Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
SlideShare a Scribd company logo

Understanding the IBM Power Systems Advantage

•Download as PPTX, PDF•
•
IBM Power Systems
IBM Power Systems

Better performance and cost effectiveness empower better results in the cognitive era. For more information, visit: http://www.ibm.com/systems/power/hardware/linux-lc.html

Related slideshows

Open Innovation with Power Systems
Open Innovation with Power Systems Open Innovation with Power Systems
Open Innovation with Power Systems
 
Ibm integrated analytics system
Ibm integrated analytics systemIbm integrated analytics system
Ibm integrated analytics system
 
Bootcamp 2017 - SQL Server on Linux
Bootcamp 2017 - SQL Server on LinuxBootcamp 2017 - SQL Server on Linux
Bootcamp 2017 - SQL Server on Linux
 
1 of 17
Better performance and cost effectiveness empower better results in the cognitive era © 2016 IBM Corporation 1
Introduction When compared with most x86 systems, IBM® Power Systems™ offer a variety of benefits, from better price performance, to better relative performance, to lower total cost of acquisition. Read this document for a closer look at these benefits, and we think you’ll see: no matter what it is you’re trying to do with your data, IBM Power Systems can put you in a position to do it better. © 2016 IBM Corporation 2
IBM Data Engine for Hadoop and Spark – Power Systems Edition With IBM Data Engine for Hadoop and Spark – Power Systems Edition, users can take advantage of workload scaling to minimize execution times and reduce batch windows. This helps lead to better price performance for a variety of Spark use cases: © 2016 IBM Corporation 3 • All results are based on IBM Internal Testing of 3 SparkBench benchmarks consisting of SQL RDD Relation, Logistic Regression, SVM, comparing IBM DE-HS with x86 cluster. • Price performance = relative performance per dollar spent. • IBM environment: 6 Data Nodes and 1 Management Node. Each node is IBM Power System S812LC 10 cores / 80 threads, POWER8; 2.92GHz, 256 GB memory, RedHat 7.2, Spark 1.5.1, OpenJDK 1.8 • x86 environment: 6 Data Nodes and 1 Management Node. Each node is x86 E5-2620V3 12 cores / 24 threads, E5-2620 V3; 2.4GHz, 256 GB memory, RedHat 7.1, Spark 1.5.1, OpenJDK 1.8 • Pricing is based on web prices of HP DL380 and list prices of IBM Power S812LC 1.7x better price performance for Spark SQL query 2.1x better price performance for Spark logistic regression- based machine learning 1.4x better price performance for Support Vector Machine (SVM)
Spark on IBM Power Systems In 10 SparkBench benchmarks, IBM Power Systems S812LC demonstrated optimized performance and price performance for Spark workloads: © 2016 IBM Corporation 4 • All results are based on IBM Internal Testing of 10 SparkBench benchmarks consisting of SQL RDD Relation, Twitter, Pageview Streaming, PageRank, Logistic Regression, SVD++, TriangleCount, SVM, MF, SQL Hive. • Price performance = relative performance per dollar spent. • IBM Power System S812LC environment: 10 cores / 80 threads, 1 X POWER8; 2.9GHz, 256 GB memory, Ubuntu 15.04, Spark 1.4, OpenJDK 1.8 • Intel Xeon HP DL380 environment: 24 cores / 48 threads, 2 X E5-2690 v3; 2.6GHz , 256 GB memory. Ubuntu 15.04, Spark 1.4, OpenJDK 1.8 • * Pricing is based on web prices for S812LC (http://www-03.ibm.com/systems/power/hardware/s812lc/buy.html) and HP DL380 (http://h71016.www7.hp.com/dstoreHPE/MiddleFrame.asp?page=config&ProductLineId=431&FamilyId=3852&BaseId=45441&oi=E9CED&BEID=19701) 1.9x better performance per system than Intel Xeon E5-2690 v2 2.3x better price performance than Intel Xeon E5-2690 v2
