lindahl at stanford.edu
Tue Jun 11 18:57:00 CEST 2002
Scott Delinger wrote:
> Forgive the newbie question, please:
> in the gmxbench-3.0.tar.gz file, there is a grompp.mdp file for DPPC,
> Poly-ch2, and Villin, but not LZM. Should there be a grompp.mdp file in
> LZM? I've run "grompp && mdrun" in the other three directories without a
> problem, but have lost my way with LZM. Any help is appreciated. I'm
> running these on Dual Athlons (on only one processor as I haven't gotten
> the MPI version to compile (I've installed both MPICH and LAM, but MPICH
> is my default)) with results similar to the published results.
There are two different mdp files for the Lysozyme benchmark; cutoff.mdp
uses cutoff and pme.mdp particle-mesh ewald electrostatics!
So, choose which setup you want and type grompp -f <name>.mdp. All the
other files should have the default names I think...
By the way - tell us if any results seem to deviate significantly from
the ones on the benchmark page; at least one users seems to have had
problems when he compiled his own version on SuSE 8.0 - it was 20%
slower than when he used our RPM package (no extra options, but compiled
with shared libraries on redhat-7.2).
More information about the gromacs.org_gmx-users