[gmx-users] Output of mpi gromacs
gmx3 at hotmail.com
Mon Jul 19 11:54:55 CEST 2004
>From: Nguyen Hoang Phuong <phuong at theochem.uni-frankfurt.de>
>Reply-To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>Subject: Re: [gmx-users] Output of mpi gromacs
>Date: Mon, 19 Jul 2004 11:41:39 +0200 (MEST)
>On Mon, 19 Jul 2004, Xavier Periole wrote:
> > Tandia, Adama wrote:
> > > Dear ALL:
> > >
> > > I have installed a MPI version of Gromacs 3.2.
> > > When I run :
> > > * grompp_d -f file.mdp -c file.gro -p file.top -o file.tpr -v*
> > > The output is file.tpr (for 1 node). Then I run:
> > > * dmpirun -np 4 mdrun_d -s file.tpr -o file.trr -x file.xtc -e
> > > file.edr -multi -v*
> > > And I will have 4 .trr, 4.xtc and 4.edr files.
> > > This does not look normal to me.
> > > Is there something that I did not do right? I want to have one for
> > > each output from mdrun_d.
> > > How can I do that? I run MD of huge systems and the output files are
> > > very big and are killing me.
> > >
>you should prepare 4 tpr files: x0.tpr,...x3.tpr
>and then run
>mdrun_d -s x.tpr -o file.trr -x file.xtc -e file.edr -multi
>you will get 4 files like file0.trr,...file3.trr
>Note that the index of the x.tpr in mdrun is asigned atomatically by the
That is correct.
But I think Adama just wants to do a normal parallel run,
in which case the -multi option should not be used.
-multi is only useful for ensemble averaging NMR restraints
and some other special things.
Talk with your online friends with MSN Messenger http://messenger.msn.nl/
More information about the gromacs.org_gmx-users