Gromacs

MPI

    Table of contents
    No headers

    The acronym MPI stands for Message Passing Interface.  Making use of the MPI libraries allows calculations to be split across many processors, leading to increased performance.  Only mdrun is MPI-enabled; none of the other utilities can be executed as parallel jobs.  With the introduction of threading in GROMACS version 4.5, external MPI libraries are not necessary to invoke a parallel calculation on a laptop, workstation or single compute node.

    Page last modified 08:43, 21 Jan 2013 by hess