CP2K performance in parallel run

[SS]

Dear CP2K users,

I was trying to run the AIMD simulation of my system containing 271 atoms(metal complex and water molecules). I am not getting a scale-up in performance in parallel runs as I go from 8 to 16 and then to 32 processors. It's taking almost the same time to complete the same number of MD steps.

I am using the following command

mpirun -np $np cp2k.popt  inp > out 

and I am running on an HPC cluster (CPU).

What could be the reason?

Expecting a reply

Thank you

[Patrick Gono]

Hello,

Could you please provide us with the input file, the output files for different runs, and the job submission script? Without this information it is impossible to know what the issue could be.

Yours sincerely,

Patrick Gono

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/0593753b-0d2b-468e-a2bc-ef7c3f1ad37b%40googlegroups.com.

[Yingchun Zhang]

check if you have the correct mpi, you should use the mpi which cp2k was compiled with