Assistance Needed for Configuring SLURM to Utilize GPUs in VM for AlphaFold

Ask for help
,
1

I am currently trying to configure SLURM on a virtual machine to utilize its GPU resources for running AlphaFold. Despite specifying the correct parameters for GPU usage in my Nextflow process, and using --gpu=all in the Docker container setup, the job continues to run without utilizing the GPU.

Here are the details of my current setup:

  • SLURM Configuration:
    • slurm.conf and nodes.conf have been updated to recognize GPUs.
    • The partition configuration in partitions.conf is set up to allocate GPU resources.
  • Docker Configuration:
    • When creating the Docker container, we used the --gpu=all option to ensure the container has access to GPU resources.
  • Nextflow Process Configuration:
    • The SLURM clusterOptions for the process are set to --gres=gpu:1 to request GPU resources.
    • The accelerator parameter is set to 1 to indicate the use of GPU acceleration.
  • Observed Issue:
    • AlphaFold runs on the VM but does not appear to utilize the GPU, despite the process parameters and Docker configuration indicating GPU usage.
1 Like
2

Hi @GiampaoloMart , were you able to solve this yourself or do you still need help?

3

Hi @GeraldineVdA , thanks for reaching out! I was able to solve the problem myself, but I really appreciate your offer to help.

Thanks again!