[Mitgcm-support] RE: Performance on a Compaq Alpha

[mitgcm-support at dev.mitgcm.org]

mike,

 i tried to log in, but got some silly ssh message about
"ssh_exchange_identification", so i'm none the wiser
as to what is going on! i did check another benchmark
that is usually a good indicator of numerical fluid
code performance - the mccalpin stream benchmark. like
specfp and mitgcm it has the p4 quite far ahead
of the alpha 666, so the mystery remains.....

chris
-----Original Message-----
From: spall at whoi.edu [mailto:spall at whoi.edu]On Behalf Of Mike Spall
Sent: Tuesday, March 05, 2002 1:03 PM
To: Chris Hill
Subject: Re: Performance on a Compaq Alpha


Hi Chris,

I am comparing 1 processor on each machine.  The other code I
benchmarked
was a simple shallow water, finite difference code that I wrote.  Karl
Helfrich (here at whoi) has also found that the alpha is appreciably
faster
than intel machines he has compared with, but unfortunately I don't know
exactly what he ran and on what machines.  My shallow water code runs on
the alpha in half the time that it takes on the Dell (using g77 on the
Dell).

You are welcome to do some testing on pontus (alpha) or metis (Dell) if
you like.

Mike
 
-- 
==================================================

Michael A. Spall
Department of Physical Oceanography
W.H.O.I. MS #21
360 Woods Hole Road
Woods Hole, MA   02543

mspall at whoi.edu
(508) 289-3342 (office)
(508) 457-2181 (fax)

=================================================