[gmx-users] MPICH on Gbit with > 4 CPUs?

[Pim Schravendijk]

I had the same problem, and never found out what exactly goes wrong 
here. I'm not sure if it's really an MPICH related problem, currently it 
works very well here with mpich on an Infiniband-connected cluster. Maybe 
the combination of MPICH, gigabit hardware and gromacs doesn't work for 
some strange reason.

FYI The working mpich version on the infiniband cluster here:

MPICH Version:          1.2.6
MPICH Release date:     $Date: 2004/08/04 11:10:38$
MPICH Patches applied:  none
MPICH Device:           vapi


> From: Carsten Kutzner <ckutzne at gwdg.de>
> Subject: [gmx-users] MPICH on Gbit with > 4 CPUs?
> To: gmx-users at gromacs.org
> Content-Type: TEXT/PLAIN; charset=US-ASCII
> 
> Hi,
> 
> has anybody ever managed to run gromacs in parallel on more than 4 CPUs
> when using MPICH on Gigabit Ethernet? I have just searched the mailing
> lists, where this problem was also reported (but no solution!).
> 
> We have dual-CPU Xeons, Linux 2.4. With up to 4 CPUs, gromacs 3.3 and
> 3.2.1 work. With 5+ CPUs, mdrun initializes properly but never finishes the first
> time step. An MPI test program (that does an all-to-all communication)
> works happily on up to 40 CPUs with MPICH. With LAM, mdrun works fine
> on any number of CPUs as well. Why not with MPICH?
> 
>    Carsten

--
Pim Schravendijk - PhD Student
Max Planck Institute for Polymer Research
http://www.mpip-mainz.mpg.de/~schraven/