB LLC cache - 2.5MB/core 9.6 GT/s QPI DDR4-2133 Intel® Hyper-Threading Intel® Turbo boost

14C (1U) 120W 2.3 GHz

14C (2U) 145W 2.6 GHz

14C (1U) 120W 2.0 GHz

E5-2603 v3

E5-2670 v3

E5-2690 v3

E5-2680 v3

E5-2650 v3

E5-2660 v3

8C 90W 2.6 GHz

6C 85W 2.4 GHz

8C (20MB) 135W 3.2 GHz (2U)

6C (20MB) 135W 3.4 GHz (2U)

4C (15MB) 135W 3.5GHz (2U)

6C 85W 1.9 GHz E5-2609 v3

E5-2620 v3

E5-2640 v3

E5-2630 v3

E5-2697 v3

E5-2695 v3

E5-2683 v3

E5-2637 v3

E5-2667 v3

E5-2687W v3

E5-2643 v3

E5-2630Lv3

E5-2650L v3

Legend

Medium Core Count (MCC)

High Core Count (HCC) 4C (10MB) 105W

3.0GHz (1U) E5-2623 v3

Low Core Count (LCC)

10MB LLC cache 8.0GT/s QPI DDR4-1866 Intel® Hyper-Threading Intel® Turbo boost 1U cooling

25MB LLC cache 9.6 GT/s QPI DDR4-2133 Intel® Hyper-Threading Intel® Turbo boost

18C (2U) 145W 2.3 GHz

16C (1U) 135W 2.3 GHz

E5-2699 v3

E5-2698 v3

15 MB LLC cache - 2.5MB/core 6.4 GT/s QPI DDR4-1600

15-20 MB LLC cache - 2.5MB/core 8.0 GT/s QPI DDR4-1866 Intel® Hyper-Threading Intel® Turbo boost

5

PUBLIC

1.20 1.24 1.28 1.30 1.41 1.42 1.44 1.60 1.60 1.67 1.70

1.91 1.97

2.45

3.02

0

0.5

1

1.5

2

2.5

3

3.5

E5

-26

97

v2

Ba

seli

ne

SP

EC

mp

iM*_

ba

se2

00

7

SP

EC

po

we

r*_

ssj2

00

8+

TP

C*-

H

SP

EC

fp*_

rate

_

ba

se2

00

6

SP

EC

int*

_ra

te_

ba

se2

00

6

SP

EC

om

pG

*_

ba

se2

01

2+

+

ST

RE

AM

(Tri

ad

)

VM

ma

rk*

2.5

SA

P S

D*

2-T

ier

mw

eb

SP

EC

vir

t*_

sc2

01

3

Lin

pa

ck

me

mca

che

d*

SP

EC

jbb

20

13

*

Mu

ltiJ

VM

crit

ica

l-jO

Ps

SP

EC

jbb

20

13

*

Mu

ltiJ

VM

ma

x-jO

Ps

Re

lati

ve

Pe

rfo

rma

nce

Intel® Xeon® Processor E5-2699 v3 (18C, 2.3GHz, 145W) vs. Intel® Xeon® Processor E5-2697 v2 (12C, 2.7GHz, 130W)

Intel® Xeon® Processor E5-2600 v3 Product Family Generational Performance Summary

+Baseline is E5-2660 v2 (10C, 2.2GHz, 95W) ++New configuration is E5-2690 v3 (12C, 2.6GHz, 135W)

Source as of Sept 8 2014, see next slide. 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 http://www.intel.com/performance Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in

this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. *Other names and brands may be claimed as the property of others

6

Intel® Xeon® Processor E5-2699 v3

PUBLIC


Intel® Xeon® Processor E5-2600 v3 Product Family 4-Year Refresh Performance Summary

7

3.1 3.2 3.2 3.3 3.3 3.4

6.3

8.7

0

1

2

3

4

5

6

7

8

9

10

X5690

2011 Launch

SPECpower*_

ssj2008+

STREAM

(Triad)