PostgreSQL on IBM POWER8® For PostgreSQL workloads, IBM POWER8 delivers high per-core performance, a better revenue-generating architecture, and better price performance: © 2016 IBM Corporation 5 • Results are based on IBM internal testing of single system running multiple virtual machines with pgbench select only work load and are current as of October 5, 2015. Performance figures are based on running a 300 scale factor. Individual results will vary depending on individual workloads, configurations and conditions. • Price performance = relative performance per dollar spent. • IBM Power System S822LC; 16 cores / 128 threads, POWER8; 3.6GHz, 256 GB memory, PostgreSQL 9.5 Alpha2, RHEL 7.1, PowerKVM • Competitive stack: HP Proliant DL380 Gen9; 36 cores / 72 threads; Intel E5-2699 v3; 2.3 GHz; 256 GB memory, PostgreSQL 9.5 Aplha2, RHEL 7.1, RHEV • Transactions per $ graph compares virtual machine at comparable per VM transactions per second using S822LC running 8 vcpu (1 core equiv.) and DL380 GEN9 ran 4 vcpu (2 core equiv.) VM configurations. S822LC produced 26, 781 average TPS per VM @ 20 VMs; DL380 produced 26,793 average TPS per VM @ 16 VMs. 76 more revenue- generating VMs per rack than HP DL830 2.7x better per-core performance than Intel Xeon E5-2699 v3 1.74x better price performance than HP DL830
EnterpriseDB on IBM POWER8 IBM POWER8 offers an ideal platform for running EnterpriseDB technology, with better price performance and higher per- core performance: © 2016 IBM Corporation 6 • Results are based on IBM internal testing of single system and OS image running with pgbench work load at scale factor of 1000 and are current as of March 16, 2016. Individual results will vary depending on individual workloads, configurations and conditions. OS and EDB support price is for 1 year duration and Hardware is 3 yr standard support. • Price performance = relative performance per dollar spent. • IBM Power System S822LC; 16 cores / 64 threads, POWER8; 3.3 GHz, 128 GB memory, EDB 9.4, RHEL 7.1 • Competitive stack: HP Proliant DL380 Gen9; 36 cores / 72 threads; Intel E5-2699 v3; 2.3 GHz; 128 GB memory, EDB 9.4, RHEL 7.1 • Both tested configurations represent the highest processor frequency for that specific processor 60% better price performance than HP DL380 1.8x higher per-core performance than HP DL380
IBM InfoSphere® Streams on IBM POWER8 IBM POWER8 can provide higher performance for IBM InfoSphere Streams, empowering organizations to gain insights faster, even while running the same software. In addition, IBM POWER8 also offers a lower total cost of acquisition: © 2016 IBM Corporation 7 • 1.39x better performance is based on IBM internal testing of the InfoSphere Streams’ LogAnalysis benchmark; current as of November 16, 2015. Performance improvement figures are cumulative of all 20 tests in the benchmark. Individual results will vary depending on individual workloads, configurations and conditions. • Same performance is based on IBM internal testing of the IBM Streams’ The NOW Factory benchmark where each job must sustain a minimum throughput of 5 million records per minute; current as of February 3, 2016. Individual results will vary depending on individual workloads, configurations and conditions. • IBM Power System S822LC; 20 cores / 80 threads, POWER8; 2.9 GHz; 64 GB memory, RHEL 7.2 Beta • Competitive stack: HP DL380; 24 cores / 48 threads; Intel E5-2690 v3; 2.6 GHz; 64 GB memory; RHEL 7.1 • HW pricing based on IBM web prices and published HP prices as of 11/16/2015 using the Simple Configurator at https://h22174.www2.hp.com/SimplifiedConfig/Index 26% lower hardware TCA for the same performance, compared with HP DL380 1.39x better performance per core than Intel E5-2690 17% lower solution TCA (HW+SW) for the same performance versus Intel Haswell E5-2690 v3
