
Tab navigation
- Overview- selected tab,
- Oracle / Sun
- HP
- Intel
Tab navigation
- On December 2, 2010, Oracle announced its SPARC SuperCluster TPC-C benchmark, claiming a world record. Oracle achieved its benchmark result by lumping together 27 systems -- a highly inefficient approach that is not practical for most clients. To achieve its result, Oracle used 27 servers, 1,728 cores and 13,824 threads.
- Oracle continues to run ads in the Wall Street Journal and elsewhere comparing a Sun SPARC computer to a previous generation IBM Power 595 benchmark result. Oracle claims the Sun computer has a seven-time faster response time, while IBM's "fastest" computer consumes six-times more energy.
- Oracle announced in October 2009 that Oracle and Sun SPARC SOLARIS achieved World Record TPC-C Performance beating IBM's best results on DB2 with Power 595 Server.
On December 2, 2010, Oracle announced its SPARC SuperCluster TPC-C benchmark, claiming a world record. Oracle achieved its benchmark result by lumping together 27 systems -- a highly inefficient approach that is not practical for most clients. To achieve its result, Oracle used 27 servers, 1,728 cores and 13,824 threads
IBM's latest TPC-C benchmark clustered system result announced in August 2010 shows better than three times the performance per processor core than the Oracle system TPC-C benchmark results with which it was compared1. And the clustered IBM TPC-C result on the IBM Power 780 (which used a total of only 24 processors with 192 cores) requires the energy, space, software licensing and systems management of just three servers, compared to Oracle's 27. In November 2010 IBM announced2 that a 256-core IBM Power 795 system with DB2 achieved the highest result ever published on the two-tier SAP Sales and Distribution (SD) standard application benchmark. The IBM configuration handled more than three times the number of SAP SD users than a 256-core Sun SPARC Enterprise M9000 (Oracle's largest system).
Oracle continues to run ads in the Wall Street Journal and elsewhere comparing a Sun SPARC computer to a previous generation IBM Power 595 benchmark result. Oracle claims the Sun computer has a seven-time faster response time, while IBM's "fastest" computer consumes six-times more energy.
A single Sun SPARC computer referenced is not capable of the performance claimed in the ad; it takes 12 nodes. The throughput of any single Sun computer referenced in the ad is less than the throughput of the IBM computer to which it is compared.3 The energy consumed by the servers in the 12-node Sun computer cluster exceeds the energy consumed by the single IBM computer to which it is being compared despite Oracle's claims that the IBM server uses six times more energy.z4 Response time referenced in the ad isn't an official TPC-C metric. IBM's Power 595 is actually over 4.7X the performance per core of the Sun system in the ad. IBM's POWER7® processor-based Power 780 has over 7.5X the performance per core of the Sun system.3 The IBM Power 595 benchmark results compared in Oracle's ad were posted in June, 2008. IBM published benchmark results on February 8, 2010, for the POWER7 announcement that show the Power 780 is faster than the Power 595.5 Still, Oracle's ad persists in claiming Power 595 is IBM's fastest computer. The fact is, there is no commercially available single computer from Oracle that provides faster throughput than IBM’s fastest Power Systems computer making the claim in Oracle's ad misleading.
Oracle announced in October 2009 that Oracle and Sun SPARC SOLARIS achieved World Record TPC-C Performance beating IBM's best results on DB2 with Power 595 server.
This benchmark from Oracle/Sun is the same TPC-C benchmark that Oracle promised to reveal on October 14, which led to a reprimand by the Transaction Processing Council (TPC) (link resides outside of ibm.com) for having made an unsubstantiated superior performance claim. In order to demonstrate "the world’s fastest database performance with a TPC-C benchmark" Oracle's used an Oracle RAC configuration (database for clustered servers) and a 12-Node Sun SPARC Enterprise T5440 cluster. It achieved 7,717,510 tpmC, with $2.34/per tpmC. Oracle compared this result to IBM’s single system 64-core IBM Power 595 with IBM DB2 9.5, based on a TPC benchmark submitted on June 10, 2008.
