Date post: | 09-Jan-2017 |
Category: |
Technology |
Upload: | intel-it-center |
View: | 1,204 times |
Download: | 1 times |
Intel® Xeon® Processor E5-2600 v4 Product Family Applications ShowcasePerformance Orchestration security
Enterprise Database
March 2016
* Other Names and Brands May Be Claimed as the Property of Others 2
Software Partners Enabling on the Intel® Xeon® processor E5-2600 v4 Family
Intel® Xeon® processor E5-2600 v4platform value proposition
3
* Other Names and Brands May Be Claimed as the Property of Others 4
+
Intel® Xeon® processor E5-2600 v4
product family
Intel® SSD DC Family D3700/3600 & P3320
+Intel® Ethernet
OR
Intel® omni-path architecture
The New 2016 Platform Optimized for the Cloud
* Other Names and Brands May Be Claimed as the Property of Others 5
Intel® Xeon® Processor E5-2600 v4 Product FamilyNew Processor Technologies
• New Features and Enhancements• Intel® Transactional Synchronization Extensions (TSX)
• Optimized Intel® AVX 2.0
• Gather Improvements
• Posted Interrupts
• Page Modification Logging
• VM Enter/Exit latency Reduction
Performance
• New Capabilities• Cache Monitoring Technology (introduced on Xeon® E5-2600 v3)
• Cache Allocation Technology
• Code and Data Prioritization
• Memory Bandwidth Monitoring
Orchestration
• New Features• Crypto Speedup
• New Random Seed Generator (RDSEED)
• #VE (Processor Virtualization exception)
• Supervisor Mode Access Protection (SMAP)
Security
Performance info
Orchestration info
Security info
Inte
l® R
eso
urc
e
Dir
ect
or
Te
chn
olo
gy
* Other Names and Brands May Be Claimed as the Property of Others
Software platform value propositionIntel® Xeon® Processor E5-2600 v4 product Family
6
With software solutions running on the Intel® Xeon® processor E5-2600 v4 product family, you can:
Accelerate cloud deployments through a robust portfolio of capabilities in the software defined infrastructure space, to better manage shared platform resources. New features such as Intel® RDT can offer higher throughput for higher priority workloads for better utilization overall.
Catalyze business innovation with cloud-based service and application delivery models that increase efficiency and lower TCO. Software optimizations for this latest generation processor can deliver improved user experiences.
Break through the “performance wall” with the Intel Xeon processor E5-2600 v4 product family for many High Performance Computing applications.
Stay competitive by enabling the IT organization to be more agile, flexible and responsive to evolving business needs and to scale on demand.
* Other Names and Brands May Be Claimed as the Property of Others 7
Enterprise database
* Other Names and Brands May Be Claimed as the Property of Others
Enterprise Database
8
Up to
Faster1
54%
FASTER DATA FLOWS TO POWER YOUR BUSINESS
From high-volume transactions to real-time
analytics, today’s data platforms have to do
more and do it faster. The Intel® Xeon®
processor E5-2600 v4 product family is helping
database vendors meet the challenge by
processing more simultaneous transactions and
queries while delivering faster response times.
Hardware and software work together to deliver
a smarter foundation for high-speed business.
See the following slides for more information
and a look at the compelling performance
gains.1 – Demonstrated with Oracle MySQL* proof point, slide 9.
Telecom.
& Cloud
Financial
Services &
Security
Big Data
AnalyticsCore Business
Technical
Computing
* Other Names and Brands May Be Claimed as the Property of Others
0
1
2S Intel® Xeon® processor E5-2697 v3
2S Intel® Xeon® processor E5-2699 v4
MySQL* 5.7 increased transactions/second by up to 1.54X with the 2S Intel® Xeon® processor E5-2699 v4
Oracle*Oracle MySQL*
“The new Intel® Xeon® processor E5-2699 v4 gives a visible extra performance gain for MySQL 5.7 running OLTP workloads. We see more than 50K TPS and up to 1.54X performance gain for read-only scenario compared to the previous processor generation.1
Tomas Ulin - VP MySQL Engineering, Oracle
9
The Intel® Xeon® processor E5-2699 v4 improved MySQL transactions per second from 32,350 to 50,000, up to 1.54X more
Intel® Xeon® processor E5-2699 v4 with enhanced microarchitecture, additional enhanced cores and larger cache gives the performance boost, delivers a significant performance gain for database applications like MySQL, the world's most popular open source database
The Intel® Xeon® processor E5-2697 v4 delivers about 34%2 improvement in performance compared to the E5-2697 v3
.
Enterprise Database
Up to 54% more
1 - Testing conducted on ISV* software comparing 2S Intel® Xeon® Processor E5-2699 v4 to 2S Intel® Xeon® Processor E5-2699 v3. Testing done by Intel. For complete testing configuration details, SEE SLIDE 32. 2 - Estimated increase for 2S Intel® Xeon® Processor E5-2697 v4 to 2S Intel® Xeon® Processor E5-2697 v3. 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 complete information visit http://www.intel.com/performance.
