Hi,
First, the GIST command is not MPI-parallelized. Cpptraj should be
printing a message that looks like this:
Error: 'gist' action does not work with > 1 thread (64 threads currently).
The GIST action *is* OpenMP parallelized, but may not scale to 64
cores. I recommend running some benchmarks (using only 1-10 frames or
so) to see what kind of speedup you can achieve, e.g.
OMP_NUM_THREADS=1 cpptraj.OMP -i gist.in
OMP_NUM_THREADS=2 cpptraj.OMP -i gist.in
OMP_NUM_THREADS=4 cpptraj.OMP -i gist.in
and so on. Also keep in mind that the memory requirements for OpenMP
are a little larger - to prevent in-memory clashes, each thread has to
have certain data structured allocated separately from other threads.
Also note that if you're calculating water-water interactions (doeij
keyword) this is *very* memory hungry and may require a coarser grid.
Hope this helps,
-Dan
On Thu, Sep 20, 2018 at 5:57 AM M RCC <mkr3j2c1.gmail.com> wrote:
>
> Dear Amber Users,
>
> I am trying to run GIST calculation in parallel with 64 core AMD system
> using Ambertools18.
> when am using "mpirun -np 64 cpptraj.MPI -i gist.in" , i got error with
> core dump.
>
> then i am using "cpptraj.MPI -i gist.in" (even assigning
> OMP_NUM_THREADS=64) program is running in only one core.
>
> How i can utilise all 64 core for GIST calculation?
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Sep 20 2018 - 09:00:03 PDT