IBM Data Engine for Analytics With IBM Data Engine for Analytics, organizations can experience 1.5x better throughput per server than Intel: © 2016 IBM Corporation 8 • Based on a Hadoop-DS benchmark 6-stream run conducted on a POWER cluster of 9 S822L servers and an Intel cluster of 17 servers with E5-2680 v2 processors. • The POWER cluster used 16 logical DataNodes (12 cores, 120 GB RAM each) and 1 NameNode. Intel cluster used 16 physical DataNodes (20 cores, 128 GB RAM each) and 1 NameNode. • Both POWER and Intel cluster used the same software stack running IBM Open Platform with Apache Hadoop v4.1 and IBM BigInsights Analyst v4.1
IBM Power Systems with NVIDIA Tesla K80 GPUs By running IBM Power Systems integrated with NVIDIA Tesla K80 GPUs, organizations can achieve results even better than those achieved running Power Systems alone. With the accelerated performance NVIDIA GPUs provide, organizations are empowered to run fewer systems, which in turn can lead to reduced costs: © 2016 IBM Corporation 9 • Results are based on IBM internal testing of systems running NAMD version 2.10 APOA1, F1ATPASE, STMV code benchmarked on POWER8 systems installed each with 2 NVIDIA Tesla K80 GPUs.. Individual results will vary depending on individual workloads, configurations and conditions. • Price performance = relative performance per dollar spent. • IBM Power System S822LC; 16 cores / 128 threads, POWER8; 3.3GHz, 128 GB memory • IBM Power System S822LC; 16 cores / 128 threads, POWER8; 3.3GHz, 128 GB memory, 2 NVIDIA K80 GPUs 6.7x better performance than Power Systems without NVIDIA Tesla K80 GPUs 2.6x better price performance than Power Systems without NVIDIA Tesla K80 GPUs
IBM POWER8 with fully populated DIMMs IBM POWER8 with 32 DIMMs is able to provide significantly more memory bandwidth than Intel Haswell with 24 DIMMs, enabling faster memory access: © 2016 IBM Corporation 10 • IBM Power System S822LC results are based on IBM internal measurements of STREAM Triad; 20 cores / 20 of 160 threads active, POWER8; 3.5GHz, up to 1TB memory, • Intel Xeon data is based on published data running STREAM Triad; 24 cores / 24 of 48 threads active, E5-2390 v3; 2.3GHz up to 1.5 TB memory. For more details see http://www.intel.com/content/www/us/en/benchmarks/server/xeon-e5-2600-v3/xeon-e5-2600-v3-stream.html 2.2x more memory bandwidth than Intel Haswell E5-2600 v2 with 24 DIMMs
Little Endian DB2® with BLU Acceleration® on IBM Power Systems Taking advantage of IBM Power Systems together with Little Endian DB2 with BLU Acceleration offers a number of performance and cost benefits: © 2016 IBM Corporation 11 • 2.03X more query results is based on IBM internal testing of a sample analytic workload; current as of October 20, 2015. Performance improvement figures are cumulative of all queries in the workload. Individual results will vary depending on individual workloads, configurations and conditions. • IBM Power System S822LC; 20 cores / 80 threads, POWER8; 3.5GHz, 768 GB memory, DB2 10.5 / Ubuntu 14.04 • Competitive stack: HP DL380p; 36 cores / 72 threads; Intel E5-2699 v3; 2.3 GHz; 768 GB; DB2 10.5 / RHEL 7.2 • HW pricing based on IBM web prices and published HP prices as of 11/9/2015 using the Simple Configurator at https://h22174.www2.hp.com/SimplifiedConfig/Index 36% lower hardware TCA for equal performance, compared with Intel E5-2699 v3 2.03x more query results per hour per core than Intel Haswell E5-2699 v3 36% lower solution TCA (HW+SW) for equal performance versus Intel E5-2699 v3