Intel® Xeon® Processor E5-2600 v4 Family
Visible Performance Gains for the World's Most Popular Open Source Database Workload: Sysbench OLTP, transactions per second
No
rma
lize
d P
erf
orm
an
ce
www.mysql.com
* Other Names and Brands May Be Claimed as the Property of Others
0
1
2S Intel® Xeon® processor E5-2697 v3
2S Intel® Xeon® processor E5-2699 v4
MariaDB increased performance by up to 1.51X with the 2S Intel® Xeon® processor E5-2699 v4
MariaDB Corporation*MariaDB 10.1*
“The new Intel® Xeon® processor E5-2699 v4 continues the dual-socket performance evolution. Powered by additional enhanced cores it improves performance for traditional databases by up to 1.51x vs. previous Xeon’s generation for multi-table OLTP workloads. Our users and customers will get faster responses to their complex queries, and can reduce TCO at the same time—so they can realize even higher value with MariaDB.”
Michael Widenius, CTO for MariaDB Corporation
10
Robust, scalable, and reliable SQL server. Drop-in replacement for MySQL.
Intel® Xeon® processor E5-2600 v4 with enhanced microarchitecture, additional enhanced cores and larger cache gives the performance boost, delivers a significant performance gain for traditional database applications like MariaDB
Proved scalability for 2S, 88 H/W threads
The Intel® Xeon® processor E5-2697 v4 delivers about 33%2 improvement in performance compared to the E5-2697 v3
Enterprise Database
Up to 51%faster
1 - Testing conducted on ISV* software comparing 2S Intel® Xeon® Processor E5-2600 v4 to 2S Intel® Xeon® Processor E5-2600 v3. Testing done by ISV. For complete testing configuration details, SEE SLIDE 32. 2 - Estimated increase for 2S Intel® Xeon® Processor E5-2697 v4 to 2S Intel® Xeon® Processor E5-2697 v3. 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 complete information visit http://www.intel.com/performance.
Intel® Xeon® Processor E5-2600 v4 Family
Faster response times and lower TCOon a popular, open-source database Workload: Sysbench OLTP
No
rma
lize
d p
erf
orm
an
ce
https://mariadb.org/
* Other Names and Brands May Be Claimed as the Property of Others
0
1
2S Intel® Xeon® processor E5-2699 v3
2S Intel® Xeon® processor E5-2699 v4
NRI BOS performance increased by up to 1.39X with the 2S Intel® Xeon® processor E5-2699 v4, Intel®
AVX2, and cluster-on-die technology
Nomura Research Institute, Ltd.*Business Oriented Solution (BOS)
“NRI is excited that the Intel® Xeon® processor E5-2699 v4 improved BOS performance with more CPU cores and higher cache hit rate. BOS’s NUMA optimized deployment strategy perfectly matches the Cluster-on-Die technology in relaxing the memory bandwidth bottleneck, which directly boosted performance to amazing levels”1
Yuzo Ishida – Senior Application Engineer, NRI (BOS Architect)
11
www.nri.com
The Intel® Xeon® processor E5-2699 v4 with Intel® AVX2 reduced processing time from 53 to 38 seconds, up to 1.39X faster.
BOS handles both OLTP and OLAP workloads, sharing “single source of truth” over the extremely normalized single database schema.
End-Users can benefit from this application in turning operational data into real-time business insight without the delays of traditional data warehouses and data marts.
The increased number of CPU cores per NUMA node directly contributed for quick response time with more concurrent queries in building software-defined logical views on Java* Virtual Machine on the fly.
Enterprise Database
Up to 39%faster
1 - Testing conducted on ISV* software comparing 2S Intel® Xeon® Processor E5-2699 v4 to 2S Intel® Xeon® Processor E5-2699 v3. Testing done by NRI. For complete testing configuration details, SEE SLIDE 32.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 complete information visit http://www.intel.com/performance.
Operational cost can be significantly reduced by managing data as Single Source of Truth Workload: Real-time stock calculation; Benchmark schema derived
from TPC-H* queries.
No
rma
lize
d P
erf
orm
an
ce
Intel® Xeon® Processor E5-2600 v4 Product Family
Intel® Xeon® + AVX2 + 10GbE adapter + CoD
* Other Names and Brands May Be Claimed as the Property of Others
0
1
2S Intel® Xeon® processor E5-2699 v3
2S Intel® Xeon® processor E5-2699 v4
Workload: HDB in-memory DBMS
Altibase HDB* improved performance by up to 1.23X with the 2S Intel® Xeon® processor E5-2699 v41
Altibase*In-memory DBMS for the Enterprise
“HDB* is aiming for real-time access to time critical hot data. The Intel® Xeon® processor E5-2699 v4 enables HDB to achieve more analytics results and faster transactions storing and manipulating data.”
Kim, Gang-Seok – Head of Research and Development Dept., Altibase
12
http://altibase.com
HDB provides fast data processing speeds for online transactional processing and online analytical processing
In-memory DBMS is required to deal with low latency and high throughput. Intel® Xeon® Processor E5 2699 v4 can achieve this mission-critical transactions with more stabilities
Enterprise Database
1 - Testing conducted on ISV* software comparing 2S Intel® Xeon® Processor E5-2699 v4 to 2S Intel® Xeon® Processor E5-2699 v3. Testing done by ISV. For complete testing configuration details, SEE SLIDE 32.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 complete information visit http://www.intel.com/performance.
