Tutorial IV of The “Scale-Up” HPC & AI Tutorial Series: The Shift to AI (GPUs)

This tutorial introduces how to shift from CPU-based workflows to GPU-accelerated computing for AI and deep learning. We will discuss the core ideas behind moving code from CPU to GPU, including how to configure and select GPU resources on the cluster and what needs to change in your code or framework to take advantage of these new resources. You will learn how to set up deep learning environments using tools (Conda and Jupyter Notebook), so you can develop and run GPU-enabled models in a reproducible way. We will also cover practical techniques for monitoring GPU and CUDA memory usage with tools such as nvidia-smi and simple Python snippets, to help you understand and troubleshoot performance and out-of-memory issues.

For the hands-on portion, you will train a basic deep learning model on a standard built-in dataset such as MNIST, with a focus on the end-to-end workflow rather than advanced model tuning. You will create or activate a GPU-ready environment, launch a Jupyter Notebook or script on the cluster, run the training on a GPU, and monitor GPU utilization and memory usage while the job runs. Throughout the exercise, you will practice saving model outputs, logs, and metrics in a structured way so they can be reported or reused later. By the end of the tutorial, you should feel more confident setting up GPU environments, running and monitoring simple deep learning jobs, and interpreting basic GPU usage patterns.

Similar to other tutorials of the series, the training will follow the "flipped-classroom" active learning style. Before coming to the lab, participants will complete a self‑paced Blackboard module that includes short hands‑on activities (called "DoIT Yourself Activities"). During the in‑person lab session, we will focus on questions, troubleshooting, and practicing examples so that you can further your understanding and strengthen the workflow.

These events are hybrid, but there is limited support for online participants. For those attending in-person: Be sure to bring your laptop!

Note that it's important to RSVP by Monday, 4/13/2026, so that we can provision accounts on the chip HPC and make Blackboard content available to you.

Remember that it's important that each participant review the available content before the event to make the best use of the synchronous classroom session (this event). We'll send a reminder to those who have RSVPed a day or so before the event!