Skip to content

A place to host notes/guides on usage of Applied Math nodes on the Pinnacles Cluster

Notifications You must be signed in to change notification settings

Deep-Learning-at-UC-Merced/AppliedMathHPCNodes

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 

Repository files navigation

Applied Math HPC Nodes

UC Merced's Applied Mathematics (AM) hosts dedicated nodes on both MERCED and Pinnacles clusters, which can be utilized by AM faculty and graduate students (or those who have been granted permission).

Pinnacles Nodes

The AM department has a dedicated partition, dept.appliedmath, which hosts 6 nodes, with 2 being GPU nodes (one A100 GPU). The walltime limit on all nodes is 7 days.

Submitting Jobs to dept.appliedmath Generic Nodes

To submit your jobs to the normal AM nodes (without others being able to join the queue), your Slurm bash script must include:

#SBATCH -p dept.appliedmath
#SBATCH --constraint=6330

Submitting Jobs to GPU Nodes on dept.appliedmath:

Similar to the generic AM nodes, your script must first specify the dept.appliedmath partition, followed by adding the constraint to include GPUs. Just as in submitting your GPU jobs to the other nodes, you must also specify which GPU(s) you are requsting. This can be done as the follwing:

#SBATCH -p dept.appliedmath
#SBATCH --constraint=a100
#SBATCH --gres=gpu:<ReplaceWithGPUNumber(s)>

MERCED Nodes

Coming soon.

About

A place to host notes/guides on usage of Applied Math nodes on the Pinnacles Cluster

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published