High-performance computing (HPC) means using many computers together to solve problems too large or too slow for a single workstation. An HPC cluster links many servers (nodes) with a fast network and shared storage, so one job can use many CPU cores, GPUs or large amounts of memory at once, and many jobs can run side by side.
How a cluster is used
- You log in to a head node over SSH and prepare your files and scripts.
- You describe what your job needs (cores, memory, GPUs, time) and submit it to a scheduler, such as Slurm.
- The scheduler runs the job on compute nodes when the resources become available and writes the results to shared storage.
When HPC helps
- Work that can be split into many independent pieces, such as running the same analysis on hundreds of samples.
- Programs that use many cores or GPUs at once, such as simulations, genome assembly or model training.
- Jobs that need more memory or storage than a laptop or workstation has.
HPC at UCR
- HPCC: the campus HPC cluster run by the High-Performance Computing Center. See HPCC, the HPCC website at hpcc.ucr.edu, and Connecting to the HPC cluster for a first job.
- Ursa Major: UCR's Google Cloud program, a separate service run by Research Computing. It covers cloud resources, including hardware not available on the HPCC. See Ursa Major and Ursa Major service tiers.
- National resources such as NSF ACCESS and the NAIRR Pilot.
For an overview of all options, see Computing resources.
Owner: Research Computing
Reviewed: 4 Oct 2026