What do you guys thing about the P4

Nordwall, Douglas J Nordwall at pnl.gov
Mon Apr 16 10:30:13 EDT 2001


Dave told me over the weekend that this code runs on his wifes machine. So it's
at least not tuned much ;)

Here's his exact phrasing:

My understanding was that Gaussian, Inc. compiled the code for
the lowest common denominator Intel platform.  My wife used to
run it on a 386 at her high school.  So, I don't know what
comiler was used, but whatever one it was they probably didn't
turn on any P4 hardware specific optimization.

The question I am not sure of is whether it was using this exact cd...



> -----Original Message-----
> From: Ivan Rossi [mailto:ivan at biocomp.unibo.it]
> Sent: Saturday, April 14, 2001 5:22 AM
> To: Nordwall, Douglas J
> Cc: beowulf at beowulf.org
> Subject: RE: What do you guys thing about the P4
> 
> 
> On Fri, 13 Apr 2001, Nordwall, Douglas J wrote:
> 
> > I spoke with Dave Feller and he does not know what compiler 
> was used. Since we
> > do not have a source license to Gaussian (our lab makes a 
> competing product ;),
> > we only get binaries. Since these were the same binaries I 
> have used on the last
> > 6 machines I have tested on, and were made _at least_ a 
> year ago, I am guessing
> > they were not made with special p4 compilers. However, Dave 
> and I can ask the
> > gaussian folks.
> 
> I would like to point out that Gaussian's stated requirement 
> for G98/linux is
> the usage of PGI Fortran compiler, release 3.1 or later. 
> Furthermore it is
> suggested the usage of the UTK BLAS libraries, tuned for the 
> Pentium Pro.
> See http://www.gaussian.com/g98_req.htm
> 
> 
> -- 
> Dr. Ivan Rossi - PNACoS Consulting & CIRB Biocomputing Unit
> PNACoS, Via della Grada 4/F, I-40122 Bologna, ITALY
> Phone (GSM): +39-0335-456177    Fax: +39-051-520257 (box 102)
> e-mail: ivan at biocomp.unibo.it   Web: http://www.biocomp.unibo.it/ivan
> 

_______________________________________________
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