MariaDB on IBM Power Systems IBM Power Systems provides high performance and reduced operating costs for MariaDB workloads: © 2016 IBM Corporation 12 • Results are based on IBM internal testing of single system running multiple virtual machines with Sysbench read only work load and are current as of October 18, 2015. Performance figures are based on running 24 M record scale factor per VM. Individual results will vary depending on individual workloads, configurations and conditions. • IBM Power System S822LC; 20 cores / 160 threads, POWER8; 2.9 GHz, 256 GB memory MariaDB 10.0.19. Ubuntu 14.04.03, PowerKVM 3.1 • Competitive stack: HP Proliant DL380 Gen9; 24 cores / 48 threads; Intel E5-2690 v3; 2.6 GHz; 128 GB memory, MariaDB 10.0.20. Ubuntu 14.04.03, KVM • Each system was configured to run at similar per VM throughput levels and number of VMs were increased for each system until total system throughput showed maximum throughput levels. Competitive pricing was taken from available web-based pricing. • For more information about MariaDB running on IBM Power Systems, go to: http://www-304.ibm.com/partnerworld/wps/servlet/ContentHandler/stg_com_sys-ibm-power-systems-and-mariadb 45% better price performance than Intel Xeon E5-2690 v3 69% more system- level throughput than Intel Xeon E5-2690 v3 66% more VMs than Intel Xeon E5-2690 v3 2x better performance per core than Intel Xeon E5-2690 v3
IBM POWER8 for the STAC benchmark In the STAC benchmark, IBM POWER8 set a new public record, making it the performance standard for organizations operating in the financial industry. The level of performance was also significantly higher than that offered by competing systems: © 2016 IBM Corporation 13 • https://stacresearch.com/news/2015/03/12/stac-report-stac-a2-ibm-power8-solution • All IBM Data is found in STAC Configuration Disclosure for this SUT: www.STACresearch.com/IBM150305 • Based on STAC-A2.β2.GREEKS.TIME.WARM. STAC Configuration Disclosure for this SUT: www.STACresearch.com/INTC140814 2.3x better performance than the best-performing 2-socket x86 solution 1.7x better performance than the best-performing solution with two x86 CPUs and one Xeon Phi co-processor
Entity Analytics on IBM POWER8 IBM POWER8 is able to deliver faster transactions for entity analytics. This empowers organizations to find entity relationships faster and react before the competition does: © 2016 IBM Corporation 14 • 2.05X faster result is based on IBM internal testing of a sample workload; current as of May 29, 2015. Performance improvement figures are based on multiple Entity Analytics processes running a 1 million record workload . Individual results will vary depending on individual workloads, configurations and conditions. • IBM Power System S824; 24 cores / 192 threads, POWER8; 3.5GHz, 256 GB memory • Competitive stack: HP DL380 Gen 9; 36 cores / 72 threads; Intel E5-2699 v3; 2.3 GHz; 768 GB memory 2.05x faster per core than Intel Haswell E5-2699 v3
Learn more With all the benefits named here and more, IBM Power Systems deliver the platforms you need to put data to work, make the most of the business insights available to you, and prepare yourself for success in the cognitive era. To find out more about what Power Systems can do for your organization, contact your IBM representative today, or visit ibm.com/systems/power. © 2016 IBM Corporation 15
© 2016 IBM Corporation 16
Trademarks and notes © Copyright IBM Corporation 2016 IBM Corporation IBM Systems Route 100 Somers, NY 10589 Produced in the United States of America April 2016 IBM, the IBM logo, ibm.com, Power Systems, POWER8, InfoSphere, DB2, and BLU Acceleration are trademarks of International Business Machines Corp., registered in many jurisdictions worldwide. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on the web at “Copyright and trademark information” at ibm.com/legal/copytrade.shtml. This document is current as of the initial date of publication and may be changed by IBM at any time. Not all offerings are available in every country in which IBM operates. The performance data discussed herein is presented as derived under specific operating conditions. Actual results may vary. THE INFORMATION IN THIS DOCUMENT IS PROVIDED “AS IS” WITHOUT ANY WARRANTY, EXPRESS OR IMPLIED, INCLUDING WITHOUT ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND ANY WARRANTY OR CONDITION OF NON- INFRINGEMENT. IBM products are warranted according to the terms and conditions of the agreements under which they are provided. © 2016 IBM Corporation 17

