right, it is not attribute of mxm, but general effect. and you are right again - performance engineering will always be needed for best performance in some cases.
OMPI, mxm trying to address out of the box performance for any workload, but OS tuning, hw tuning, OMPI or mxm tuning may be needed as well. (there is a reason that any MPI have hundreds of knobs) On Thu, Oct 1, 2015 at 1:50 PM, Dave Love <d.l...@liverpool.ac.uk> wrote: > Mike Dubman <mi...@dev.mellanox.co.il> writes: > > > we did not get to the bottom for "why". > > Tried different mpi packages (mvapich,intel mpi) and the observation hold > > true. > > Does that mean it's a general effect, unrelated to mxm, or that it is > related? > > > it could be many factors affected by huge heap size (cpu cache misses? > > swapness?). > > I'm sure we're grateful for any information, and I don't mean to be > rude, but this could be frustrating to people told they should do > performance engineering and trying to understand what might be going on. > [Was "heap" a typo?] > > > On Wed, Sep 30, 2015 at 1:12 PM, Dave Love <d.l...@liverpool.ac.uk> > wrote: > > > >> Mike Dubman <mi...@dev.mellanox.co.il> writes: > >> > >> > Hello Grigory, > >> > > >> > We observed ~10% performance degradation with heap size set to > unlimited > >> > for CFD applications. > >> > >> OK, but why? It would help to understand what the mechanism is, and why > >> MXM specifically tells you to set the stack to the default, which may > >> well be wrong for the application. > > _______________________________________________ > users mailing list > us...@open-mpi.org > Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users > Link to this post: > http://www.open-mpi.org/community/lists/users/2015/10/27759.php > -- Kind Regards, M.
