Ask the administrators.
Nodes are named by their GPU resources.
frl1g[035-106]: each with 256 AMD cores, 512 GB RAM, 1 NVIDIA GEFORCE 3090.frl2g[004-034]: each with 256 AMD cores, 512 GB RAM, 2 NVIDIA GEFORCE 3090.frl8g[134-137]: each with 128 AMD cores, 512 GB RAM, 8 NVIDIA GEFORCE 3090.frl8a[138-141]: each with 128 AMD cores, 512 GB RAM, 8 NVIDIA TESLA A100.
We currently use Slurm.
To view your tasks, run:
squeue # See all.
squeue -u <USER> # See only a user.
To cancel a task, run:
scancel <JOBID>
To cancel all your tasks, run:
scancel -u <USER>