Re: AMBER PC-linux/O2 SGI

From: Mark Forster <>
Date: Wed 02 Feb 2000 10:00:15 +0000

Dear Fdo

Please check out the SPECfp95, QM and MM benchmarks at the Daresbury site

They indicate that an 700 MHz AMD K7 Athlon syetem would have
far better floating point performance than an R5K based O2. The quantum
code used is GAMESS-UK while the MM code is DL-POLY. One
needs to bear in mind the usual caveats that benchmark rankings may
ultimately be somewhat different from rankings with a real world

It is of interest to compare the performance of the R12K based O2
and Octane systems from SGI. At the same clock rate (300MHz)
the Octane has a SPECfp of around 27 while that of the O2 is
less than half this value. One factor affecting this is probably
the cache size, which is 2MB on the Octane vs 1MB on the O2.

I would be interested to see any AMBER
benchmarks that you receive.


Fdo Danilo Gonzalez Nilo wrote:

> Hi all!
> I would like to know if anybody have some experience using amber
> in PC-linux and O2 SGI (benchmarks),
> What is the difference in performance between a PC-linux (700 MHz) and a
> O2 R5000 SGI, using AMBER?
> Is possible to use two CPU in parallel in a Dual Pentium (linux, 2x700
> MHz))?
> Thanks a lot!
> Fernando Danilo Gonzalez N.
> University of Santiago de Chile
> Faculty of Chemistry and Biology, Computational Chemistry Lab.
> Casilla 40, Correo 33, Santiago, Chile Fono: (562) 681 1542 Anexo:799
> E-mail : Fax : (562) 681 2108
> **************************************************************************

  Dr Mark J Forster Ph.D.
  Principal Scientist
  Informatics Laboratory
  National Institute for Biological Standards and Control
  Blanche Lane, South Mimms,
  Hertfordshire EN6 3QG, United Kingdom.
  Tel  +44 (0)1707 654753
  FAX  +44 (0)1707 646730
Received on Wed Feb 02 2000 - 02:00:15 PST
Custom Search