HPC Systems

ARCC offers several HPC systems for researchers at the University of Wyoming (Uwyo) and their collaborators. This core service is performed primarily over the on-campus HPC systems Medicinebow and GannettPeak. However ARCC maintains a number of specialty HPC resources for special use cases. ARCC maintained resources are listed below and links provided for more detailed information. Additionally ARCC has relationships with external HPC resource providors to enable UWyo researchers with support on these external HPC Resources including allocations on NCAR Wyoming Supercomputing Center (NWSC) and Argone National Lab Leadership Computing Facility (ALCF).
-
MedicineBow
MedicineBow is ARCC's current flagship HPC system. It is a heterogenous cluster with many different hardware configurations to meet a variety of research workflows. Including AMD and Intel CPUs and a variety of NVIDIA CPU nodes.
-
GannettPeak
GannettPeak (Artifical Intelligence for Wyoming) is ARCC's newest HPC system that is a purpose-built cluster for optimizing AI applications. Utilizing NVIDIA GraceHopper Superchip
Common HPC Information
While MedicineBow and GannettPeak are individual HPC systems, there are several features to using them that apply to both, or other systems that may be available to researchers. Listed below are these features that are documented in a generic sense.
-
Logging Into HPC
ARCC provides two different methods for logging into HPC systems. Including a web-based portal and Secure Shell (ssh) using ssh keys.
-
Using Software on HPC
General information on how to use research software on HPC systems. Including what is available as system-wide modules and methods for researchers to install thier own software.
-
Running HPC Jobs with Slurm
General information on how to schedule and run jobs on compute nodes using Slurm on HPC systems.
External HPC Systems
For information on how to get access and use external HPC systems, please see the External HPC Resources page on the ARCC website.
Learn More About HPC
Get familiar with HPC through ARCC's Intro to HPC