Database performance matters
Whether you are running transactional or analytical workloads, database performance is a key driver to overall system performance. IBM® DB2® is designed and developed in the labs to deliver industry-leading performance to IBM clients. The results speak for themselves.
Leading benchmark records
As of March 1, 2012, DB2 holds the following leading benchmark records for database performance.
- DB2 achieves the first clustered database result published for the SAP® Transaction Banking standard application benchmark[i]
- DB2 holds the #1 single system TPC-C result in throughput for transactional performance[ii]
- DB2 holds the #1 4-socket x86-64 TPC-C result in throughput for transactional performance[iii]
- DB2 holds the #1 SAP 3-tier SD (Sales & Distribution) result[iv], #1 SAP 2-tier SD single server result[v], and #1 SAP 4-processor result on Linux on 2-tier SD[vi] in SAP Standard Application Benchmarks
Over the years, DB2 has delivered many ground breaking performance results, well ahead of the competition.
Raw performance is critical to our clients and DB2 delivers the best performance numbers. However, the real value of our performance results lies in how they translate into lower operating costs for our clients. For instance, DB2 engineers strive to deliver the best per-core performance, which is important for clients in industries such as energy, mobile communications and financial services, who run database application workloads that use per-core pricing when determining total solution cost. Fewer cores for the same performance can translate into significant direct savings in licensing and ongoing annual software maintenance costs.
Energy Efficiency
Given the focus on lowering energy consumption, companies are keen to learn how they can reduce the carbon footprint of their IT. IBM calculations on the configuration used in the recently announced benchmark show it completed 53% more work per unit of power compared to power requirements published by Oracle[v] .
Longevity – the complete story
Industry benchmarks can be like a game of leapfrog, with vendors often passing one another. A more telling statistic is the overall amount of time a vendor is in the lead. The following diagrams were drawn from information on public Web sites and show the numbers of days of leadership in key industry benchmarks from 1 January 2003.
click to enlargeWhat Customers Are Saying About DB2 Performance
Benchmarks are a great way to compare database performance across vendors. However real world workloads can vary significantly in performance and high availability characteristics. DB2 is the clear leader in performance on production systems as well.
[i] This benchmark fully complies with the SAP Benchmark Council regulations and has been audited and certified by SAP AG (certification number 2011035). Details can be obtained from IBM and SAP.
[ii] A 64-core IBM Power 595 (5 GHz, 32 chips, 128 threads) with IBM DB2 9.5 is the overall single system (6,085,166 tpmC, $2.81/tpmC, configuration available 12/10/08).
[iii] System x 3850 X5 can grow to 8 socket 80 cores, 4 socket 40 core configuration used for this result. Best tpmC results listed for non-IBM 4socket systems as of 2011/07/26
DL585G7 Result: HP ProLiant DL585 G7: 1,193,472 tpmC at $0.69USD/tpmC, avail 2010/09/01, (4proc/48core/48thread)
DL580G7 Result: HP ProLiant DL580 G7: 1,807,347 tpmC at $0.49USD/tpmC, avail 2010/10/15, (4proc/32core/64thread)
x3850 X5 Result: IBM x3850 X5: 3,014,684 tpmC at $0.59USD/tpmC, avail 2011/09/22, (4proc/40core/80thread)
[iv] The 32-core IBM eServer p5 595 (1.9 GHz) achieved the best three-tier SAP SD Standard Application Benchmark result of 168,300 benchmark running IBM DB2 Universal Database v8.2.2, AIX 5L™ V5.3, SAP R/3® Enterprise Release 4.70 solution. SAP Certification number 2005021.
[v] IBM has the best two-tier SAP SD single server application benchmark result.
Configuration and results of the IBM Power System 795 on the two-tier SAP SD standard application benchmark running SAP enhancement package 4 for SAP ERP 6.0 (Unicode): 32 processors / 256 cores / 1024 threads, POWER7, 4.0 GHz, 4096 GB memory, 126,063 SAP SD benchmark users, dialog resp.: 0.98s, line items/hour: 13,772,670 Dialog steps/hour: 41,318,000, SAPS: 688,630, DB resp. time (dialog/ update):0.011s / 0.024s, CPU utilization: 96%, OS: AIX 7.1, DB2 9.7; cert# 2010046. For more details, see http://www.sap.com/benchmark (link resides outside ibm.com).
[vi] IBM achieved a world-record 4-processor result on Linux® on the two-tier SAP® Sales and Distribution (SD) standard application benchmark.
Configuration and results of the IBM System x®3850 X5, running IBM DB2® 9.7 and Red Hat® Enterprise Linux 6.2, and SAP enhancement package 4 for the SAP ERP application Release 6.0. The x3850 X5 achieved 12,560 SAP SD benchmark users with 0.99 seconds average dialog response time, 68,580 SAPS, measured throughput of 4,115,000 dialog steps per hour (or 1,371,670 fully processed line items per hour), and an average CPU utilization of 98% for the central server. The x3850 X5 was configured with four Intel® Xeon® E7-8870 processors at 2.40GHz with 30MB shared L3 cache per processor (4 processors/40 cores/80 threads), 512GB of memory, 64-bit DB2 9.7, Red Hat Enterprise Linux 6.2, and SAP enhancement package 4 for SAP ERP 6.0. The server accessed the DB2 9.7 database on an IBM System Storage® DS4800 disk system. For more details, see http://www.sap.com/benchmark (link resides outside ibm.com).
TPC-C is a trademark of the Transaction Performance Processing Council (TPPC). The TPPC web page is located at http://www.tpc.org (link resides outside ibm.com)
SAP and all other SAP product and service names mentioned herein are trademarks or registered trademarks of SAP AG in Germany and several other countries around the world. SAP benchmark information is located at http://www.sap.com/benchmark (link resides outside ibm.com)
SPECjEnterprise2010 is a trademark of the Standard Performance Evaluation Corporation (SPEC).
All benchmark claims are based on published information as of March 2012.