High Performance Computing

Run scientific workloads that need more cores, memory or runtime than a personal computer or lab workstation can provide.

Use the HPC service when your analysis takes too long, needs more memory, must run many jobs, or should be executed close to research data.

Use this service when you need to:

Run analyses that take hours or days.

Process many samples, images, simulations or datasets.

Execute pipelines repeatedly or in parallel.

Use more CPU cores or memory than your workstation has.

Keep computation close to GPFS research storage.

Run reproducible workflows through Slurm.

Bioinformatics, genomics, imaging pipelines and batch data processing.

Physical, biological, molecular, population or engineering simulations.

Large R, Python, MATLAB, Stata or command-line analyses.

Many independent jobs, parameter sweeps and sample-based workflows.

Jobs that need stable execution beyond an interactive desktop session.

Purpose: General-purpose CPU workloads, pipelines, batch jobs and routine scientific computing.

Partition: std-cpu

Use when: Your workload fits standard CPU nodes and does not require very large memory or high core counts per job.

Purpose: Large CPU jobs, memory-intensive analysis and demanding parallel workloads.

Partition: high-cpu

Use when: Your workflow requires more cores, more memory or larger parallel execution than the standard CPU partition.

If you are not sure which partition to use, contact SCC before running large workloads.