[Beowulf] computer performance growth

Eugen Leitl eugen at leitl.org
Wed Oct 12 08:58:01 PDT 2005


I'm trying to get a mental picture of application-level
performance (say, CFD, MD, QM or similiar application classes)
growth in recent years.

http://www.ce.chalmers.se/research/group/hpcag/publ/2004/EWN04/performancegrowth_tr-2004-9.pdf
takes a look at SPEC CINT, which makes it IMO way too optimistic
for memory bottlenecked codes, especially those with
an unpredictable access pattern.

What are your experiences in the trenches? This is single-node,
of course, as parallel scaling is an issue of its own.

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a>
______________________________________________________________
ICBM: 48.07100, 11.36820            http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://www.scyld.com/pipermail/beowulf/attachments/20051012/70e8cc30/attachment.bin


More information about the Beowulf mailing list