We're happy to announce the availability of new Nvidia V100 GPU in the gpu_ram partition of the cluster together with the cuda/10.2 (environmental software) module.
Each node in that partition now contains two GPU. Please note, that node cs-f04c01b10 holds the K40 which came with the first cluster installation.
Nodes cs-f04c01b[01-04,09] contain two V100 each, while
Node cs-f04c01b10 contains two Tesla K40m
Scientist may load environmental software module cuda/10.2 to use the latest features of this programming environment.
Please see the updated Cluster User Guide for information about requesting specific GPU resources in your job submissions: https://www.pik-potsdam.de/services/it/hpc/user-guides/slurm/
