Re: TMD: different behaviour between serial and mpi versions

From: Jerome Henin (jhenin_at_cmm.chem.upenn.edu)
Date: Tue Jun 05 2007 - 11:27:40 CDT

Thomas,

Thank you for reporting this, it confirms what our tests indicated. We have
reported it to Jim, and indeed he seemed to have found some problem with the
code.

One thing I should check: in your simulations, do you use any other type of
scripted forces, restraints etc?

Bonne soirée,
Jerome

On Tuesday 05 June 2007 10:57, Thomas Gaillard wrote:
> Targeted Molecular Dynamics works well in serial versions of NAMD, with
> a nice linear decrease of the RMSD.
>
> With exactly the same parameters, the evolution of the RMSD is totally
> different with a mpi version of NAMD. The current RMSD is not able to
> follow the target RMSD and the simulation eventually stops with errors
> (Bad Global Exclusion Count).
>
> This observation has been made on different architectures (x86_64 and
> ibm aix), with different TMD force constants, and either in vacuo or
> with PBC.
>
> It seems that the TMD implementation with MPI is buggy.
>
> How is it possible that the evolution of the current RMSD is different
> between serial and mpi NAMD ?
>
> I am available for more informations or tests that you could suggest.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:44:47 CST