User Tools

Site Tools



This shows you the differences between two versions of the page.

Link to this comparison view

Next revision
Previous revision
lsf_gpu [2013/11/19 12:20]
martin created
lsf_gpu [2018/05/08 08:36] (current)
meesters [GPU Queues] - fixed broken link
Line 1: Line 1:
-To use a GPU you have to explicitely reserv it in the bsub call.+====== ​GPU Queues ======
--R "​select[cuda=1],rusage[cuda=1]"+There are three different ​[[partitions|partitions (SLURM lingo for '​queues'​)]] inside the cluster that support gpu usage: The titan-Queues (''​titanshort/​long''​) currently include hosts i0001-i0009while the gpu-Queues (''​infogpu''​) include the hosts g0001-g0009 ((formally there have been the ''​gpushort/​long''​ queues on these nodes - access however is restricted.)). The titan-hosts carry 4 GeForce GTX TITAN, hence a usage request up to ''​cuda=4''​ can be selected (see below). In contrast the GeForce GTX 480 is installed on the gpu-hosts (for the ''​gpushort/​long''​ queues). Finally, for the tesla-Queues (''​teslashort/​long''​) 4 Tesla K20m cards are installed.
 +====== GPU Usage ======
 +To use a GPU you have to explicitly reserve it as a resource in the submission script:
 +<code bash>
 +# ... other SBATCH statements
 +#SBATCH --gres=gpu:<​number>​
 +#SBATCH -p <​appropriate partition>​
 +Number can be anything from 1-4 on our GPU nodes. In order to use more than 1 GPU the application needs to support using this much, of course.
 +===== Using multiple nodes and multiple GPUs =====
 +In order to use multiples nodes, you have to request more than one node.
lsf_gpu.1384860032.txt.gz · Last modified: 2013/11/19 12:20 by martin