More Related Content

Understanding the IBM Power Systems Advantage

  • 1. Better performance and cost effectiveness empower better results in the cognitive era © 2016 IBM Corporation 1
  • 2. Introduction When compared with most x86 systems, IBM® Power Systems™ offer a variety of benefits, from better price performance, to better relative performance, to lower total cost of acquisition. Read this document for a closer look at these benefits, and we think you’ll see: no matter what it is you’re trying to do with your data, IBM Power Systems can put you in a position to do it better. © 2016 IBM Corporation 2
  • 3. IBM Data Engine for Hadoop and Spark – Power Systems Edition With IBM Data Engine for Hadoop and Spark – Power Systems Edition, users can take advantage of workload scaling to minimize execution times and reduce batch windows. This helps lead to better price performance for a variety of Spark use cases: © 2016 IBM Corporation 3 • All results are based on IBM Internal Testing of 3 SparkBench benchmarks consisting of SQL RDD Relation, Logistic Regression, SVM, comparing IBM DE-HS with x86 cluster. • Price performance = relative performance per dollar spent. • IBM environment: 6 Data Nodes and 1 Management Node. Each node is IBM Power System S812LC 10 cores / 80 threads, POWER8; 2.92GHz, 256 GB memory, RedHat 7.2, Spark 1.5.1, OpenJDK 1.8 • x86 environment: 6 Data Nodes and 1 Management Node. Each node is x86 E5-2620V3 12 cores / 24 threads, E5-2620 V3; 2.4GHz, 256 GB memory, RedHat 7.1, Spark 1.5.1, OpenJDK 1.8 • Pricing is based on web prices of HP DL380 and list prices of IBM Power S812LC 1.7x better price performance for Spark SQL query 2.1x better price performance for Spark logistic regression- based machine learning 1.4x better price performance for Support Vector Machine (SVM)
  • 4. Spark on IBM Power Systems In 10 SparkBench benchmarks, IBM Power Systems S812LC demonstrated optimized performance and price performance for Spark workloads: © 2016 IBM Corporation 4 • All results are based on IBM Internal Testing of 10 SparkBench benchmarks consisting of SQL RDD Relation, Twitter, Pageview Streaming, PageRank, Logistic Regression, SVD++, TriangleCount, SVM, MF, SQL Hive. • Price performance = relative performance per dollar spent. • IBM Power System S812LC environment: 10 cores / 80 threads, 1 X POWER8; 2.9GHz, 256 GB memory, Ubuntu 15.04, Spark 1.4, OpenJDK 1.8 • Intel Xeon HP DL380 environment: 24 cores / 48 threads, 2 X E5-2690 v3; 2.6GHz , 256 GB memory. Ubuntu 15.04, Spark 1.4, OpenJDK 1.8 • * Pricing is based on web prices for S812LC (http://www-03.ibm.com/systems/power/hardware/s812lc/buy.html) and HP DL380 (http://h71016.www7.hp.com/dstoreHPE/MiddleFrame.asp?page=config&ProductLineId=431&FamilyId=3852&BaseId=45441&oi=E9CED&BEID=19701) 1.9x better performance per system than Intel Xeon E5-2690 v2 2.3x better price performance than Intel Xeon E5-2690 v2
  • 5. PostgreSQL on IBM POWER8® For PostgreSQL workloads, IBM POWER8 delivers high per-core performance, a better revenue-generating architecture, and better price performance: © 2016 IBM Corporation 5 • Results are based on IBM internal testing of single system running multiple virtual machines with pgbench select only work load and are current as of October 5, 2015. Performance figures are based on running a 300 scale factor. Individual results will vary depending on individual workloads, configurations and conditions. • Price performance = relative performance