Hey Rob,
Perhaps something in the direction of srun --ntasks=2 --gres=gpu:4
nvidia-smi , help you?
this will run two tasks each with 4 gpu and execute nvidia-smi,
the output should be similar of doing nvidia-smi on one 8 gpu server
On 22/02/2018 01:26, Rob
Hello,
I'm relatively new to administering slurm, so my apologies if I've missed
something obvious.
We have nodes of 4 GPU and nodes of 8 GPU. I would like users to be able to
request a total number of GPUs they require. The MPI software is not fussed
how many nodes it spans.
I had hoped request
