Joe Landman wrote:
We are working on a machine with 2 Opteron 2218s. For laughs, I ran
streams on it. Here are 1,2,3,4 processor data.
1 processor:
Copy: 5713.9944 0.0591 0.0560 0.0600
Scale: 5713.9822 0.0587 0.0560 0.0600
Add: 5454.2389 0.0911 0.0880 0.0920
Triad: 5454.1576 0.0916 0.0880 0.0920
Mmmm ... these actually seem low considering the bus bandwidth is
supposed to be 10.67 GB/sec
in the socket F. Using the 75% rule (1 silent read, 2 actual
reads, and 1 write) from the triad shouldn't
we get:
10.67 * .75 = ~8 GBs/sec
2 processors:
Function Rate (MB/s) Avg time Min time Max time
Copy: 5713.8788 0.0591 0.0560 0.0640
Scale: 5713.9822 0.0596 0.0560 0.0640
Add: 5713.9538 0.0858 0.0840 0.0880
Triad: 5713.9538 0.0871 0.0840 0.0880
3 processors:
Copy: 8888.4000 0.0391 0.0360 0.0400
Scale: 8888.4147 0.0391 0.0360 0.0400
Add: 8570.9733 0.0573 0.0560 0.0600
Triad: 8570.9551 0.0587 0.0560 0.0600
4 processors:
Copy: 11428.1834 0.0293 0.0280 0.0320
Scale: 11428.1347 0.0298 0.0280 0.0320
Add: 10908.5812 0.0449 0.0440 0.0480
Triad: 10908.6107 0.0440 0.0440 0.0440
Mmm ... similarly ... this seems to be only 50% of the full bandwidth
of 2 socket F buses?
I did get one outlier at 4 CPUs. Could be a timing glitch.
4 processors:
Copy: 13332.7765 0.0302 0.0240 0.0360
Scale: 13332.7765 0.0262 0.0240 0.0320
Add: 13332.6220 0.0400 0.0360 0.0480
Triad: 14999.0756 0.0391 0.0320 0.0480
These are the numbers I would expect here ... about 70% of 2 *
10.67 GBs/sec.
I think that your out-"liar" is actually giving you the correct
numbers.
This is a dual socket 1207 motherboard. DDR2/533 ECC Registered
memory (will look back and check to be sure).
Initial benchmarks (2.6 GHz clock) put it at about 17% faster than
Opteron 275 and Woodcrest 5150 (2.66 GHz) on a GAMESS test we use
(1h41m for Woodcrest and 275, and 1h26m for this unit).
AMDs premium socket F is the 2220 SE which runs at 2.8 GHz. The
new AMD 4 digit naming convention
is discussed in the presentation I posted earlier.
rbw
--
Richard B. Walsh
"The world is given to me only once, not one existing and one
perceived. The subject and object are but one."
Erwin Schroedinger
Project Manager
Network Computing Services, Inc.
Army High Performance Computing Research Center (AHPCRC)
[EMAIL PROTECTED] | 612.337.3467
-----------------------------------------------------------------------
This message (including any attachments) may contain proprietary or
privileged information, the use and disclosure of which is legally
restricted. If you have received this message in error please notify
the sender by reply message, do not otherwise distribute it, and delete
this message, with all of its contents, from your files.
-----------------------------------------------------------------------
_______________________________________________
Beowulf mailing list, [email protected]
To change your subscription (digest mode or unsubscribe) visit
http://www.beowulf.org/mailman/listinfo/beowulf