Intel® Xeon® Processor E5-2600 v4 Family
Optimized performance for maximizing in-memory DBMS workloads
No
rma
lize
d P
erf
orm
an
ce
ISV logo
Up to 23% Faster
* Other Names and Brands May Be Claimed as the Property of Others
0
1
2S Intel® Xeon® processor E5-2699 v3
2S Intel® Xeon® processor E5-2699 v4
IBM DB2 increased performance with the 2S Intel® Xeon® processor E5-2699 v4
IBM* DB2 10.5 (BLU*)
DB2 is IBM's flagship database product offers industry leading performance, scale and reliability, in the cloud and on premise. DB2 v10.5 adds support for in-memory column store tables (BLU) for analytics workloads that vastly accelerates performance while lowering costs. IBM DB2 10.5 (BLU) produced up to a 18% improvement with the Intel® Xeon® processor E-2699 v4 compared to the previous generation processor.
13
Boosted by Intel® C++ Compiler for faster execution.
BLU* acceleration optimized for Intel® SIMD instructions.
BLU optimization for highly-parallel processing on Intel® Xeon® processors.
Enterprise Database
Up to 18%faster
1 - Testing conducted on ISV* software comparing 2S Intel® Xeon® Processor E5-2699 v4 to 2S Intel® Xeon® Processor E5-2699 v3. Testing done by ISV. For complete testing configuration details, SEE SLIDE 32. 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 complete information visit http://www.intel.com/performance.
Intel® Xeon® Processor E5-2600 v4 Product Family
DB2: Accelerated for scalability and performance todayWorkload: IBM Proof of Performance and Scalability (POPS) 2TB database.
No
rma
lize
d P
erf
orm
an
ce
www.ibm.com
* Other Names and Brands May Be Claimed as the Property of Others
0
1
2S Intel® Xeon® processor E5-2699 v3
2S Intel® Xeon® processor E5-2699 v4
IBM IWA increased performance with the 2S Intel® Xeon® processor E5-2699 v4
IBM* Informix Warehouse Accelerator* (IWA)
The IBM Informix Warehouse Accelerator is an in-memory database accelerator for the IBM Informix database server, which includes deep columnar and compression technology to improve query performance and reduce typical query tuning and database tuning. Multi-core and vector optimized algorithms paired with massive parallelism exploit the increased cores of the Intel® Xeon® processor E5-2600 v4 product family with SIMD technology. Frequency partitioning in combination with predicate evaluation on compressed data allow optimal access of the data residing in memory. IBM IWA produced up to a 16% improvement with the Intel® Xeon® processor E-2699 v4 compared to the previous generation processor.
14
www.ibm.com
IWA optimized for Intel® SIMD instructions.
IWA optimization for highly-parallel processing on Intel® Xeon® processors.
Enterprise Database
Up to 16%faster
1 - Testing conducted on ISV* software comparing 2S Intel® Xeon® Processor E5-2699 v4 to 2S Intel® Xeon® Processor E5-2699 v3. Testing done by ISV. For complete testing configuration details, SEE SLIDE 32. 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 complete information visit http://www.intel.com/performance.
Intel® Xeon® Processor E5-2600 v4 Product Family
Brilliant marriage of the Intel® Xeon® processor E5-2699 v4 hardware and database software Workload: Analytical queries on noSQL sensor Time Series data (2.5
billion records)
No
rma
lize
d p
erf
orm
an
ce
* Other Names and Brands May Be Claimed as the Property of Others
2x2 and Better together stories
15
* Other Names and Brands May Be Claimed as the Property of Others 16
More Transactions per Second with Microsoft SQL Server* and Intel® Xeon® Processor E5-2699 v4
Handling More Peak Database Transactions Together
2012 2016
Software UpgradeMicrosoft SQL Server* 2016
Performance and Scalability Enhancements
In-Memory Online Transaction Processing
Faster query response times
Hardware UpgradeIntel® Xeon® processor E5-2699 v4 Increased parallelism - up to 22 cores, up to 2-sockets per server
natively
Up to 1.5TB of DDR4 memory supported per socket1
Reduce transaction bottlenecks with newer
Intel® SSD DC S3700 Series
SoftwareBaseline Microsoft
SQL Server2012*
Enterprise Edition
Better Together Up to 2.6x More On-Line Transaction Processing
requests per second for peak times2
Delivers in-memory analytics and business insight through a modern, fully-integrated enterprise-class data platform
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.
1 Up to 3TB of memory supported per 2-socket server. Each socket supports 4 memory channels @ 3 DIMMs per channel = 12 DIMM slots per socket. Using 128GB memory DIMMS = 1.5TB per socket = 3 TB of memory per 2P server.