SPECint*_rate_

base2006

SAP SD*

2-Tier

VMmark* 2.x SPECfp*_rate_

base2006

Linpack SPECjbb2013*

MultiJVM

max-jOPs

Preliminary

Re

lati

ve

Pe

rfo

rma

nce

Intel® Xeon® E5-2699 v3 (18C, 2.3GHz, 145W) vs. X5690 (6C, 3.46GHz, 130W)

Intel® Xeon® Processor E5-2699 v3 – 2014 Launch

PUBLIC

+Baseline is X5675 (6C, 3.06GHz, 95W)

Source as of Sept 8 2014, see next slide. 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 http://www.intel.com/performance Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in

this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. *Other names and brands may be claimed as the property of others


0

200

400

600

800

1000

1200

1400

1600

2006

Xeon 5160

(2C,

3GHz, 80W)

2007

X5272

(2C, 3.4GHz,

80W)

2007

X5365

(4C,

3GHz, 150W)

2008

X5492

(4C, 3.4GHz,

150W)

2009

X5570

(4C, 2.93GHz,

95W)

2010

X5680

(6C, 3.33GHz,

130W)

2012

E5-2690

(8C, 2.9GHz,

135W)

2013

E5-2697 v2

(12C, 2.7GHz,

130W)

2014

E5-2699 v3

(18C, 2.3GHz,

145W)

SP

EC

int*

_ra

te_b

ase

20

06

pe

rfo

rma

nce

Intel® Xeon® Processor E5-2600 v2 Product Family Historical 2S Integer Throughput Performance

8

PUBLIC

Up to 22x increase in integer throughput on Intel® Xeon® since 2006

Higher is better Up to 22X

Source as of Sept 8 2014: results published or submitted to www.spec.org. 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 http://www.intel.com/performance Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. *Other names and brands may be claimed as the property of others



Intel® Xeon® Processor E5-2600 v2 Product Family Historical 2S Energy Efficient Performance

9

Up to 27x

Higher is better

Up to 27x increase in energy efficiency on Intel® Xeon® since 2006

Source as of Sept 8 2014: results published or submitted to www.spec.org. 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 http://www.intel.com/performance Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. *Other names and brands may be claimed as the property of others

0

2,000

4,000

6,000

8,000

10,000

12,000

2006

Xeon 5160

(2C,

3GHz, 80W)

2007

X5272

(2C, 3.4GHz,

80W)

2007

L5335

(4C,

2GHz, 50W)

2008

L5430

(4C, 2.66GHz,

50W)

2009

X5570

(4C, 2.93GHz,

95W)

2010

X5675

(6C, 3.06GHz,

95W)

2012

E5-2660

(8C, 2.2GHz,

95W)

2013

E5-2660 v2

(10C, 2.2GHz,

95W)

2014

E5-2699 v3

(18C, 2.3GHz,

145W)

SP

EC

po

we

r*_s

sj2

00

8

(Ov

era

ll s

sj_o

ps/

wa

tt)

PUBLIC



Intel® Xeon® Processor E5-2600 v3 Product Family Die Configurations

HCC consists of 4 columns and 2 Memory Controllers. MCC consists of 3 columns 4 (core + LLC slices) and 2 Memory Controllers. LCC consists of 2 columns 4 (core + LLC slices) and single Memory Controller.

Not representative of actual die-sizes, orientation and layouts – for informational use only.

10

14-18 Core (HCC) 4-8 Core (LCC) 10-12 Core (MCC)

10

PUBLIC

Cluster on Die (COD) Mode

11

Cluster0

Cbo LLC

Cbo LLC

Sbo

Sbo

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

HA0

QPI 0/1 IIO

HA1

Core

Core Core

Core

Core Core

Core

Core Core

Core

Core

Core

Core

Core

Cluster1

Cbo LLC

Cbo LLC

Cbo LLC

Core

Core

Core

Core Cbo LLC

COD Mode for 18C HSW-EP

Supported on 2S HSW-EP SKUs with 2 Home Agents (10+ cores)

