Re: AMBER: sander.MPI / parallel

From: David A. Case <case.scripps.edu>
Date: Mon, 16 Jun 2008 11:33:35 -0700

On Mon, Jun 16, 2008, Arturas Ziemys wrote:
>
> Yes - I tried whatever. According to openMPI, mpirun should get PBS
> environment and all assigned CPUs.

In every PBS environment I have used, you have to *both* tell PBS how many
nodes you want, and tell mpirun now many cpus you are using. On our openMPI
+ PBS system, you also have to tell mpirun about the hostfile that PBS
creates, so I write something like this:

   mpirun -np 8 -hostfile $PBS_NODEFILE /path/to/sander.MPI flags go here

Every PBS installation might be set up in a different fashion, but something
like that listed above is likely to be required.

...good luck...dac

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
      to majordomo.scripps.edu
Received on Wed Jun 18 2008 - 06:07:33 PDT
Custom Search