2 Up to 2.6x more transactions per second comparing HPE* Proliant DL380p Gen8 Server with two Intel® Xeon® processor E5-2690, 256GB Memory (source: http://www.tpc.org/4060) to a Lenovo* x3650 M5 Server with two Intel® Xeon® processor E5-2699 v4, 512GB Memory (source: to be updated on 31 March 2016).
For more information, visit intel.com/performance
HardwareBaseline Intel® Xeon®
processor
E5-2690
Benchmark : TPC-E*
* Other Names and Brands May Be Claimed as the Property of Others
1,881
4,938
0
1,000
2,000
3,000
4,000
5,000
6,000
Intel® Xeon® processor E5-2690 + Microsoft SQL Server* 2012 EE Intel® Xeon® processor E5-2699 v4 + Microsoft SQL Server* 2016 EE
Tra
nsa
ctio
ns
pe
r S
eco
nd
17
More Transactions per Second with Microsoft SQL Server* and Intel® Xeon® Processors E5-2699 v4
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.
Configuration details on next slide.
For more complete information visit http://www.intel.com/performance.
Higher is better
Up to 2.6x moreTransactions/Second
* Other Names and Brands May Be Claimed as the Property of Others 18
Configuration DetailsRefresh Microsoft SQL Server* OLTP with Intel® Xeon® Processor E5-2699 v4
ProcessorIntel® Xeon® processor E5-2690
(8 cores, 16 threads)Intel® Xeon® processor E5-2699 v4
(22 Cores, 44 Threads)
Memory 256 GB (16x16 GB DDR3-1600) 512 GB (16 x 32 GB DDR4-2400)
Disk32 x 400GB SAS SSD (Database)2 x 500GB SAS 7.2K RPM HDD
2 x 900GB SAS HDD (Tempdb)4 x 800GB SAS SSD (DB Log)
53 x 800GB SAS SSD (Database)
OS Distribution Microsoft Windows Server* 2012 SE Microsoft Windows Server* 2012 SE
Software Microsoft SQL Server* 2012 EE Microsoft SQL Server* 2016 EE
OLTP Throughput 1,881 K transactions per second 4,938 K transactions per second
Source http://www.tpc.org/4060 To be update 31 March 2016
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 complete information visit http://www.intel.com/performance.
* Other Names and Brands May Be Claimed as the Property of Others 19
Deliver Faster Insights with Real Value with Microsoft SQL Server* and Intel® Xeon® Processor E5-2699 v4Microsoft SQL Server* and Intel® architecture deliver a fully-integrated data platform for in-memory computing and real-time business intelligence.
2013 2016
Software UpgradeMicrosoft SQL Server* 2016 EE
Performance and Scalability Enhancements
In-Memory Online Transaction Processing
Faster query response times
SoftwareBaseline Microsoft Windows
Server* 2012 R2 SE
Microsoft SQL
Server* 2014 EE
Better Together Up to 2.2X improvement in virtualized
performance2
Delivers deeper analytics and business insight through a modern, fully-integrated enterprise-class data platform
HardwareBaseline Intel® Xeon®
processor E5-
2690 v2
Hardware UpgradeIntel® Xeon® processor E5-2699 v4
Built on Intel’s industry-leading 14nm second-
generation Tri-Gate transistor technology
Increased parallelism—up to 22 cores/44 threads per
socket
Up to 1.5TB of DDR4 memory supported per socket1
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 complete information visit http://www.intel.com/performance. 1. Up to 3TB of memory supported per 2-socket server. Each socket supports 4 memory channels @ 3 DIMMs per channel = 12 DIMM slots per socket. Using 128GB memory DIMMS = 1.5TB per socket = 3 TB of memory per 2P server2 Up to 2.2x performance improvement based on TPC-H workload comparing baseline Cisco UCS C240 M3 Model Enterprise Server with two Intel® Xeon® processor E5-2590 v2 3GHz, 768GB memory scoring 304,361.7 QphH, costing $221,937, $0.73/QphH (source: http://tpc.org/3301 ) to HPE* Proliant DL380 Gen9 Server with two Intel® Xeon® processor E5-2699 v4 2.2GHz, 768GB memory scoring 678,492.3 QphH costing $428,737, $0.64/QphH (Source: to be updated on 31 March 2016)
Benchmark : TPC-H*
* Other Names and Brands May Be Claimed as the Property of Others
304,361
678,492
0
100,000
200,000
300,000
400,000
500,000
600,000
700,000
800,000
Intel® Xeon® processor E5-2690 + Microsoft SQL
Server* 2014 EE
Intel® Xeon® processor E5-2699 v4 + Microsoft SQL
Server* 2016 EE
Qp
hH
@ 1
00
0G
B
20
Deliver Faster Insights with Real Value with Microsoft SQL Server* and Intel® Xeon® Processor E5-2699 v4
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.
Configuration details on next slide.
For more complete information visit http://www.intel.com/performance.
Higher is BetterUp to 2.2X more
composite queries/hour
Microsoft SQL Server* and Intel® architecture deliver a fully-integrated data platform for in-memory computing and real-time business intelligence.
