[gmx-users] parallel run crashes only when shuffled

[David van der Spoel]

On Tue, 2002-10-01 at 15:43, Mikko Huhtala wrote:
> 
> I tried to use grompp with '-shuffle' to setup up a simulation for 4
> processors. The system is a protein dimer, one small-molecule ligand
> in each monomer, water and some ions. The suffled simulation crashes
> after 15 steps and some LINCS warnings with a fatal pressure scaling
> error (= system explodes). The exact same starting conformation runs
> on 4 processors just fine when it is set up without the suffle.
> 
> I know that shuffle probably does not improve things much for a system
> like this, so I am happily running the unshuffled configuration. I was
> simply wondering if I missed something obvious or is this a bug in
> Gromacs? I am running version gromacs-3.1.4-1 (rpm) with
> lam-6.5.4-usysv.1 (rpm) on Pentium 4 Xeons.
Could be a bug. You could improve your performance a bit by shuffling
manually:
protein 1
sol 10000
ions 999
protein 1
sol 10000

> 
> Mikko
> 
> 
> _______________________________________________
> gmx-users mailing list
> gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the list. Use the 
> www interface or send it to gmx-users-request at gromacs.org.
> 
-- 
Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, 	Biomedical center, Dept. of Biochemistry
Husargatan 3, Box 576,  	75123 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://zorn.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++