This description is meant to serve as a template that can be plugged into grant sections calling for descriptions of the computational resources available at Montana State University.

For questions, email: tempest-support@montana.edu

Research Cyberinfrastructure High Performance Computing

MSU maintains a priority access research computing cluster called Tempest, which has 128 nodes with 12264 logical CPU cores, 55926 gigabytes of RAM, 26 Nvidia A40 GPUs, 14 Nvidia A100 GPUs, 12 Nvidia H100 GPUs, and 600 terabytes of flash storage. Tempest is partially funded by and available via a buy-in model with free tier in which research groups receive access commensurate to their level of contribution. In 2021, MSU installed a large scale, scalable, high availability, backed up data storage system called Blackmore. This system was expanded in 2022 to a total capacity of 1.7 petabytes with throughput of up to 80 gigabits per second. In addition to Blackmore, RCI also provides support for SharePoint cloud storage, OURRstore archival storage, data publication repositories (in partnership with the MSU library), and controlled research storage (in partnership with UIT Security).

Dr. XXX has purchased access to both the large scale storage Blackmore and High Performance Computing cluster, Tempest. This grants me priority access to XX GPU nodes and XX CPU cores with XX TB of RAM and access-as-available to XX additional GPU nodes (XX A40 and XX A100), and an additional XXX CPU nodes, with a total of XX TB of memory. These resources are critical for XXX data analysis. Additionally, Dr. XXX has purchased XXX TB of high performance storage, which stores all XXX data and functions as a backup for all data generated in the XXX lab. These compute and storage resources are connected by a minimum 10 Gb/s network, and an institutional Globus subscription ensures efficient, secure, and uncorrupted data transfers between storage and compute nodes.