Thank you so much !
It worked for me !
we have to give nvidia-sim
Then we will know the jobID running on GPU, It also lists the GPU devices
with number from 0, 1 & 2
then we need to export as you said with device number which is free of job.
then launch the job with PMEMD.cuda
now of we give nvidia-sim , once can see volatile GPU utility as 99% for
the device 02.
This might be useful for others who are doing for the firsts time
thank you once again
On Sun, Sep 6, 2015 at 12:14 PM, Ross Walker <ross.rosswalker.co.uk> wrote:
> Hi Lara,
>
> Please take a read of http://ambermd.org/gpus/ which should hopefully
> tell you what you need to no. In short what you would do is:
>
> cd run1
> export CUDA_VISIBLE_DEVICES=0
> $AMBERHOME/bin/pmemd.cuda -O -i ... &
> cd ../run2
> export CUDA_VISIBLE_DEVICES=1
> $AMBERHOME/bin/pmemd.cuda -O -i ... &
> cd ../run3
> export CUDA_VISIBLE_DEVICES=2
> $AMBERHOME/bin/pmemd.cuda -O -i ... &
>
> All the best
> Ross
>
> > On Sep 5, 2015, at 4:32 PM, Lara rajam <lara.4884.gmail.com> wrote:
> >
> > Dear Amber
> >
> > I am running my jobs in multi gnu machine ( it has 3 GPU cards) .
> > If I want to use a single GPU for each job so that I can use 3 gpu for 3
> > different job
> > as a user how can i give that and how can i check that. Since I am not
> > aware of root password
> >
> > thank you
> > _______________________________________________
> > 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
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Sep 06 2015 - 10:00:03 PDT
