Aw: [gmx-users] Free Energy Simulations in Parallel

[Justin Lemkul]

On 7/27/13 7:02 AM, lloyd riggs wrote:
> Play with the domain decomposition, lincs itr/order, -ntomp and -ntmpi, etc...
> I was able to get a 4 day simulation which often gave that error to speed up to
> 12 hours on 24 CPU/3 nodes/144 cores but it took 2 days of submitting, checking
> speed, and killing jobs to try another grid routine.  My problem related to this
> is it wont go any faster (more nodes, cpu's or cores and starts to hit limits
> where there are only 100 atoms in a unit cell) if anyone knows a way to make it
> go faster.

In the context of free energy simulations (and the original post on this topic), 
the limiting factor is generally the use of couple_intramol.  If set to "no," as 
is fairly common, exclusions are generated that then limit the number of DD 
cells that can be used.  More detail on this specific topic can be found in the 
list archive.

-Justin

> Sincerely,
> Stephan Watkins
> *Gesendet:* Freitag, 26. Juli 2013 um 22:13 Uhr
> *Von:* "Quintin Sheridan" <qsherida at nd.edu>
> *An:* "gmx-users at gromacs.org" <gmx-users at gromacs.org>
> *Betreff:* [gmx-users] Free Energy Simulations in Parallel
> Dear Gromacs Users,
>
> Is it possible to run free energy calculations in parallel using mpirun?
> If not, what is the fastest way to run free energy calculations. I am
> trying to us the Bennet's Accepetance Ratio (g_bar) to get the free energy
> of solvation for an ionic liquid based on the tutorial by Justin Lemkul. I
> hav tried to decouple an ion pair as well as individual ions. In either
> case the simulations run locally but when I try to run them in parrallel I
> get the error:
>
> Fatal error:
> There is no domain decomposition for 8 nodes that is compatible with the
> given box and a minimum cell size of 2.26125 nm
>
> Thank You
> Quintin Sheridan
> --
> gmx-users mailing list gmx-users at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> * Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
>

-- 
==================================================

Justin A. Lemkul, Ph.D.
Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441

==================================================