[gmx-users] GROMACS performance :installation

[Szilárd Páll]

What do you want to do? For single node runs, except multi-runs, you
do not need MPI.
--
Szilárd


On Thu, Aug 18, 2016 at 6:03 PM, Nikhil Maroli <scinikhil at gmail.com> wrote:
> Dear all,
> We have following  workstations
>
> 2 x 12 core processor + 3 x GTX 980 Ti
> 2 x 12 core processor + 3 x GTX 1070
> 2 x 8 core processor + 2 xGTX 1070 and Tesla K20
>
> and a Desktop system with Single  i7 processor (8 core) + GTX 1070
>
> Previously I installed with flag DGMX_GPU=ON
> -DCUDA_TOOLKIT_ROOT_DIR=/usr/local/cuda    (8 core System with tesla k20)
>
> i used to run single job;
>  gmx mdrun -nt X -deffnm XX -gpu_id 0
>
> can I proceed with the same installation set for the above systems? or
> should I compile with CUDA+MPI? or any other options?
>
> I planned to run at least two Gromacs jobs in one workstation ,Which will
> give the best performance.?
>
>
>
>
> --
> Regards,
> Nikhil Maroli
> --
> 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-request at gromacs.org.