Targeted at NUMA workloads where latency is more important than sharing data across Caching Agents (Cbo)

Reduces average LLC hit and local memory latencies

HA mostly sees requests from reduced set of threads which can lead to higher memory bandwidth

OS/VMM own NUMA and process affinity decisions

PUBLIC

Cluster on Die (COD) Mode

Cbo LLC

Cbo LLC

Sbo

Sbo

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

HA0

QPI 0/1

IIO

HA1

Core

Core Core

Core

Core Core

Core

Core

Core Core

Core

Core Core

Core


12

Cluster0

Cluster1


Cbo LLC

Cbo LLC

Sbo

Sbo

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

Cbo LLC

HA0

QPI 0/1

IIO

HA1

Core

Core Core

Core

Core Core

Core

Core

Core

Core

Core Cbo LLC

Core

PUBLIC

Refresh Performance

1.41

3.2

0

0.5

1

1.5

2

2.5

3

3.5

E5-2697 v2

(12C, 2.7GHz, 130W)

Q3'13 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

X5690

(6C, 3.46GHz, 130W)

Q1'11 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

Re

lati

ve

Pe

rfo

rma

nce

Generational Performance

13

Enterprise Benchmark Performance General Computing – SPECint*rate_base2006

Source as of Sept 8 2014. Baseline: Supermicro SuperServer 6027AX-TRF platform with two Intel Xeon Processor E5-2697 v2, Intel Compiler version 14.0.0.080, SPECint_rate_base2006 score: 981. Source.

New configuration: Supermicro SuperServer 6028UX-TR4 with two Intel Xeon Processor E5-2699 v3, 16x16GB DDR4-2133 DR-RDIMM, RHEL 6.5, Intel Compiler version 14.0.0.080, SPECint_rate_base2006

score: 1,390. Source, results submitted as of Sept. 8 2014. 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. Any difference in system hardware or software design or configuration may affect actual performance. Intel does not control or audit the

design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance

benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. For more information go to

http://www.intel.com/performance *Other names and brands may be claimed as the property of others.

• More cores & larger cache

• DDR4 memory • Intel® AVX2 • Cluster on Die snoop

mode

PUBLIC

http://www.spec.org/cpu2006/results/res2014q1/cpu2006-20131117-27458.html



Refresh Performance

1.41

3.2

0

0.5

1

1.5

2

2.5

3

3.5

E5-2697 v2

(12C, 2.7GHz, 130W)

Q3'13 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

X5690

(6C, 3.46GHz, 130W)

Q1'11 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

Re

lati

ve

Pe

rfo

rma

nce


14

Enterprise Benchmark Performance General Computing – SPECint*rate_base2006

Source as of Sept 8 2014 on SPECint*rate_base2006. Baseline Configuration: 2-socket Intel® Xeon® processor X5690 platform, 48GB memory, RHEL 6.2. Compiler version: Intel Compiler 12.1.0.

Source. Score: 425. 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. Any difference in system hardware or

software design or configuration may affect actual performance. Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel

encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and

reflect performance of systems available for purchase. For more information go to http://www.intel.com/performance *Other names and brands may be claimed as the property of others.

PUBLIC



Refresh Performance

1.30

3.4

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

E5-2697 v2

(12C, 2.7GHz, 130W)

E5-2699 v3

(18C, 2.3GHz, 145W)

X5680

(6C, 3.33GHz, 130W)

E5-2699 v3

(18C, 2.3GHz, 145W)

Re

lati

ve

Pe

rfo

rma

nce


15

Technical Computing Benchmark Performance Floating Point Throughput – SPECfp*rate_base2006

Source as of Sept 8 2014. New configuration: Supermicro SuperServer 6028UX-TR4 with two Intel Xeon Processor E5-2699 v3, 16x16GB DDR4-2133 DR-RDIMM, RHEL 6.5, Intel Compiler version 14.0.0.080,

SPECfp_rate_base2006 score: 935. Source, results submitted as of Sept. 8 2014. Baseline: Supermicro SuperServer 6027AX-TRF platform with two Intel Xeon Processor E5-2697 v2, Intel Compiler version

14.0.0.080, SPECfp_rate_base2006 score: 714. Source. 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. Any difference in

system hardware or software design or configuration may affect actual performance. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data

are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. For more information go to http://www.intel.com/performance

*Other names and brands may be claimed as the property of others.