* Other Names and Brands May Be Claimed as the Property of Others 21
Configuration Details
Deliver Faster Insights with Real Value with Microsoft SQL Server* and Intel® Xeon® Processor E5-2699 v4
Processor Intel® Xeon® processor E5-2690 v2 Intel® Xeon® processor E5-2699 v4
Cores/Threads 10/20 22/44
Memory 768GB (24 x 32GB DDR3-1866) 768GB (16 x 32GB PC4-2400T)
Operating System Microsoft Windows Server* 2012 R2 SEMicrosoft Windows Server* 2012 R2
SE
Software Microsoft SQL Server* 2014 EE Microsoft SQL Server* 2016 EE
QphH@1000GB 304,361.7 678,492.3
$/QphH@1000GB $0.73 USD $0.64 USD
Source http://www.tpc.org/3301 To be updated on 31 March 2016
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, visit intel.com/performance
* Other Names and Brands May Be Claimed as the Property of Others
Maximize Throughput with Intel® Xeon® Processor E5-2699 v4 and SAP* ERP
2012 2016
Software BaselineSAP* EHP5 for SAP ERP 6.0 Sybase* ASE 16.0
Fast queries with automatic optimizations
Efficient Indexing
Hardware UpgradeIntel® Xeon® processor E5-2699 v4
Increased parallelism - up to 22 cores, up to 2-sockets
per server natively
Up to 1.5TB of DDR4 memory supported per socket1
Software Baseline• SAP* EHP5 for
SAP ERP 6.0 • Sybase ASE 15.7
Better Together
Up to 3.2x more benchmark users2
Improve querying speed
Support more users3
Hardware Baseline• Intel® Xeon®
processor E5-2690
SAP and SAP NetWeaver are the registered trademarks of SAP AG in Germany and in several other countries. See www.sap.com/benchmark for more information. 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, visit intel.com/performance*1,2,3 - Additional information on the following page.
Benchmark : SAP SD 2-Tier
Process more Orders per Hour
85
* Other Names and Brands May Be Claimed as the Property of Others
Maximize Throughput with Intel® Xeon® Processor E5-2699 v4 and SAP* ERP
SAP and SAP NetWeaver are the registered trademarks of SAP AG in Germany and in several other countries. See www.sap.com/benchmark for more information. 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, visit intel.com/performance1. Up to 3TB of memory supported per 2-socket server. Each socket supports 4 memory channels @ 3 DIMMs per channel = 12 DIMM slots per socket. Using 128GB memory DIMMS = 1.5TB per socket = 3 TB of memory per 2P server2,3 Up to 3.2x more SAP SD 2-tier benchmark users running SAP SD 2-tier workload comparing Cisco* UCS B200 M3 server with two Intel® Xeon® processor E5-2690, 256GB memory running SAP* EHP 5 for ERP 6.0 + Sybase* ASE 15.7 scoring 6,530 benchmark users (Source) to a Dell* PowerEdge R730 server with two Intel Xeon processor E5-2699 v4, 512GB memory running SAP* enhancement package 5 for SAP ERP 6.0 + Sybase* ASE 16.0, scoring 21,050 benchmark users (Source)Configuration Details on next slide
Higher is Better
6,530
21,050
0
5,000
10,000
15,000
20,000
25,000
Intel® Xeon® processor E5-2690 Intel® Xeon® processor E5-2699 v4
Nu
mb
er
of
SA
P*
SD
Be
nch
ma
rk U
sers
Up to 3.2X more SAP SD Benchmark Users
86
* Other Names and Brands May Be Claimed as the Property of Others
Configuration Details: Maximize Throughput with Intel® Xeon® Processor E5-2699 v4 and SAP* ERP
ProcessorIntel® Xeon® processor E5-2690
(8 cores, 16 threads)Intel® Xeon® processor E5-2699 v4
(22 cores, 44 threads)
Memory 256GB 512GB
SAP enhancement package SAP* enhancement package 5 for SAP ERP 6.0 SAP* enhancement package 5 for SAP ERP 6.0
OS Distribution Red Hat* Enterprise Linux 6.3 Red Hat* Enterprise Linux 7.2
Database Sybase* ASE 15.7 Sybase* ASE 16.0
Number of SAP SD benchmark users 6,530 21,050
Source Source To be updated on 31 March 2016
SAP and SAP NetWeaver are the registered trademarks of SAP AG in Germany and in several other countries. See www.sap.com/benchmark for more information. 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, visit intel.com/performance
Process more Orders per Hour
87
* Other Names and Brands May Be Claimed as the Property of Others
More Queries per Second with MariaDB* and Intel® Xeon® Processor E5-2697 v4 ̂
25
Handling More Peak Database Queries Together
2012 2016
Software UpgradeUbuntu* 15.10
MariaDB* 10.1Robust, scalable, and reliable SQL server
Faster query response times
In-Memory Online Transaction Processing
Hardware Upgrade^Intel® Xeon® processor E5-2697 v4 Increased parallelism - up to 22 cores, up to 2-sockets per server
natively
Up to 1.5TB of DDR4 memory supported per socket1
Reduce transaction bottlenecks with newer
Intel® SSD Data Center Series
SoftwareBaseline Ubuntu* 12.04
MariaDB* 10.0.4
Better Together Up to 2.8X (estimated)^ more queries per second
for peak times2
Intel® and MariaDB delivers faster response times to customer’s complex queries
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 or simulated using internal Intel analysis or architecture simulation or modeling, and provided to you for informational purposes. Any differences in your system hardware, software or configuration may affect your actual performance.
