[gmx-users] multi option

[Nathan Moore]

Its a shame that the velocity generation is done in grompp (although I
understand how it might make the mdrun integrator easier to write and also
more efficient).  The minimum number of nodes one allocates on Blue Gene
is 1024 (half of those are often used for communication).  1024 serial
jobs * (100MB/tpr + 100MB/trr) =~ 200GB, This is a big amount of data
right? (3% of the total scratch space at my local supercomputing
center...)

Thanks for your comments,

NT


> On Tue, 2005-07-26 at 11:46 -0500, Nathan Moore wrote:
>> I'd like to use the Multi option to submit a bunch of serial jobs
>> simulataneously (at least 512 at once).  512 tpr files takes up a lot of
>> space if the system consists of even a moderate number of atoms.  Is
>> there
>> a way to submit just one tpr file and have np copies of it run with
>> multi?
>> (each needs to have an independant velocity initialization and rendom
>> seed
>> though..)
> This is not possible because velocity generation is done in grompp.
> I hope a machine like the Blue Gene comes with sufficient storage? Your
> trajectories will be a lot bigger anyway.
>
> --
> David.
> ________________________________________________________________________
> David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
> Dept. of Cell and Molecular Biology, Uppsala University.
> Husargatan 3, Box 596,          75124 Uppsala, Sweden
> phone:  46 18 471 4205          fax: 46 18 511 755
> spoel at xray.bmc.uu.se    spoel at gromacs.org   http://xray.bmc.uu.se/~spoel
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
> _______________________________________________
> gmx-users mailing list
> gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
>