Academic Technology offers faculty and students access to a centrally managed High-Performance Computing (HPC) cluster designed to support large-scale workflows that benefit from enhanced computational power.

Cluster Specifications

Additional Features

Network: InfiniBand (HDR/Ethernet 200 Gb/s).

Operating System: Rocky Linux 8.9.

Software: Slurm, OpenMPI, GCC, and CUDA (multiple versions available via modules; available versions may change), Kraken, SPAdes, LAMMPS, Vasp, hdf5-2.1.1, MATLAB.

How to Get Started

• Accessing the HPC cluster
• Job submission using Slurm
• Getting Started with JupyterHub

Why Use HPC?

• Speed: Process data and run computations faster than traditional computing environments.
• Efficiency: Handle large datasets and complex simulations more effectively.
• Parallel workloads: Run workloads that can be divided into many tasks (e.g., MPI, large analyses, ML training).
• Resource-intensive applications: Support applications requiring significant compute, memory, or GPU resources (e.g., large-scale genome sequencing data).

Recommended Uses

• Scientific research: Bioinformatics, chemistry, astrophysics, and other research requiring simulation or large-scale analysis.
• Engineering: Design, modeling, and testing workflows.
• Large-scale data processing: Big data analytics, machine learning, and AI workloads.

If you are interested in accessing HPC for your research, submit a request via ServiceNOW or email at@sfsu.edu. This will usually lead to a brief introductory meeting with our systems team to discuss your research needs.

Note: All students require faculty sponsorship. Faculty must submit a request on the student's behalf for access to be granted.

Usage Policies

• Protected and sensitive data: The Polaris cluster does not support the storage or processing of HIPAA, FERPA, or PII protected data. Please do not store or process Level 1 data on the cluster, as Polaris is not designed to meet the security requirements associated with this type of data or applicable data-use requirements.
• Storage: Storage on the cluster is shared and limited. Users are expected to regularly review and remove data that is no longer needed and should not use the cluster for long-term or archival storage. If your project requires more than 1 TB of storage, please contact us to discuss your storage needs.
• Backups: Data stored on the cluster is not backed up. Academic Technology cannot guarantee that data will never be lost or corrupted. Users are responsible for maintaining backups of important research data and should regularly copy critical data to an appropriate backup or archival storage location.