1 Up to 3TB of memory supported per 2-socket server. Each socket supports 4 memory channels @ 3 DIMMs per channel = 12 DIMM slots per socket. Using 128GB memory DIMMS = 1.5TB per socket = 3 TB of memory per 2P server.
2 Up to 2.8x estimated more queries per second comparing baseline Intel® Server platform with two Intel® Xeon® processor E5-2680, 64GB Memory (source: Intel® technical report #1942) to an Intel® Server platform with two Intel® Xeon® processor E5-2697 v4, 64GB Memory (source: Intel® technical report #1942).
For more information, visit intel.com/performance
HardwareBaseline Intel® Xeon®
processor
E5-2680
Benchmark : Sysbench OLTP
+
^ The results shown for the Intel® Xeon® processor E5-2697 v4 have been estimated based on measured data obtained using the Intel® Xeon® processor E5-2699 v4
* Other Names and Brands May Be Claimed as the Property of Others
14,393
41,412
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
Intel® Xeon® processor E5-2680 Intel® Xeon® processor E5-2697 v4
Qu
eri
es
pe
r S
eco
nd
More Queries per Second with MariaDB* and Intel® Xeon® Processors E5-2697 v4^
26
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 or simulated using internal Intel analysis or architecture simulation or modeling, and provided to you for informational purposes. Any differences in your system hardware, software or configuration may affect your actual performance.
Configuration details on next slide.
For more complete information visit http://www.intel.com/performance.
Higher is better Up to 2.8X^ moreQueries/Second
+
^ The results shown for the Intel® Xeon® processor E5-2697 v4 have been estimated based on measured data obtained using the Intel® Xeon® processor E5-2699 v4
* Other Names and Brands May Be Claimed as the Property of Others
Configuration Details
27
More Queries per Second with MariaDB* and Intel® Xeon® Processor E5-2697 v4^
ProcessorIntel® Xeon® processor E5-2680
(16 cores, 32 threads)Intel® Xeon® processor E5-2697 v4
(18 Cores, 36 Threads)
Memory 64 GB (8x 8GB DDR3-1600) 64 GB (8x 8GB DDR4-2133)
OS Distribution Ubuntu* 12.04 Ubuntu* 15.10
Database Version MariaDB* 10.0.4 MariaDB* 10.1
OLTP Throughput 14,393 queries per second 41,412^ queries per second
Source Intel Internal Testing Intel Technical Report #1942
+
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 or simulated using internal Intel analysis or architecture simulation or modeling, and provided to you for informational purposes. Any differences in your system hardware, software or configuration may affect your actual performance.
For more complete information visit http://www.intel.com/performance.
^ The results shown for the Intel® Xeon® processor E5-2697 v4 have been estimated based on measured data obtained using the Intel® Xeon® processor E5-2699 v4
* Other Names and Brands May Be Claimed as the Property of Others
More Queries per Second with MySQL* and Intel® Xeon® Processor E5-2697 v4^
28
Handling More Peak Database Queries Together
2012 2016
Software UpgradeUbuntu* 15.10
MySQL* 5.7 Faster query response times
New Optimizer
Native JSON Support
Hardware Upgrade^Intel® Xeon® processor E5-2697 v4 Increased parallelism - up to 22 cores, up to 2-sockets per server
natively
Up to 1.5TB of DDR4 memory supported per socket1
Reduce transaction bottlenecks with newer
Intel® SSD Data Center Series
SoftwareBaseline Ubuntu* 12.04
MySQL* 5.6.12
Better Together Up to 2.8X^ (estimated) more queries per second
for peak times2
Intel® Xeon® processors and MySQL provides customers with more queries per second and faster response times.
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 or simulated using internal Intel analysis or architecture simulation or modeling, and provided to you for informational purposes. Any differences in your system hardware, software or configuration may affect your actual performance.
1 Up to 3TB of memory supported per 2-socket server. Each socket supports 4 memory channels @ 3 DIMMs per channel = 12 DIMM slots per socket. Using 128GB memory DIMMS = 1.5TB per socket = 3 TB of memory per 2P server.
2 Up to 2.8x^ more estimated queries per second comparing baseline Intel® Server platform with two Intel® Xeon® processor E5-2680, 64GB Memory (source: Intel® technical report #1942) to an Intel® Server platform with two Intel® Xeon® processor E5-2697 v4, 64GB Memory (source: Intel® technical report #1952).