• More Cores & Larger Cache

• DDR4 Memory • Intel® AVX2 • Cluster on Die Snoop

Mode

PUBLIC




Refresh Performance

1.30

3.4

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

E5-2697 v2

(12C, 2.7GHz, 130W)

E5-2699 v3

(18C, 2.3GHz, 145W)

X5690

(6C, 3.46GHz, 130W)

E5-2699 v3

(18C, 2.3GHz, 145W)

Re

lati

ve

Pe

rfo

rma

nce


16

Technical Computing Benchmark Performance Floating Point Throughput – SPECfp*rate_base2006

Source as of Sept 8 2014 on SPECfp*rate_base2006. Baseline Configuration: 2-socket Intel® Xeon® processor X5690 platform, 48 GB memory, SLES11* SP1, Intel Compiler 12.1.0. Source. Score: 271.

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. Any difference in system hardware or software design or configuration may affect

actual performance. Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the

referenced Web sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for

purchase. For more information go to http://www.intel.com/performance *Other names and brands may be claimed as the property of others.

PUBLIC



Ad

va

nce

d

Sta

nd

ard

B

asi

c

Se

gm

en

ted

O

pti

miz

ed

L

ow

P

ow

er

12C 120W 2.3 GHz

10C 105W 2.3 GHz

12C 120W 2.5 GHz

8C 85W 2.4 GHz

12C 65W 1.8 GHz

8C 55W 1.8 GHz

6C 85W 1.6 GHz

Intel® Xeon® Processor E5-2600 v3 Product Family Internal SPECCPU*rate_base2006 Performance Estimates

12C 135W 2.6 GHz

10C 105W 2.6 GHz

14C 120W 2.3 GHz

14C 145W 2.6 GHz

14C 120W 2.0 GHz

E5-2603 v3

E5-2670 v3

E5-2690 v3

E5-2680 v3

E5-2650 v3

E5-2660 v3

8C 90W 2.6 GHz

6C 85W 2.4 GHz

8C (20M) 135W 3.2 GHz

6C (20M) 135W 3.4 GHz

4C (15M) 135W 3.5GHz

6C 85W 1.9 GHz E5-2609 v3

E5-2620 v3

E5-2640 v3

E5-2630 v3

E5-2697 v3

E5-2695 v3

E5-2683 v3

E5-2637 v3

E5-2667 v3

E5-2643 v3

E5-2630L v3

E5-2650L v3

4C (10M) 105W 3.0GHz E5-2623 v3

18C 145W 2.3 GHz

16C 135W 2.3 GHz

E5-2699 v3

E5-2698 v3

Source as of July 2014: Platform with two E5 v3, Prefetchers, HT & Turbo Enabled, NUMA & COD (LCC-ES) mode, 8x16GB DDR4-2133, RHEL 6.4, IC14.0-AVX2, BIOS 27.R01. SIR for E5-2699 v3 used 8x32GB

LR-DIMM DDR4-2133. 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. 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 information go to http://www.intel.com/performance


Fre

qu

en

cy

Op

tim

ize

d

1060 771

1000 747

960 740

873 686

817 668

695 567

661 534

502 437

305 317 726 577

514 445

1290 870

1190 835

1190 831

1080 782

1050 781

797 632

SPECint*_rate _base2006 estimates

SPECfp*_rate _base2006 estimates

266 290

661 557

454 412

414 380

PUBLIC NEW/UPDATED


HSW IVB

Up to 42%, 32%

Up to 31%, 27%

Up to 30%, 26%

Up to 21%, 22%

Up to 35%, 28%

Up to 17%, 14%

Up to 24%, 22%

SPECCPU*base2006 Relative Performance

18



in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. 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 information go to http://www.intel.com/performance


