Re: your mail

From: Brian Bennion (brian_at_youkai.llnl.gov)
Date: Sun Feb 22 2004 - 00:44:16 CST

Hi Min,

>From my conversations with Jim Phillips the optimal number of atoms per
cpu is around 600. How many cpus per node? 4cpus sharing a 200 atom
system is overkill...

Brian

On Sun, 22 Feb 2004 yu275197_at_yorku.ca wrote:

>
> Hi
> I'm running a simulations of a small peptide (~200atoms) and i've noticed that
> it actually runs slower on two nodes compared to one. I know that it takes the
> simulation some time to distribute the load and speed up but my simulation
> seems to run slower the more nodes i use.
> I was wondering if there are ways to improve scaling.
>
> Cheers
> min
>

*****************************************************************
**Brian Bennion, Ph.D. **
**Computational and Systems Biology Division **
**Biology and Biotechnology Research Program **
**Lawrence Livermore National Laboratory **
**P.O. Box 808, L-448 bennion1_at_llnl.gov **
**7000 East Avenue phone: (925) 422-5722 **
**Livermore, CA 94550 fax: (925) 424-6605 **
*****************************************************************

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:38:26 CST