Standard STREAM Submission for IBM Power 755

From: Peter Wong <wpeter@us.ibm.com>
Date: Fri Feb 05 2010 - 17:05:21 CST

John,

I would like to submit this STREAM standard result for IBM Power 755.

This system has four 3.3 GHz POWER7 processors with a total of 32 cores,
256 GB (32 x 8 GB) DDR3 1066 MHz, 32KBI+32KBD L1 on chip per core,
256KB L2 on chip per core, and 4MB L3 on chip per core.

This system also runs the following software:
- SUSE Linux Enterprise Server 11 for IBM POWER
- Libhugetlbfs 2.0
- IBM XL C/C++ for Linux, V10.1 updated with the Oct2009 PTF
- IBM XL Fortran for Linux, V12.1 updated with the Oct2009 PTF

Note: The measurement was done on IBM Power 750 Express (3.3 GHz) which is

      electronically equivalent to IBM Power 755 (3.3 GHz).

Computer Array size SMT Triad score
------------------------------ ------------- --- -----------
IBM Power 755 (3.3 GHz POWER7) 1,789,569,700 off 121360.3214

Please make this result publicly available at your earliest convenience
for highlighting IBM Power 755 to be announced on February 8, 2010.

Let me know if you have any questions.

Thanks,
Peter
--------------------------
Peter Wai Yee Wong, Ph.D.
IBM Linux Performance Team
wpeter@us.ibm.com

==========================================================================

-------------------------------------------------------------
STREAM version $Revision: 5.9 $
-------------------------------------------------------------
This system uses 8 bytes per DOUBLE PRECISION word.
-------------------------------------------------------------
Array size = 1789569700, Offset = 0
Total memory required = 40960.0 MB.
Each test is run 10 times, but only
the *best* time for each is used.
-------------------------------------------------------------
Number of Threads requested = 32
-------------------------------------------------------------
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
Printing one line per active thread....
-------------------------------------------------------------
Your clock granularity/precision appears to be 1 microseconds.
Each test below will take on the order of 187200 microseconds.
   (= 187200 clock ticks)
Increase the size of the arrays if this shows that
you are not getting at least 20 clock ticks per test.
-------------------------------------------------------------
WARNING -- The above is only a rough guideline.
For best results, please be sure you know the
precision of your system timer.
-------------------------------------------------------------
Function Rate (MB/s) Avg time Min time Max time
Copy: 105418.1004 0.2719 0.2716 0.2721
Scale: 104864.7057 0.2733 0.2730 0.2735
Add: 121341.1929 0.3542 0.3540 0.3543
Triad: 121360.3214 0.3543 0.3539 0.3545
-------------------------------------------------------------
Solution Validates
-------------------------------------------------------------

Received on Fri Feb 05 17:34:42 2010

This archive was generated by hypermail 2.1.8 : Mon Feb 08 2010 - 06:14:52 CST