1.00

0.95

0.66

0.44

0.58

0.48

1.00

0.97

0.77

0.57

0.68

0.59

0.56

0.39

1.00

0.92

0.92

0.84

0.81

1.00

0.95

0.95

0.89

0.89

0.66

0.51

12C 130W 2.7GHz E5-2697 v2

12C 115W 2.4GHz E5-2695 v2

14C 120W 2.3 GHz

14C 145W 2.6 GHz

14C 120W 2.0 GHz

E5-2697 v3

E5-2695 v3

E5-2683 v3

18C 145W 2.3 GHz

16C 135W 2.3 GHz

E5-2699 v3

E5-2698 v3

6C 135W 3.4 GHz

4C 135W 3.5GHz E5-2637 v3

E5-2643 v3

4C 105W 3.0GHz E5-2623 v3

8C 135W 3.2 GHz E5-2667 v3

4C 130W 3.5GHz E5-2637 v2

8C 130W 3.3GHz E5-2667 v2

6C 130W 3.5GHz E5-2643 v2

6C 60W 2.4GHz E5-2630L v2

10C 70W 1.7GHz E5-2650L v2

12C 65W 1.8 GHz

8C 55W 1.8 GHz E5-2630Lv3

E5-2650L v3

0.89

IVB-to-HSW %gains on SIR, SFR

SPECint*_rate _base2006

Relative Performance

SPECfp*_rate _base2006


0.83 Up to 4%, 8% 0.61 0.72

Up to 9%, 10% 0.51 0.64

Up to 12%, 11% 0.35 0.47

Up to 2%, 2% 0.32 0.43

PUBLIC

IVB SIR relative to

E5-2697 v2

IVB SFR relative to

E5-2697 v2

HSW SIR relative to

E5-2699 v3

HSW SFR relative to

E5-2699 v3


HSW IVB SPECint*_rate

_base2006 Relative Performance

Up to 25%, 24%

Up to 23%, 23%

Up to 24%, 26%

Up to 26%, 26%

Up to 35%, 27%

Up to 39%, 29%

Up to 23%, 18%

Up to 24%, 23%

SPECfp*_rate _base2006


19

0.92

0.89

0.84

0.77

0.70

0.56

0.52

0.44

0.26

0.19

4C 80W 2.5GHz E5-2609 v2

4C 80W 1.8GHz E5-2603 v2

10C 115W 2.5GHz E5-2670 v2

10C 130W 3.0GHz E5-2690 v2

10C 115W 2.8GHz E5-2680 v2

8C 95W 2.6GHz E5-2650 v2

10C 95W 2.2GHz E5-2660 v2

6C 80W 2.1GHz E5-2620 v2

8C 95W 2.0GHz E5-2640 v2

6C 80W 2.6GHz E5-2630 v2

0.94

0.92

0.89

0.84

0.80

0.67

0.63

0.56

0.38

0.30

12C 120W 2.3 GHz

10C 105W 2.3 GHz

12C 120W 2.5 GHz

8C 85W 2.4 GHz

6C 85W 1.6 GHz

12C 135W 2.6 GHz

10C 105W 2.6 GHz

E5-2603 v3

E5-2670 v3

E5-2690 v3

E5-2680 v3

E5-2650 v3

E5-2660 v3

8C 90W 2.6 GHz

6C 85W 2.4 GHz

6C 85W 1.9 GHz E5-2609 v3

E5-2620 v3

E5-2640 v3

E5-2630 v3

0.82

0.78

0.74

0.67

0.63

0.53

0.51

0.38

0.88

0.85

0.85

0.78

0.76

0.65

0.61

0.50



in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. 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 information go to http://www.intel.com/performance


