Full course description
Term: Spring 2026
Date: March 9th, 2026
Time: 2:00 p.m. to 4:00 p.m.
Location: Online Only
Instructor: Matthew Brown
Presented By: Advanced Research Computing (ARC)
Description:
Prerequisites:
- A very basic understanding of computer hardware components like computing units (e.g., CPUs, cores) and memory, and of concurrency (e.g., threading) will be useful.
- An account on ARC clusters and an assignment to a project (so that you can do computations).
- Some basic familiarity with ARC clusters would be useful: clusters, compute nodes, head nodes, the Slurm job scheduler.
Description:
Graphics Processing Units (GPUs) are accelerators that are attached/integrated into computers (or compute nodes on clusters). They are finding ever greater use in research computing, as they continue to evolve. In this workshop, we provide an overview of GPU architectures and conceptual views of how they operate. Some current GPU advancements, like unified memory will be highlighted. This will also serve as a contrast to conventional multi-core compute nodes. Finally, illustrative uses of GPUs on ARC clusters will be demonstrated. Whether you write GPU code using programming languages or use (commercial) software that exploits GPUs, this workshop will give you greater insight into their purposes and use.