Here's a deeper look at Oracle's claim and the comparison to IBM's system:
- This is not a straightforward "apples to apples" comparison: IBM system is a single server while the Sun system is a configuration with a 12 node cluster.
- The Sun cluster had a total of 384 processor cores and 3,072 threads compared to 64 cores and 128 threads in the IBM system.
- With 512GB of memory per node the Sun cluster had a total of 6TB of memory, compared to 4TB in the IBM Power 595 system.
- While the Sun tpmC per core is 20,097; the IBM tpmC per core is 95,080. The IBM result had 4.7 times higher performance.
- The Sun system will not available until 12/14/09. The IBM system has been available since December 10, 2008.
- The flash technology used in the Sun result was integral to the performance and price/performance. Flash technology wasn't even being used in benchmarks when the TPC-C published the IBM system result on June 10, 2008. IBM has leadership results using SSDs in benchmarks including the recent #1 SPC-1 benchmark (link resides outside of ibm.com)
Additional information
Power Systems give you 3x per core performance versus Oracle/Sun
A new generation of systems that provide compelling new choices for companies of all sizes
1 IBM Power 780 cluster (3 x 64 core) (24 chips, 192 cores, 768 threads) with DB2 9.7 TPC-C result of 10,366,254 tpmC, $1.38/tpmC, configuration available 10/13/10) vs. Oracle's SPARC SuperCluster with T3-4 Servers (27 x 64 core) (108 chips, 1728 cores, 13824 threads) TPC-C result of 30,249,688 tpmC, $1.01/tpmC, configuration available 6/1/11). Source: www.tpc.org. Results current as of 12/02/10. TPC-C is a trademark of the Transaction Performance Processing Council (TPPC).
2 Configuration and results of the IBM Power 795 on the two-tier SAP SD standard application benchmark running SAP enhancement package 4 for the SAP ERP 6.0 application (Unicode): 32 processors / 256 cores / 1,024 threads, POWER7, 4.0 GHz, 4,096 GB memory, 126,063 SAP SD benchmark users, 0.98 seconds dialog response time, 13,772,670 line items per hour, 41,318,000 dialog steps per hour, 688,630 SAPS, 0.011 seconds/0.024 seconds database response time (dialog/ update), 96 percent CPU utilization running AIX® 7.1 and DB2® 9.7. Certification #2010046. Results valid as of 11/15/2010. Configuration and results of the Sun SPARC Enterprise Server M9000 on the two-tier SAP SD standard application benchmark running SAP ERP 6.0: 64 processors / 256 cores / 512 threads, 1,024 GB memory, 39,100 SAP SD benchmark users, SPARC64 VII, 2.52 GHz, Solaris 10, Oracle 10g, Certification # 2008042. http://www.sap.com/benchmark.
3 64-core IBM Power 595 (5 GHz, 32 chips, 128 threads) with IBM DB2 9.5 is the best overall single system (6,085,166 tpmC, $2.81/tpmC, configuration available 12/10/08) vs. 12-node 384-core Sun SPARC Enterprise T5440 Cluster (1.6 GHz UltraSPARC T2 Plus, 4 processor) with Oracle 11g Enterprise Edition with Real Application Clusters and Partitioning, 7,646,486.7 tpmC, $2.36/tpmC, available 3/19/10 vs. 8-core IBM Power 780 (4.14 GHz, 2 chips, 32 threads) with IBM DB2 9.5 is the best 8-core system (1,200,011 tpmC, $.69/tpmC, configuration available 10/13/10). Source: www.tpc.org (link resides outside of ibm.com). Results current as of 5/27/10. TPC-C is a trademark of the Transaction Performance Processing Council (TPC).
4 http://www.ibm.com/systems/power/migratetoibm/footnotes/efficiency.html
5 http://www.ibm.com/systems/power/hardware/reports/system_perf.html
