Sun Fire X2270 M2 Achieves Leading Single Node Results on ABAQUS Benchmark
By 11111 on Jun 29, 2010
- The Sun Fire X2270 M2 server performed up to 1.25 times faster than the previous generation Sun Fire X2270 server on the ABAQUS/Standard Server Benchmark test suite.
Comparisons below are against the top results
found at the ABAQUS website.
For the complete set of results, please go to the
ABAQUS/Standard Server Benchmark Test Suite (Single Platform)
Results are total elapsed run time in seconds
|Sun Fire X2270 M2||12||256||1033||730||1234|
|Sun Fire X2270 M2||8||313||1248||834||1340|
|Sun Fire X2270||8||319||1280||832||1360|
|HP BL460c G6||8||324||1309||843||1322|
|Sun Fire X2270 M2||4||544||1981||1230||1794|
|Sun Fire X2270||4||546||1983||1210||1771|
|HP BL460c G6||4||561||2062||1234||1812|
|Ratio 8-core X2270: 12-core X2270 M2||1.25||1.24||1.14||1.10|
Results and Configuration Summary
48 GB memory
4 x 24 GB striped SSDs
48 GB memory
2 x 24 GB internal striped SSDs
ABAQUS Standard Module V6.9
ABAQUS Standard Benchmark Test Suite - Server test cases
ABAQUS/Standard server benchmark problems provide an estimate of the
performance that can be expected when running the ABAQUS/Standard
module. The tests include analysis of a flywheel with centrifugal
loading (S2A), bolting a cylinder head onto an engine block (S4B &
S4D), and determining the footprint on an automobile tire (S6).
Key Points and Best Practices
The memory requirements for the test cases in the ABAQUS/Standard server benchmark test suite are substantial with some of the test cases requiring over 20 GB of memory. There are two memory limit parameters one can set to tune performance. One is to control when out-of-core memory when be used. You want to tune this to avoid excessive use of the out-of-core algorithms. The other control concerns I/O. This control will minimize the amount of unneeded disk activity. These limits may be tuned before a full run by doing a preliminary diagnostic mode run.
The test cases for the ABAQUS standard module all have a substantial I/O component where 15% to 25% of the total run times are associated with I/O activity (primarily scratch files). Performance will be enhanced by using a high performance file system.
The following are trademarks or registered trademarks of Dassault Systemes, or its subsidiaries in the United States and/or other countries: Simulia, ABAQUS, ABAQUS/Standard, ABAQUS/Explicit. All information on the ABAQUS website is Copyrighted 2004-2010 by Dassault Systemes. Results from http://www.simulia.com/support/v69/v69_performance.php as of June 28, 2010.