per dollar spent. • IBM Power System S822LC; 16 cores / 128 threads, POWER8; 3.6GHz, 256 GB memory, PostgreSQL 9.5 Alpha2, RHEL 7.1, PowerKVM • Competitive stack: HP Proliant DL380 Gen9; 36 cores / 72 threads; Intel E5-2699 v3; 2.3 GHz; 256 GB memory, PostgreSQL 9.5 Aplha2, RHEL 7.1, RHEV • Transactions per $ graph compares virtual machine at comparable per VM transactions per second using S822LC running 8 vcpu (1 core equiv.) and DL380 GEN9 ran 4 vcpu (2 core equiv.) VM configurations. S822LC produced 26, 781 average TPS per VM @ 20 VMs; DL380 produced 26,793 average TPS per VM @ 16 VMs. 76 more revenue- generating VMs per rack than HP DL830 2.7x better per-core performance than Intel Xeon E5-2699 v3 1.74x better price performance than HP DL830
  • 6. EnterpriseDB on IBM POWER8 IBM POWER8 offers an ideal platform for running EnterpriseDB technology, with better price performance and higher per- core performance: © 2016 IBM Corporation 6 • Results are based on IBM internal testing of single system and OS image running with pgbench work load at scale factor of 1000 and are current as of March 16, 2016. Individual results will vary depending on individual workloads, configurations and conditions. OS and EDB support price is for 1 year duration and Hardware is 3 yr standard support. • Price performance = relative performance per dollar spent. • IBM Power System S822LC; 16 cores / 64 threads, POWER8; 3.3 GHz, 128 GB memory, EDB 9.4, RHEL 7.1 • Competitive stack: HP Proliant DL380 Gen9; 36 cores / 72 threads; Intel E5-2699 v3; 2.3 GHz; 128 GB memory, EDB 9.4, RHEL 7.1 • Both tested configurations represent the highest processor frequency for that specific processor 60% better price performance than HP DL380 1.8x higher per-core performance than HP DL380
  • 7. IBM InfoSphere® Streams on IBM POWER8 IBM POWER8 can provide higher performance for IBM InfoSphere Streams, empowering organizations to gain insights faster, even while running the same software. In addition, IBM POWER8 also offers a lower total cost of acquisition: © 2016 IBM Corporation 7 • 1.39x better performance is based on IBM internal testing of the InfoSphere Streams’ LogAnalysis benchmark; current as of November 16, 2015. Performance improvement figures are cumulative of all 20 tests in the benchmark. Individual results will vary depending on individual workloads, configurations and conditions. • Same performance is based on IBM internal testing of the IBM Streams’ The NOW Factory benchmark where each job must sustain a minimum throughput of 5 million records per minute; current as of February 3, 2016. Individual results will vary depending on individual workloads, configurations and conditions. • IBM Power System S822LC; 20 cores / 80 threads, POWER8; 2.9 GHz; 64 GB memory, RHEL 7.2 Beta • Competitive stack: HP DL380; 24 cores / 48 threads; Intel E5-2690 v3; 2.6 GHz; 64 GB memory; RHEL 7.1 • HW pricing based on IBM web prices and published HP prices as of 11/16/2015 using the Simple Configurator at https://h22174.www2.hp.com/SimplifiedConfig/Index 26% lower hardware TCA for the same performance, compared with HP DL380 1.39x better performance per core than Intel E5-2690 17% lower solution TCA (HW+SW) for the same performance versus Intel Haswell E5-2690 v3
  • 8. IBM Data Engine for Analytics With IBM Data Engine for Analytics, organizations can experience 1.5x better throughput per server than Intel: © 2016 IBM Corporation 8 • Based on a Hadoop-DS benchmark 6-stream run conducted on a POWER cluster of 9 S822L servers and an Intel cluster of 17 servers with E5-2680 v2 processors. • The POWER cluster used 16 logical DataNodes (12 cores, 120 GB RAM each) and 1 NameNode. Intel cluster used 16 physical DataNodes (20 cores, 128 GB RAM each) and 1 NameNode. • Both POWER and Intel cluster used the same software stack running IBM Open Platform with Apache Hadoop v4.1 and IBM BigInsights Analyst v4.1
