On 7/29/15 4:25 PM, Elton Carvalho wrote:
On Thu, Jun 25, 2015 at 7:39 PM, Justin Lemkul <jalem...@vt.edu> wrote:
On 6/25/15 2:46 PM, John Degenstein wrote:
I suppose in retrospect that I could simply use a single processor which I
think would reduce the number of DD cells and thus prevent this error.
Or just use OpenMP parallelization. Any time you drastically change the
size of the system, this will be the case.
Is DD treated differently if you run MPI or OpenMP parallelization? If
so, how? I don't recall reading that in the docs and I always assumed
that DD worked distributing one domain per thread (be it OpenMP or
MPI).
There are as many DD cells as there are (thread-)MPI processes. These can be
further subdivided into multiple OpenMP threads. What I was talking about above
is the use of only OpenMP, e.g.
mdrun -ntmpi 1 -nt X
There is no use of DD in this case.
-Justin
--
==================================================
Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow
Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201
jalem...@outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul
==================================================
--
Gromacs Users mailing list
* Please search the archive at
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a
mail to gmx-users-requ...@gromacs.org.
