Frontier, located at Oak Ridge National Laboratory in Oak Ridge, TN, is one of the world's most powerful supercomputers dedicated to open scientific research. It has lots of unique architectural features that we'll explore today, including:
- 64-core AMD “Optimized 3rd Gen EPYC” CPU (x9,408)
- AMD MI250X GPU (x37,632)
- Single-Node performance of 212TF (x 9,408)
- Node-local Memory / "Burst Buffer" of 3.84TB/Node
Here you will find the session agenda along with various resources to help get you up and running on Frontier. We'll take a deep look at the system's design, discuss how research teams write programs for and run workloads on Frontier, and you'll get to run on your own. The session is guided by hands-on challenges.
The first challenge is to successfully log into Frontier and clone this repository.
Each challenge has its own sub-directory under /challenges/, and includes a
README.md file and any additional files for that challenge. Challenges range in
difficulty from introductory to very complex, and may require research and
reference materials outside of what's provided in this repository.
Official Frontier User Guide & Documentation
SchedMD Slurm Documentation (scheduler)