  • 9. IBM Power Systems with NVIDIA Tesla K80 GPUs By running IBM Power Systems integrated with NVIDIA Tesla K80 GPUs, organizations can achieve results even better than those achieved running Power Systems alone. With the accelerated performance NVIDIA GPUs provide, organizations are empowered to run fewer systems, which in turn can lead to reduced costs: © 2016 IBM Corporation 9 • Results are based on IBM internal testing of systems running NAMD version 2.10 APOA1, F1ATPASE, STMV code benchmarked on POWER8 systems installed each with 2 NVIDIA Tesla K80 GPUs.. Individual results will vary depending on individual workloads, configurations and conditions. • Price performance = relative performance per dollar spent. • IBM Power System S822LC; 16 cores / 128 threads, POWER8; 3.3GHz, 128 GB memory • IBM Power System S822LC; 16 cores / 128 threads, POWER8; 3.3GHz, 128 GB memory, 2 NVIDIA K80 GPUs 6.7x better performance than Power Systems without NVIDIA Tesla K80 GPUs 2.6x better price performance than Power Systems without NVIDIA Tesla K80 GPUs
  • 10. IBM POWER8 with fully populated DIMMs IBM POWER8 with 32 DIMMs is able to provide significantly more memory bandwidth than Intel Haswell with 24 DIMMs, enabling faster memory access: © 2016 IBM Corporation 10 • IBM Power System S822LC results are based on IBM internal measurements of STREAM Triad; 20 cores / 20 of 160 threads active, POWER8; 3.5GHz, up to 1TB memory, • Intel Xeon data is based on published data running STREAM Triad; 24 cores / 24 of 48 threads active, E5-2390 v3; 2.3GHz up to 1.5 TB memory. For more details see http://www.intel.com/content/www/us/en/benchmarks/server/xeon-e5-2600-v3/xeon-e5-2600-v3-stream.html 2.2x more memory bandwidth than Intel Haswell E5-2600 v2 with 24 DIMMs
  • 11. Little Endian DB2® with BLU Acceleration® on IBM Power Systems Taking advantage of IBM Power Systems together with Little Endian DB2 with BLU Acceleration offers a number of performance and cost benefits: © 2016 IBM Corporation 11 • 2.03X more query results is based on IBM internal testing of a sample analytic workload; current as of October 20, 2015. Performance improvement figures are cumulative of all queries in the workload. Individual results will vary depending on individual workloads, configurations and conditions. • IBM Power System S822LC; 20 cores / 80 threads, POWER8; 3.5GHz, 768 GB memory, DB2 10.5 / Ubuntu 14.04 • Competitive stack: HP DL380p; 36 cores / 72 threads; Intel E5-2699 v3; 2.3 GHz; 768 GB; DB2 10.5 / RHEL 7.2 • HW pricing based on IBM web prices and published HP prices as of 11/9/2015 using the Simple Configurator at https://h22174.www2.hp.com/SimplifiedConfig/Index 36% lower hardware TCA for equal performance, compared with Intel E5-2699 v3 2.03x more query results per hour per core than Intel Haswell E5-2699 v3 36% lower solution TCA (HW+SW) for equal performance versus Intel E5-2699 v3
  • 12. MariaDB on IBM Power Systems IBM Power Systems provides high performance and reduced operating costs for MariaDB workloads: © 2016 IBM Corporation 12 • Results are based on IBM internal testing of single system running multiple virtual machines with Sysbench read only work load and are current as of October 18, 2015. Performance figures are based on running 24 M record scale factor per VM. Individual results will vary depending on individual workloads, configurations and conditions. • IBM Power System S822LC; 20 cores / 160 threads, POWER8; 2.9 GHz, 256 GB memory MariaDB 10.0.19. Ubuntu 14.04.03, PowerKVM 3.1 • Competitive stack: HP Proliant DL380 Gen9; 24 cores / 48 threads; Intel E5-2690 v3; 2.6 GHz; 128 GB memory, MariaDB 10.0.20. Ubuntu 14.04.03, KVM • Each system