[gmx-users] Excessive memory usage of mdrun

[mpignata at bidmc.harvard.edu]

Good morning,


I'm trying to simulate a protein system on an HPC cluster called Orchestra, that use gromacs 5.0.4.

The system size is 39126 atoms and is made by the protein, 12160 water molecules, one Na+ and one Cl-. Of this system I've already made a run of 50 ns on my workstation without any problems, but every time that I've tried to simulate it on the cluster my job get killed for excessive memory usage (more than 64 GB).

I've tried to use different numbers of CPU (from 4 to 12) and I've tried to use a single node all for myself, but every time I had the same problem.

I've even tried the lysozyme benchmark, both with my .mdp and with a .mdp from a gromacs tutorial, and the same problem occurred every time.

I've talked with one of the administrator and they don't have a clue about what's happening, also because they usually don't have gromacs user on it.

Have this kind of things happened to someone before? Any clue?


Thanks,


Marcello



________________________________

This message is intended for the use of the person(s) to whom it may be addressed. It may contain information that is privileged, confidential, or otherwise protected from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution, copying, or use of this information is prohibited. If you have received this message in error, please permanently delete it and immediately notify the sender. Thank you.