[gmx-users] multi threading on rocks cluster
mbx0009 at yahoo.com
Fri Mar 14 10:47:26 CET 2014
I have a problem with multithreading on a cluster.
Our (vanilla) rocks cluster has several nodes each with 8 cores. I used to
able to start a job with "mdrun_d -nt 8" (NOT mdrun_mpi) and the directive
"-pe mpi_smp 8" in the qsub job file, and the calculations would
be distributed over the eight cores in a single node.
Now judging from the speed of calculations, it seems as if only one core was
used, but in the log file there is a statement: "Using 8 MPI threads" ... is
way to make sure that all cores are used, and is there a way to check how
many cores are really used by a given job?
ps: I used a tpr file made with grompp version 4.6.5 and the mdrun on the
is actually 4.6.3 ... mdrun does not complain, but can this be the problem?
pps: i found that when using "cutoff-scheme = Verlet" in the mdp file the
run faster than otherwise, but even then they are slower than
ppps: i these runs i do not use PME, but cut-off for electrostatics, does
perhaps affect multi-threading?
View this message in context: http://gromacs.5086.x6.nabble.com/multi-threading-on-rocks-cluster-tp5015146.html
Sent from the GROMACS Users Forum mailing list archive at Nabble.com.
More information about the gromacs.org_gmx-users