Sun server scores high on benchmark test

Sun's 10000 server scores a whopping 6569.1 QphH@1000GB on a TPC-H benchmark test.

Sun Microsystems, Inc. recently announced its performance of 6,569.1 QphH@1000GB on the industry-standard TPC-H benchmark. The benchmark was performed on a Sun Enterprise 10000 server, also known as Starfire, running Informix Extended Parallel Server 8.31.FD1, a data warehouse product from Informix, Inc.

Established by the Transaction Processing Council, the TPC-H benchmark is an industry-standard test designed to measure systems' capability to examine large volumes of data, execute queries and give answers to business questions.

Sun's Starfire server out dueled a 64-CPU Compaq AlphaServer GS140 configuration at nearly half the price and performance -- US$1,159/QphH versus Compaq's US$2,171/QphH.

The Compaq result showed a performance of 6,514.1 QphH@1000GB and had a system availability date of Nov. 1, 1999.

"This benchmark shows that the Starfire is the best single-system solution for the rough-and-tumble world of ad hoc databases typical of today's complex data center environments," said Steve Campbell, director of marketing for Enterprise System Products, Sun Microsystems. "The reliability of its Starfire outperforms its competitors, even clustered solutions, which have a difficult time with batch processing and load balancing since they only focus on local transactions."

The system, running Informix XPS software release 8.31.FD1 and Solaris[tm] 7 Operating Environment software, was configured for this benchmark with 64 UltraSPARC[tm] II 466 MHz CPUs, each with an 8 MB external cache, and 64 GB of memory combined with 32 Sun StorEdge[tm] A5200 and disk array subsystems and 2 D1000 disk arrays totaling 6.4 TB of storage. The configuration is scheduled to be available February 28, 2001.

Tests were conducted at Sun's Large Scale Computing Center in Beaverton, Oregon.

According to the Transaction Processing Council, the TPC-H benchmark evaluates a composite performance metric (QphH@size) that measures the performance of decision support systems by the execution of sets of queries against a standard database under controlled conditions.

Newsletters

You have been successfully signed up. To sign up for more newsletters or to manage your account, visit the Newsletter Subscription Center.
See All
See All