Dear HPC community,
The High Performance Computing (HPC) team is pleased to announce Juno, UTD’s new flagship HPC system that will grow over time to support research and education mission of the university.
Juno consists of 77 compute nodes with 5,184 CPU cores, 32 terabytes of memory, and 12 GPUs. Juno is accompanied by IO2, the new high-speed storage system with 2 petabytes capacity to store data actively used in computations. Juno is a community or shared cluster. All of Juno’s resources are shared among all users and there are no “condo” or private nodes or queues. (Ganymede 2 will continue to serve as our condo model HPC cluster.)
| Juno | Ganymede | Ganymede 2 |
| 77 nodes 5,184 cores 32,256 GB RAM 12 Nvidia GPUs | 137 public nodes 2,192 cores 8,900 GB RAM No public GPUs | 122 nodes Cores, memory and GPUs vary by nodes. |
Juno considerably expands the HPC resources available to the UTD community to solve its most demanding computing problems. With Juno we can support larger more computational demanding problems and support higher degree of parallel processing performance.
Juno was delivered to UTD in the fall semester and the HPC staff have been working to install and configure the system for campus use. Juno will become available to the campus community according to the following timeline:
- January 2025 – User acceptance testing with select users
- February 2025 – Early access for existing active Ganymede users
- March 2025 – General access for UTD faculty, students and staff
We will communicate again with specific Juno availability information in February. Please reach out to hpc@utdallas.edu if you have any questions about Juno or HPC in general.
Khalid Warraich
Director HPC Facilitation
Rajesh Rao
Director HPC Operations and Cyberinfrastructure
Stefano Leonardi
Sr Director HPC
Professor Dept. of Mechanical Engineering