Oracle Database 11g and Oracle Solaris 11 Running on Oracle’s SPARC T4-4 Server Outperforms HP and IBM on Industry-Standard Benchmark
Redwood Shores, Calif. – November 30, 2011
Demonstrating the outstanding performance of its SPARC T4-4 servers and the benefits of hardware and software engineered to work together, Oracle today announced a best-in-class TPC-H benchmark result at the three TB scale factor(1).
With this latest addition, Oracle’s SPARC T4 servers (www.oracle.com/us/products/servers-storage/servers/sparc-enterprise/t-series/index.html) now claim 12 world records (www.oracle.com/benchmarks) set on various business-critical enterprise workloads and industry-standard benchmarks.
In this TPC-H benchmark, which evaluates the performance of various Business Intelligence/Data Warehousing and Decision Support Systems for several standard database sizes, referred to as the scale factor (SF), Oracle Database 11g (www.oracle.com/us/products/database/index.html) Release 2 and Oracle Solaris 11 (www.oracle.com/solaris) , the first cloud OS, running on a SPARC T4-4 (www.oracle.com/us/products/servers-storage/servers/sparc-enterprise/t-series/sparc-t4-4/overview/index.html) server with four SPARC T4 3 GHz processors, achieved a stellar result of 205,792 [email protected]
with a price of $4.10/[email protected]
The SPARC T4-4 server with Oracle Database 11g demonstrated over two times better performance per processor with 36 percent lower price per query (reported as $/[email protected]
) than IBM's most recent result by an eight processor Power 780 server running Sybase. Additionally, Oracle’s solution offers better long-term value to its customers, as the TPC-H price/performance metric includes maintenance costs over a three-year period(2).
Compared to the best HP result at the three TB SF, Oracle’s solution delivers 2.5 times faster performance per processor and is capable of loading the entire dataset in half the time of an eight processor HP DL980 server running Microsoft SQL Server(3).
Oracle’s Sun Storage 2540-M2 arrays provided a well-balanced I/O configuration that allowed Oracle Database to achieve the I/O throughput of 17 GB/sec and execute refresh functions up to 3.4 times faster than IBM and up to 3.2 times faster than HP(1,2,3).
With this result, SPARC-based systems running Oracle Database 11g now hold TPC-H performance records at the 3TB SF for four processor systems, as well as an overall single system world record (1,4).
Oracle’s SPARC T4-4 server (www.oracle.com/us/products/servers-storage/servers/sparc-enterprise/t-series/sparc-t4-4-ds-486944.pdf) , running Oracle Solaris 11 is a versatile, high performance, highly scalable, two or four-socket server based on the SPARC T4 processor. Optimized for a wide range of enterprise workloads, it delivers unsurpassed single and multi-thread throughput performance and boasts an unmatched combination of speed, security, and availability in a compact 5U design.
"This outstanding benchmark result showcases the superior performance of Oracle’s SPARC T4 servers," said Juan Loaiza, senior vice president, Systems Technology, Oracle. "Delivering leading benchmark results across multiple configurations and scale factors is further proof that Oracle continues to provide customers with the best foundation and long-term value for their data warehousing solutions.”
Oracle Database (www.oracle.com/database/index.html)
Oracle Solaris (www.oracle.com/solaris)
SPARC T4 Servers (www.oracle.com/us/products/servers-storage/servers/sparc-enterprise/t-series/index.html)
Oracle engineers hardware and software to work together in the cloud and in your data center. For more information about Oracle (NASDAQ:ORCL), visit www.oracle.com (www.oracle.com/) .
Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of their respective owners.
Source: Transaction Processing Performance Council (TPC) www.tpc.org as of November 30, 2011.
(3) HP ProLiant DL980 G7 server (8 sockets/64 cores/128 threads) 162,602 [email protected]
, US$2.68/[email protected]
, 20,325 [email protected]
per processor, available 10/13/10. Database load time: 8h 35 m 17s. Refresh function 1: 173.0s. Refresh function 2: 126.3s.