was configured to run at similar per VM throughput levels and number of VMs were increased for each system until total system throughput showed maximum throughput levels. Competitive pricing was taken from available web-based pricing. • For more information about MariaDB running on IBM Power Systems, go to: http://www-304.ibm.com/partnerworld/wps/servlet/ContentHandler/stg_com_sys-ibm-power-systems-and-mariadb 45% better price performance than Intel Xeon E5-2690 v3 69% more system- level throughput than Intel Xeon E5-2690 v3 66% more VMs than Intel Xeon E5-2690 v3 2x better performance per core than Intel Xeon E5-2690 v3
  • 13. IBM POWER8 for the STAC benchmark In the STAC benchmark, IBM POWER8 set a new public record, making it the performance standard for organizations operating in the financial industry. The level of performance was also significantly higher than that offered by competing systems: © 2016 IBM Corporation 13 • https://stacresearch.com/news/2015/03/12/stac-report-stac-a2-ibm-power8-solution • All IBM Data is found in STAC Configuration Disclosure for this SUT: www.STACresearch.com/IBM150305 • Based on STAC-A2.β2.GREEKS.TIME.WARM. STAC Configuration Disclosure for this SUT: www.STACresearch.com/INTC140814 2.3x better performance than the best-performing 2-socket x86 solution 1.7x better performance than the best-performing solution with two x86 CPUs and one Xeon Phi co-processor
  • 14. Entity Analytics on IBM POWER8 IBM POWER8 is able to deliver faster transactions for entity analytics. This empowers organizations to find entity relationships faster and react before the competition does: © 2016 IBM Corporation 14 • 2.05X faster result is based on IBM internal testing of a sample workload; current as of May 29, 2015. Performance improvement figures are based on multiple Entity Analytics processes running a 1 million record workload . Individual results will vary depending on individual workloads, configurations and conditions. • IBM Power System S824; 24 cores / 192 threads, POWER8; 3.5GHz, 256 GB memory • Competitive stack: HP DL380 Gen 9; 36 cores / 72 threads; Intel E5-2699 v3; 2.3 GHz; 768 GB memory 2.05x faster per core than Intel Haswell E5-2699 v3
  • 15. Learn more With all the benefits named here and more, IBM Power Systems deliver the platforms you need to put data to work, make the most of the business insights available to you, and prepare yourself for success in the cognitive era. To find out more about what Power Systems can do for your organization, contact your IBM representative today, or visit ibm.com/systems/power. © 2016 IBM Corporation 15
  • 16. © 2016 IBM Corporation 16
  • 17. Trademarks and notes © Copyright IBM Corporation 2016 IBM Corporation IBM Systems Route 100 Somers, NY 10589 Produced in the United States of America April 2016 IBM, the IBM logo, ibm.com, Power Systems, POWER8, InfoSphere, DB2, and BLU Acceleration are trademarks of International Business Machines Corp., registered in many jurisdictions worldwide. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on the web at “Copyright and trademark information” at ibm.com/legal/copytrade.shtml. This document is current as of the initial date of publication and may be changed by IBM at any time. Not all offerings are available in every country in which IBM operates. The performance data discussed herein is presented as derived under specific operating conditions. Actual results may vary. THE INFORMATION IN THIS DOCUMENT IS PROVIDED “AS IS” WITHOUT ANY WARRANTY, EXPRESS OR IMPLIED, INCLUDING WITHOUT ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND ANY WARRANTY OR CONDITION OF NON- INFRINGEMENT. IBM products are warranted according to the terms and conditions of the agreements under which they are provided. © 2016 IBM Corporation 17