For more information, visit intel.com/performance
HardwareBaseline Intel® Xeon®
processor
E5-2680
Benchmark : Sysbench OLTP
+
^ The results shown for the Intel® Xeon® processor E5-2697 v4 have been estimated based on measured data obtained using the Intel® Xeon® processor E5-2699 v4
* Other Names and Brands May Be Claimed as the Property of Others
15,247
43,786
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
50,000
Intel® Xeon® processor E5-2680 Intel® Xeon® processor E5-2697 v4
Qu
eri
es
pe
r S
eco
nd
More Queries per Second with MySQL* and Intel® Xeon® Processor E5-2697 v4^
29
Higher is betterUp to 2.8X^ more
Queries/Second
+
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 or simulated using internal Intel analysis or architecture simulation or modeling, and provided to you for informational purposes. Any differences in your system hardware, software or configuration may affect your actual performance.
Configuration details on next slide.
For more complete information visit http://www.intel.com/performance.
^ The results shown for the Intel® Xeon® processor E5-2697 v4 have been estimated based on measured data obtained using the Intel® Xeon® processor E5-2699 v4
* Other Names and Brands May Be Claimed as the Property of Others
Configuration Details
30
More Queries per Second with MySQL* and Intel® Xeon® Processor E5-2697 v4
ProcessorIntel® Xeon® processor E5-2680
(16 cores, 32 threads)Intel® Xeon® processor E5-2697 v4
(18 Cores, 36 Threads)
Memory 64 GB (8x 8GB DDR3-1600) 64 GB (8x 8GB DDR4-2133)
OS Distribution Ubuntu* 12.04 Ubuntu* 15.10
Database Version MySQL* 5.6.12 MySQL* 5.7
OLTP Throughput 15,247 queries per second 43,786 queries per second^
Source Intel Internal Testing Intel Technical Report #1952
+
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 or simulated using internal Intel analysis or architecture simulation or modeling, and provided to you for informational purposes. Any differences in your system hardware, software or configuration may affect your actual performance.
For more complete information visit http://www.intel.com/performance.
^ The results shown for the Intel® Xeon® processor E5-2697 v4 have been estimated based on measured data obtained using the Intel® Xeon® processor E5-2699 v4
Configurations & Disclaimers
31
* Other Names and Brands May Be Claimed as the Property of Others
Proof point Configuration Details:Oracle MySQL: Sysbench OLTP, transactions per second (TPS) workload. Testing by Intel, 3/3/2016BASELINE: 1-Node, 2 x Intel® Xeon® Processor E5-2697 v3 on Grantley-EP (Wellsburg) with 64 GB Total Memory, 8 slots / 8 GB / 2133 MT/s / DDR4 RDIMM, In-memory storage (RAM disk), disk I/O is not a bottleneck, CPU bound, default BIOS settings, HT and turbo on, on Ubuntu 15.10 using Oracle's MySQL 5.7, Sysbench OLTP 0.4.12. Data Source: Request Number: 1952NEW: 1-Node, 2 x Intel® Xeon® Processor E5-2699 v4 on Grantley-EP (Wellsburg) with 64 GB Total Memory, 8 slots / 8 GB / 2133 MT/s / DDR4 RDIMM, In-memory storage (RAM disk), disk I/O is not a bottleneck, CPU bound, default BIOS settings, HT and turbo on, on Ubuntu 15.10 using Oracle's MySQL 5.7, Sysbench OLTP 0.4.12. Data Source: Request Number: 1952
MariaDB: Sysbench OLTP 0.4.12, transactions per second (TPS) workload. Testing by Intel, 3/3/2016BASELINE: 1-Node, 2 x Intel® Xeon® Processor E5-2697 v3 on Grantley-EP (Wellsburg) with 64 GB Total Memory, 8 slots / 8 GB / 2133 MT/s / DDR4 RDIMM, In-memory storage (RAM disk), disk I/O is not a bottleneck, CPU bound, default BIOS settings, HT and turbo on, on Ubuntu 15.10 using MariaDB 10.1, Sysbench OLTP 0.4.12. Data Source: Request Number: 1942NEW: 1-Node, 2 x Intel® Xeon® Processor E5-2699 v4 on Grantley-EP (Wellsburg) with 64 GB Total Memory, 8 slots / 8 GB / 2133 MT/s / DDR4 RDIMM, In-memory storage (RAM disk), disk I/O is not a bottleneck, CPU bound, default BIOS settings, HT and turbo on, on Ubuntu 15.10 using MariaDB 10.1,. Data Source: Request Number: 1942
Nomura Research Institute, Ltd.* BOS: Real-time stock calculation workload; Benchmark schema derived from TPC-H* queries. Testing done by NRI and Intel, 1/28/2016BASELINE: 1-Node, 2 x Intel® Xeon® Processor E5-2699 v3 on Grantley-EP (Wellsburg) with 64 GB Total Memory, 8 slots / 8 GB / 2133 MT/s / DDR4 RDIMM, SSD SATA P3700 400G, on Red Hat Enterprise Linux* 7.0 kernel 3.10.0-123 using Java 8u25, PostgreSQL 9.4. Data Source: Request Number: 1951NEW: 1-Node, 2 x Intel® Xeon® Processor E5-2699 v4 on Grantley-EP (Wellsburg) with 128 GB Total Memory, 8 slots / 16 GB / 2400 MT/s / DDR4 RDIMM, SSD SATA P3700 400G, on Red Hat Enterprise Linux* 7.0 kernel 3.10.0-123 using Java 8u25, PostgreSQL 9.4. Data Source: Request Number: 1951
