Issues with Simulation Speed

DL_POLY
1

Dear all,

I am trying to simulate a solvated system containing metal organic frameworks and organic analytes (such as dopamine and serotonin). The systems have around 30,000 atoms and the simulations are running incredibly slow (the speed I am getting is around 2 picoseconds/hour using a timestep of 0.001 picosecond/step). I believe the issue stems from the fact that DL_POLY 4 does not seem to be able to run parallel processes at my university's HPC. This is the shell script I have been using to run the simulations in the cluster:

2

Hi Bruno,

that indicates to me the code was not built correctly.

there is no openmp support in dlpoly... so there is only MPI.

in cmake needs to be build with cmake -DWITH_MPI=on (which is the default.)

also you can ask the sysadmin to build the last version

Regards,
Alin

Without Questions there are no Answers!