John,
I would like to submit STREAM benchmark results for
an IBM RS/6000 model 397. The entry in the table would look
something like this:
ncpus COPY SCALE ADD TRIAD
IBM_RS6000-397 1 778.8 777.5 883.1 882.4
The table entry labled
IBM-SP_P2SC-thin 1 690.6 684.6 787.2 786.8
should probably be changed to
IBM-SP_P2SC_120MHz-thin 1 690.6 684.6 787.2 786.8
since there is a 160 MHz P2SC thin SP node which is equivalent to the
RS/6000 397.
Thanks,
Frank Johnston
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
P.S. Here is the output from Stream on the IBM RS/6000 model 397.
The compiler used was xlf 5.1.0.0 with options -O3 -qarch=pwr2
----------------------------------------------
Double precision appears to have 16 digits of accuracy
Assuming 8 bytes per DOUBLE PRECISION word
----------------------------------------------
Array size = 2000000
Offset = 0
The total memory requirement is 45 MB
You are running each test 10 times
The *best* time for each test is used
----------------------------------------------------
Your clock granularity/precision appears to be 1 microseconds
The tests below will each take a time on the order
of 28828 microseconds
(= 28828 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) RMS time Min time Max time
Copy: 778.7826 .0415 .0411 .0419
Scale: 777.5306 .0414 .0412 .0420
Add: 883.0675 .0554 .0544 .0626
Triad: 882.3979 .0545 .0544 .0547
Sum of a is = 0.230660156259187354E+19
Sum of b is = 0.461320312485643840E+18
Sum of c is = 0.615093750014125568E+18
This archive was generated by hypermail 2b29 : Tue Apr 18 2000 - 05:23:07 CDT