Altibase HDB*: In-memory DBMS queries per second workload. Testing by Intel, 2/16/2016BASELINE: 1-Node, 2 x Intel® Xeon® Processor E5-2699 v3 on Grantley-EP (Wellsburg) with 128 GB Total Memory, 8 slots / 16 GB / 2133 MT/s / DDR4 RDIMM, 1TB HDD (SATA 7200rpm), HT and turbo on, on Red Hat Enterprise Linux* 6.4 kernel 2.6.32-358 using Intel Compiler 16 . Data Source: Request Number: 1904NEW: 1-Node, 2 x Intel® Xeon® Processor E5-2699 v4 on Grantley-EP (Wellsburg) with 128 GB Total Memory, 8 slots / 16 GB / 2400 MT/s / DDR4 RDIMM, 1TB HDD (SATA 7200rpm), HT and turbo on, on Red Hat Enterprise Linux* 7.1 kernel 3.10.0-229 using Intel Compiler 16 . Data Source: Request Number: 1904
IBM DB2: POPS queries workload. Testing by Intel, 2/11/2016BASELINE: 1-Node, 2 x Intel® Xeon® Processor E5-2699 v3 on Grantley-EP (Wellsburg) with 192 GB Total Memory, 24 slots / 8 GB / 2133 MT/s / DDR4 RDIMM / running at 1600MT/s, 2x 800GHz Intel S3700 SSD, 30R3 C-states off, turbo on, on SUSE Linux Enterprise Server* 11 SP3 using IBM DB2 10.5 FP7 running 36 queries. Data Source: Request Number: 1932NEW: 1-Node, 2 x Intel® Xeon® Processor E5-2699 v4 on Grantley-EP (Wellsburg) with 192 GB Total Memory, 24 slots / 8 GB / 2133 MT/s / DDR4 RDIMM / BIOS reports 1600MHz due to full memory population, 2x 800GHz Intel S3700 SSD, 271.RR00 C-States off, turbo on, on SUSE Linux Enterprise Server* 11 SP3 using IBM DB2 v10.5 FP7 running 36 queries. Data Source: Request Number: 1932
IBM* Informix Warehouse Accelerator (IWA): In-memory accelerator total time workload. Testing by Intel, 2/11/2016BASELINE: 1-Node, 2 x Intel® Xeon® Processor E5-2699 v3 on Grantley-EP (Wellsburg) with 192 GB Total Memory, 24 slots / 8 GB / 2133 MT/s / DDR4 RDIMM / BIOS reports 1600MHz due to full memory population, BIOS reports 1600MHz due to full memory population, 1x 800GB Intel S3700 SSD, HT and turbo on, on Red Hat Enterprise Linux* 6.5 kernel 2.6.32-431 using Informix 12.10. Data Source: Request Number: 1933NEW: 1-Node, 2 x Intel® Xeon® Processor E5-2699 v4 on Grantley-EP (Wellsburg) with 192 GB Total Memory, 24 slots / 8 GB / 2133 MT/s / DDR4 RDIMM / BIOS reports 1600MHz due to full memory population, BIOS reports 1600MHz due to full memory population, 1x 800GB Intel S3700 SSD, HT and turbo on, on Red Hat Enterprise Linux* 6.5 kernel 2.6.32-431 using Informix 12.10. Data Source: Request Number: 1933
32
* Other Names and Brands May Be Claimed as the Property of Others
Legal Notices and DisclaimersIntel technologies’ features and benefits depend on system configuration and may require enabled hardware, software or service activation. Learn more at intel.com, or from the OEM or retailer.
No computer system can be absolutely secure.
Tests document performance of components on a particular test, in specific systems. Differences in hardware, software, or configuration will affect actual performance. Consult other sources of information to evaluate performance as you consider your purchase. For more complete information about performance and benchmark results, visit http://www.intel.com/performance.
Cost reduction scenarios described are intended as examples of how a given Intel-based product, in the specified circumstances and configurations, may affect future costs and provide cost savings. Circumstances will vary. Intel does not guarantee any costs or cost reduction.
This document contains information on products, services and/or processes in development. All information provided here is subject to change without notice. Contact your Intel representative to obtain the latest forecast, schedule, specifications and roadmaps.
No license (express or implied, by estoppel or otherwise) to any intellectual property rights is granted by this document.
Statements in this document that refer to Intel’s plans and expectations for the quarter, the year, and the future, are forward-looking statements that involve a number of risks and uncertainties. A detailed discussion of the factors that could affect Intel’s results and plans is included in Intel’s SEC filings, including the annual report on Form 10-K.
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.
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 complete information visit http://www.intel.com/performance.
Intel, the Intel logo and others are trademarks of Intel Corporation in the U.S. and/or other countries. *Other names and brands may be claimed as the property of others.
© 2016 Intel Corporation.
33
Copyright © 2016, Intel Corporation. All rights reserved.