Running two jobs at once doesn't make the GPU any more efficient, it just
makes the performance of both of them awful from all the context switching.
Only a process that horrendously underutilized the GPU would benefit from
this. A person promoting this approach is a person who did a lousy job with
the GPU part of their code.
In summary, what Ross said...
On Mon, Jun 6, 2011 at 3:24 AM, Marek Maly <marek.maly.ujep.cz> wrote:
> Dear all,
> some time ago I asked here question, as a part of some discussion, about
> the possibility to overload GPU card (here TESLA 2050) with calculations
> of two
> simultaneous jobs. If I remember Ross answered that it should be OK
> although,
> performance will decrease significantly comparing to just one job
> calculation.
>
> Today I tried to put 2 jobs on one Tesla 2050 and when I wanted to launch
> the second job
> there it was crashed with this error message:
>
>
> -----------------------------------------------------------------------------------------------
> cudaMemcpyToSymbol: SetSim copy to cSim failed all CUDA-capable devices
> are busy or unavailable
>
> -----------------------------------------------------------------------------------------------
>
> From the memory point of view it should be OK (enough for both jobs).
>
> I would be grateful for any useful suggestions !
>
> Thanks in advance,
>
> Best wishes,
>
> Marek
>
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Jun 06 2011 - 10:00:03 PDT