Up to 28%, 25% 0.23 0.36

Up to 50%, 45% 0.20 0.33

SPECCPU*base2006 Relative Performance IVB-to-HSW %gains

on SIR, SFR

IVB SIR relative to

E5-2697 v2

IVB SFR relative to

E5-2697 v2

HSW SIR relative to

E5-2699 v3

HSW SFR relative to

E5-2699 v3

PUBLIC


Refresh Performance

1.24

3.1

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

E5-2660 v2

(10C, 2.6GHz, 95W)

Q3'13 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

X5675

(6C, 3.06GHz, 95W)

Q1'11 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

Re

lati

ve

Pe

rfo

rma

nce


20

Enterprise Benchmark Performance Energy Efficiency – SPECPower*_ssj2008

PUBLIC

Source as of Sept 8 2014. Sugon I620-G20 platform with two Intel Xeon Processor E5-2699 v3, 8x16GB DDR4-2133 DR-RDIMM, Microsoft Windows Server 2008 R2, IBM J9 VM, 10,599 overall ssj_ops/watt.

Source, results submitted as of Sept. 8 2014. Dell Inc. PowerEdge T620 platform with two Intel Xeon Processor E5-2660 v2, IBM J9 VM, 8,533 overall ssj_ops/watt. Source. HP ProLiant DL360 G7* with two

Intel® Xeon® Processor X5675, Source, Score: 3,329 overall ssj_ops/watt. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. 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. Any difference in system hardware or software design or configuration may affect actual

performance. Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web

sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. For more

information go to http://www.intel.com/performance *Other names and brands may be claimed as the property of others.


http://www.spec.org/power_ssj2008/results/res2013q4/power_ssj2008-20131029-00645.html

http://www.spec.org/power_ssj2008/results/res2011q4/power_ssj2008-20110713-00386.html


Refresh Performance

1.60

3.3

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

E5-2697 v2

(12C, 2.7GHz, 130W)

Q3'13 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

X5690

(6C, 3.46GHz, 130W)

Q1'11 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

Re

lati

ve

Pe

rfo

rma

nce


21

Enterprise Benchmark Performance Virtualization – VMmark 2.x

Source as of Sept 8 2014. New Configuration: Fujitsu PRIMERGY RX2540 M1 platform with two Intel Xeon Processor E5-2699 v3, VMware ESXi 5.5.0 U2, VMmark v2.5.2 score: 26.48 @ 22 tiles. Source.

Baseline: HP ProLiant DL380p Gen8 platform with two Intel Xeon E5-2697 v2, VMWare ESXi 5.1.0 U1, VMmark v2.5.1 score: 16.54 @ 14 tiles. Source .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. Any difference in system hardware or software design or configuration may affect actual performance. Intel does not control or audit the design or

implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data

are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. For more information go to http://www.intel.com/performance


• Higher CPU utilization • DDR4 Memory • Faster QPI Speed

PUBLIC

http://www.vmware.com/a/vmmark/

http://www.vmware.com/a/assets/vmmark/pdf/2013-09-10-HP-ProLiantDL380pG8.pdf


Refresh Performance

1.60

3.3

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

E5-2697 v2

(12C, 2.7GHz, 130W)

Q3'13 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

X5690

(6C, 3.46GHz, 130W)

Q1'11 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

Re

lati

ve

Pe

rfo

rma

nce


22

Enterprise Benchmark Performance Virtualization – VMmark2.x

Source as of Sept 8, 2014. Baseline: Fujitsu PRIMENERGY RX300 S6 with two Intel® Xeon® Processor X5690, VMware ESXi 4.1 U1, VMmark* v2.1.1 score: 7.59 @ 7 tiles, source. 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. Any difference in system hardware or software design or configuration may affect actual

performance. Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web

sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. For more

information go to http://www.intel.com/performance *Other names and brands may be claimed as the property of others.

