Re: [AMBER] sander.MPI

From: Fabian Glaser <fglaser.technion.ac.il>
Date: Thu, 29 Nov 2012 15:54:06 +0200

>
> #PBS -l select=1:ppn=12

This does not work....

Anyway the following

#PBS -l select=1:ncpus=12:mpiprocs=12

and

mpirun -hostfile $PBS_NODEFILE pmemd.MPI -O -i prod.in -p protein.prmtop -c protein_equil.rst -o protein_prod.out -x protein_prod.mdcrd -r protein_prod.rst

Seems to be working VERY FAST, I'll see how it goes.

Thanks to everybody for your great help,

Fabian


_______________________________
Fabian Glaser, PhD
Bioinformatics Knowledge Unit,
The Lorry I. Lokey Interdisciplinary
Center for Life Sciences and Engineering

Technion - Israel Institute of Technology
Haifa 32000, ISRAEL
fglaser.technion.ac.il
Tel: +972 4 8293701
Fax: +972 4 8225153

On Nov 29, 2012, at 3:36 PM, Jason Swails wrote:

>
>
> On Thu, Nov 29, 2012 at 8:18 AM, Fabian Glaser <fglaser.technion.ac.il> wrote:
> Thanks Jason, much more clear now...
>
> So the following command is fine?
>
> mpirun -np 12 -hostfile $PBS_NODEFILE pmemd.MPI -O .... etc.
>
> The -np 12 is redundant if you are using $PBS_NODEFILE. In fact, if you omit the "-np 12", then you can simply change the line
>
> #PBS -l select=1:ppn=12
>
> line when you want to change the number of processors used (instead of also having to change the mpirun command as well).
>
> Good luck,
> Jason
>
> --
> Jason M. Swails
> Quantum Theory Project,
> University of Florida
> Ph.D. Candidate
> 352-392-4032


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Nov 29 2012 - 06:00:04 PST
Custom Search