[Beowulf] Register article on Opteron - disagree
landman at scalableinformatics.com
Mon Nov 22 00:57:35 EST 2004
Opteron is a newcomer to the stage. A stable OS for it has been
available for a relatively short period of time (SuSE, and very late
model RHEL, though it does not shine until you get a 2.6 kernel on
there, so RHEL is out of serious contention until RHEL 4). SuSE with a
reasonable kernel has only been available for a few months.
EM64t uses the same northbridge/southbridge design as the Xeon. This
means that multiprocessors are going to be not very good in SMP
performance, just like the Xeon's (and I am sure some folks will
disagree, but I am thinking of memory bandwidth hungry apps). The
chipset is the bottleneck.
Compare that to the Opteron. SC'04 was chock full of folks showing
off 4 and 8 way Opteron's. There were other things there not on the
show floor that were *very* interesting.
I am not sure what impact Nocona will have. I reserve judgement until
I get to see one/play with one, and see its future path. Lots of
interesting conversations were had at SC. The quote from Mark Twain may
be apt. If a Nocona follow-on follows the design of the Opteron (IOMMU
and HT, among other things), then things could get quite interesting.
john.hearns at clustervision.com wrote:
>I spotted this article on The Register.
>As someone who has installed Opteron clusters, including what I believe was the
>first in the UK (used for computational chemistry, and going strong),
>I disagree with this article.
>Opteron are alive and kicking.
>What's more the article confuses 'supercomputer' with '(super)computer in
>the Top 500'
>and also makes no mention of EMT64/Nocona. The artcle says Xeon/Itanium,
>then goes on to talk about 32 bit Xeon.
>The fact that there are fewer Opteron based systems in the Top 500 is
>irrefutable (I didn't know this) but it makes me uneasy to extrapolate
>this to the impending death of a CPU.
>I DO agree (and let's have some debate here) that Nocona is bound to make
>Beowulf mailing list, Beowulf at beowulf.org
>To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
Joseph Landman, Ph.D
Scalable Informatics LLC,
email: landman at scalableinformatics.com
web : http://www.scalableinformatics.com
Beowulf mailing list, Beowulf at beowulf.org
To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
More information about the Beowulf