Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!uwm.edu!ux1.cso.uiuc.edu!moskva!banko
From: banko@moskva.ks.uiuc.edu (Brad Banko)
Newsgroups: comp.sys.alliant
Subject: Fortran benchmarks on an alliant fx2800 with 24 CEs?
Message-ID: <1991May15.155432.15315@ux1.cso.uiuc.edu>
Date: 15 May 91 15:54:32 GMT
Sender: usenet@ux1.cso.uiuc.edu (News)
Distribution: usa
Organization: University of Illinois at Urbana-Champaign
Lines: 19

I have recently been trying to run some benchmark calculations with
the Xplor2.1 molecular dynamics software (by Axel Brunger) on an Alliant fx2800 
with 24 CEs.
The sources are written in fortran, but are not optimized for the Alliant
at all.  The dynamics calculations involve 2-body force calculations
between all pairs of N atoms.  The fortran compiler is FX-Fortran v1.1.30,
a beta release.  Can anybody offer me some rough idea of how fast such
a machine might be relative to other machines (such as a Silicon Graphics 4D 
machine) on some well-defined benchmark calculations?

Thanks!

Brad Banko

-- 
==========================================================================
Brad Banko, NIH Center for Parallel Computation for Molecular Dynamics
Beckman Institute & Dept of Physics, University of Illinois--Urbana-Champaign
banko@lisboa.ks.uiuc.edu, (217)-244-1851, 328-4932