[gmx-users] GPU metadynamics

[Jacopo Sgrignani]

Dear Albert
to run parallel jobs on multiple GPUs you should use something like this:

mpirun -np (number of parallel sessions on CPU) mdrun_mpi ...... -gpu_id 0000111122223333

so you will have 4 calculations for GPU.


Jacopo

Inviato da iPad

Il giorno 15/ago/2013, alle ore 10:56, Albert <mailmd2011 at gmail.com> ha scritto:

> Hello:
> 
> I've got two GTX690 GPU in a workstation, and I compiled gromacs-4.6.3 with plumed and MPI support. I am trying to run some metadynamics with mdrun with command:
> 
> mdrun_mpi -s md.tpr -v -g md.log -o md.trr -x md.xtc -plumed plumed2.dat -e md.edr
> 
> but mdrun can only use 1 GPU as indicated in the log file:
> 
> 
> 
> 4 GPUs detected on host node3:
>  #0: NVIDIA GeForce GTX 690, compute cap.: 3.0, ECC:  no, stat: compatible
>  #1: NVIDIA GeForce GTX 690, compute cap.: 3.0, ECC:  no, stat: compatible
>  #2: NVIDIA GeForce GTX 690, compute cap.: 3.0, ECC:  no, stat: compatible
>  #3: NVIDIA GeForce GTX 690, compute cap.: 3.0, ECC:  no, stat: compatible
> 
> 
> NOTE: potentially sub-optimal launch configuration, mdrun_mpi started with less
>      PP MPI process per node than GPUs available.
>      Each PP MPI process can use only one GPU, 1 GPUs per node will be used.
> 
> 1 GPU auto-selected for this run: #0
> 
> 
> 
> I am just wondering how can we use multiple GPU for such kind of jobs?
> 
> THX
> Albert
> -- 
> 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