High Performance Computing Resources

NOTS

NOTS (Night Owls Time-Sharing Service) is a general-purpose Intel-based cluster with both conventional and GPU processors. NOTS is designed for both high performance computing (HPC) and high throughput computing (HTC), and has 159 nodes interconnected with high-speed Ethernet and either Omni-Path or InfiniBand low-latency networking. An expansion in 2024 provided a significant boost in capacity and capability for CPU and GPU compute power, with a total of 88 million CPU core hours and 750 thousand GPU hours annually.

Nodes
Millions CPU Hours/Yr
thousand GPU Hours/Yr

ARIES

Health Research

The ARIES cluster is a donation from AMD and intended for health research. It uses AMD GPUs providing 12 million CPU hours and 1.5 million GPU hours annually.

Millions CPU Hours/Yr
Millions GPU Hours/Yr

RAPID

RAPID (Rice Analysis Platform for Interactive Data) is a small GPU-only cluster with 80 Nvidia GPUs, and is set up for interactive compute jobs and container-based execution.

Nvidia GPUs
Interactive
Optimized

Shared High-Speed Storage

Several hundred terabytes of VAST-based high-performance scratch space is available across all HPC clusters, ensuring rapid data access for compute jobs.