PUBLIC

http://www.vmware.com/a/assets/vmmark/pdf/2011-10-18-Fujitsu-RX300S6.pdf


Refresh Performance

1.60

3.30

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

E5-2697 v2

(12C, 2.7GHz, 130W)

Q3'13 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

X5680

(6C, 3.33GHz, 130W)

Q1'11 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

Re

lati

ve

Pe

rfo

rma

nce


23

Enterprise Resource Planning Performance SAP Sales & Distribution* (2 Tier) on Linux

Source as of Sept 8 2014. results submitted or published to www.sap.com/benchmark, details on next slide. 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. Any difference in system hardware or software design or configuration may affect actual performance. Intel does not control or audit the design or implementation of third

party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and

confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. For more information go to http://www.intel.com/performance


• Cluster on Die snoop mode

• More cores & larger cache

PUBLIC

http://www.sap.com/benchmark


Refresh Performance

1.60

3.30

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

E5-2697 v2

(12C, 2.7GHz, 130W)

Q3'13 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

X5680

(6C, 3.33GHz, 130W)

Q1'11 Launch

E5-2699 v3

(18C, 2.3GHz, 145W)

Q3'14 Launch

Re

lati

ve

Pe

rfo

rma

nce


24

Enterprise Resource Planning Performance SAP Sales & Distribution* (2 Tier) on Linux

Source as of Sept 8 2014. results submitted or published to www.sap.com/benchmark, details on next slide. 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. Any difference in system hardware or software design or configuration may affect actual performance. Intel does not control or audit the design or implementation of third party

benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm

whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. For more information go to http://www.intel.com/performance


PUBLIC

http://www.sap.com/benchmark


Source as of Sept 8 2014: results published or submitted to www.spec.org, details on the next slide. 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





Intel® Xeon® Processor E5-2600 v3 Product Family Java Performance & Benefits of JDK8

1.45

2.62

3.58

0

0.5

1

1.5

2

2.5

3

3.5

4

SPECjbb2013-MultiJVM

max-jOPs

Preliminary

Re

lati

ve

Pe

rfo

rma

nce

E5-2697 v2

(12C, 2.7GHz, 130W)

JDK7u11

E5-2699 v3

(18C, 2.3GHz, 145W)

JDK7u11 + 2x memory

E5-2699 v3

(18C, 2.3GHz, 145W)

JDK8u5 + 2x memory

E5-2699 v3

(18C, 2.3GHz, 145W)

JDK8u11 + 2x memory + optimizations

25

PUBLIC

Up to 80% gain from JDK8



8.7

0.0

2.0

4.0

6.0

8.0

10.0

Re

lati

ve

Pe

rfo

rma

nce

Refresh Performance

X5690

(6C, 3.46GHz, 130W)

JDK7u11

E5-2699 v3

(18C, 2.3GHz, 145W)

JDK8u11

3.02

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Re

lati

ve

Pe

rfo

rma

nce


E5-2697 v2

(12C, 2.7GHz, 130W)

JDK7u45

E5-2699 v3

(18C, 2.3GHz, 145W)

JDK8u11

26

Java Performance SPECjbb*2013-MultiJVM max-jOPs

Source as of Sept 8 2014: results published or submitted to www.spec.org. Software and workloads used in performance tests may have been optimized for performance only on Intel







PUBLIC



16.3

0.0

5.0

10.0

15.0

20.0

Re

lati

ve

Pe

rfo

rma

nce

Refresh Performance

X5690

(6C, 3.46GHz, 130W)

JDK7u11

E5-2699 v3

(18C, 2.3GHz, 145W)

JDK8u20

2.45

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Re

lati

ve

Pe

rfo

rma

nce


E5-2697 v2

(12C, 2.7GHz, 130W)

JDK7u40

E5-2699 v3

(18C, 2.3GHz, 145W)

JKD8u20

27

Java Performance SPECjbb*2013-MultiJVM critical-jOPs

Source as of Sept 8 2014: results published or submitted to www.spec.org. Software and workloads used in performance tests may have been optimized for performance only on Intel







PUBLIC



Технологии Intel в центрах обработки данных

Technology

Transcript of Технологии Intel в центрах обработки данных