Seitenvergleich

...

Partition name	Node number	GPU hardware	Description
gpu-a100	42	4x A100 per Node	full node exclusive
gpu-a100:shared	FIXME	1 to 4 A100	shared node access, exclusive use of the requested GPUs
gpu-a100:shared:mig	1	1 to 28 1g.10gb A100 MIG slices	shared node access, shared GPU devices via Multi Instance GPU. Each of the four GPUs is logically split into usable seven slices with 10 GB of GPU memory associated to each slice	gpu-a100	4x A100 per Node	full node exclusive

# run command on two MPI nodes in the exclusive use partition
srun -N 2 -p gpu-a100 example_ cmd

gpu-a100:shared1 to 4 A100shared node access, exclusive use of the requested GPUs

# run command using GPUs (-G/--gpus) in the shared node partition
srun -G 2 -p gpu-a100:shared example_cmd

gpu-a100:shared:mig1 to 28 1g.10gb A100 MIG slicesshared node access, shared GPU devices via Multi Instance GPU.
Each of the four GPUs is logically split into usable seven slices with 10 GB of GPU memory associated to each slice

# run command using one Multi Instance GPU slice on the according partition
srun -G 1 -p gpu-a100:shared:mig example_cmd

Versionen im Vergleich

Alte Version 9

Neue Version 10

Schlüssel