Re: [AMBER-Developers] [AMBER] Compiling amber parallel

From: David A Case <case.biomaps.rutgers.edu>
Date: Fri, 26 Oct 2012 14:17:49 -0400

On Fri, Oct 26, 2012, Jason Swails wrote:

> I'm guessing mdgx is using some MPI-2 features that Intel MPI does not have
> implemented. OpenMPI (at least as of version 1.5.x) also does not
> implement a large number of MPI-2 features (such as those used by the
> sander/terachem interface and pmemd.cuda.MPI). And apparently MDGX, now,
> as well.

I've never had problems with OpenMPI and mdgx. Works for me, in fact, with
openmpi-1.4.4, as well as with openmpi-1.5.x. So I'm not convinced that it
has much to do with intrinsic limitations in openMPI.

Regular readers of this list know my usual advice (compile MPI yourself using
the configure scripts we provide.)

...dac


_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Fri Oct 26 2012 - 11:30:02 PDT
Custom Search