[gmx-users] tpbconv ... -np option.

David spoel at xray.bmc.uu.se
Tue Jul 8 08:14:00 CEST 2003


On Tue, 2003-07-08 at 00:35, Senthil Kandasamy wrote:
> I have been running a simulation for about 10 ns on 4 nodes of a
> cluster. I am going to lose access to this cluster soon and need to
> continue this run on a different cluster. However, on the second
> cluster, scaling over 4 nodes is really bad. I would rather run it on a
> single processor or maybe two. To get a continuous trajectory and to run
> on a different number of processor, what really are my options?  
> 
> Since tpbconv cannot change the np option, i will have to go back to
> grompp and recreate a tpr file. However, from the new input gro file,
> the information I provide contains only coordinates and velocities. The
> Forces are lost, leading to a discontinuous trajectory (am I correct in
> stating this?)
No, forces are not important...
> 
> grompp -f samemdp.mdp -c f210ns.gro -p sametop.top -np (1 or 2)
> newtpr.tpr
> 
> I see a -t option for grompp. Does this recreate a continuous
> trajectory?
Almost. This has been discussed quite a bit on the list, but for the
first thing running on a different number of processors (or other type)
yields a different trajectory due to different order of summing forces
etc. Continuing (in the 3.1.4 code) also yields a small discontinuity
due to temperature coupling, this has been fixed in the cvs code.

Anyway, check the folling variablesin your mdp file:
init_t  = 10000 (or whatever)
gen_vel = no
unconstrained-start      = yes

then grompp with the -t option to get binary precision for the
coordinates and velocities.


> 
> I have about 8 simulations that need to be migrated and it would really
> improve my efficiency if I can recompile them for two processors or even
> one.
> 
> Thanks.
> 
> Senthil
> 
> 
> _______________________________________________
> 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, 	Dept. of Cell and Molecular Biology
Husargatan 3, Box 596,  	75124 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++



More information about the gromacs.org_gmx-users mailing list