How can I see what resources are on each node/partition?

View the HPC resources type table, this gives you detailed hardware specifications of each physical server node type:

• https://hpc.researchcomputing.ncl.ac.uk/calc/resources/

View the partition definitions, this shows what default values are available in each of the Slurm resource partitions.

View the Rocket partition definitions:

• HPC Resources & Partitions - Rocket

View the Comet partition definitions:

• HPC Resources & Partitions - Comet

Use sinfo to report on current partition resource availability. On Rocket you can see the list of CPU and GPU resources on the power partition as follows:

(Rocket) $ sinfo -o "%N %c %G" -p power
NODELIST CPUS     GRES
pn001    128      gpu:Tesla-V100-SXM2-16GB:4
(Rocket) $

Repeat as needed for each partition type you are interested in; the same commands work for Comet; e.g.:

(Comet) $ sinfo -o "%N %c %G" -p gpu-s_free
NODELIST     CPUS GRES
gpu[001-003] 256  gpu:L40:8
(Comet) $

Note that all our node types of the same category are identical - we don't have mixes of different GPU types within the same node type; all small GPU hosts have the same card, and all large GPU hosts have the same.

Back to FAQ index