From owner-nwchem-users@emsl.pnl.gov Mon Sep 27 10:44:33 2004
Received: from odyssey.emsl.pnl.gov (localhost [127.0.0.1])
	by odyssey.emsl.pnl.gov (8.12.10/8.12.10) with ESMTP id i8RHiXgm028658
	for <nwchem-users-outgoing@odyssey.emsl.pnl.gov>; Mon, 27 Sep 2004 10:44:33 -0700 (PDT)
Received: (from majordom@localhost)
	by odyssey.emsl.pnl.gov (8.12.10/8.12.10/Submit) id i8RHiXcv028657
	for nwchem-users-outgoing; Mon, 27 Sep 2004 10:44:33 -0700 (PDT)
Date: Mon, 27 Sep 2004 18:43:31 +0100 (BST)
From: "B.D Allen" <B.D.Allen@newcastle.ac.uk>
Subject: NWCHEM on SGI Altix
To: nwchem-users@emsl.pnl.gov
Reply-to: b.d.allen@newcastle.ac.uk
Message-id: <32958.82.39.109.49.1096307011.squirrel@sws2.ncl.ac.uk>
MIME-version: 1.0
Content-type: text/plain; charset=iso-8859-1
Content-transfer-encoding: 8bit
User-Agent: SquirrelMail/1.4.2
X-Newcastle-MailScanner-Information: Please contact Postmaster@newcastle.ac.uk
 for more information
X-Newcastle-MailScanner: Found to be clean
X-MailScanner-From: n7027418@swsimap.ncl.ac.uk
Sender: owner-nwchem-users@emsl.pnl.gov
Precedence: bulk

Hi all,
  I'm using NWChem 4.6 on our Linux SGI Altix system. I built from source
using MPI and Intel efc. Although the executable seems to work, it seems
a bit slow to me. I've run the SCF benchmarks from the NWChem website,
and the runs take roughly twice as long as those graphed  for the Altix
system, however many processors ( I've tried up to 26 ). Are there any
tricks for building NWChem on Altix systems?

  The settings I used were:
  NWCHEM_TARGET LINUX64
  USEMPI "y"
  LIBMPI -lmpi

  make nwchem_config
  make FC=efc

  Thanks in advance for your help

Ben


Ben Allen
Molecular Photonics Laboratory
School of Natural Sciences
Bedson Building
University Of Newcastle Upon Tyne
Newcastle Upon Tyne
NE1 7RU

+44 (0)191 2226485