opteron VS Itanium 2

Jim Phillips jim at ks.uiuc.edu
Tue Oct 28 21:23:48 EST 2003


Hi,

The Athlon design has some Alpha blood in it, and in my experience they
both excel on branchy, unoptimized, float-intensive code.  The Opteron is
similar to the Athlon, but I wouldn't bother with 64-bit unless you're
actually going to use more than 2 GB of memory per node.  Athlon vs
Pentium 4 or Xeon is a closer match, and you really need to run some
benchmarks to decide between them.  If you have access to an Opteron you
should benchmark it as well, since I've heard they fly on some problems.

Itanium 2 (Madison) is the current NAMD speed champ (although it's tied
with a hyperthreaded P4 running multithreaded code), but it took some
serious work to get the inner loops to the point that the Intel compiler
could software pipeline them to get decent performance.  I've heard that
some Fortran codes had an easier time of it.  Big branches really hurt.

-Jim


On 28 Oct 2003, Butti Gabriele - Dottorati di Ricerca wrote:

> Dear all,
>       we are planning to build up a new cluster (16 nodes) before this
> year's end; we are evaluating different proposals from machine sellers,
> but the main doubt we have at this moment is whether choosing an Itanium
> 2 architecture or an AMD Opteron one.
>
> I know that ther's had already been on this list a debate on such a
> topic, but maybe some of you has some new experience to tell about.
>
> There is a wild bunch of benchmarks on these machines, but we fear that
> these are somewhat misleading and are not designed to test CPU's for
> intense scientific computing. The code we want to run on these machines
> is basically a home-made code, not fully optimized, which allocates
> around 500 Mb of RAM per node. Communication between nodes is a quite
> rare event and does not affect much computation time. In the past we had
> a very nice experience using Alpha CPU's which performed very well.
>
> To sum up, the question is: is the Itanium2 worth the price difference
> or is the Opteron the best choice?
>
> Thank you all
>
> Gabriele Butti
> --
>                                  \\|//
>                                 -(o o)-
>               /------------oOOOo--(_)--oOOOo-------------\
>               |                                          |
>               |             Gabriele Butti               |
>               |        -----------------------           |
>               |      Department of Material Science      |
>               |      University of Milano-Bicocca        |
>               |      Via Cozzi 53, 20125 Milano, ITALY   |
>               |      Tel (+39)02 64485214                |
>               |             .oooO   Oooo.                |
>               \--------------(   )---(   )---------------/
>                               \ (     ) /
>                                \_)   (_/
>
>
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org
> To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
>

_______________________________________________
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 mailing list