Hands-On Introduction to HPC and AI

SC26 Student Programming Tutorial

A full-day, hands-on path from logging into an HPC cluster to running parallel programs on CPUs and GPUs, then using AI models and agents on the same system.

Session details

• Format: Hands-on cluster exercises and challenges

• Venue: SC26, Chicago · Room TBD

Start the Tutorial

Getting Help

If you get stuck, ask for help early. Instructors will be in the room throughout the tutorial, and we may also provide an online help channel such as Slack. Final help-channel details will be added before the tutorial.

Course Path

Module	Topic	What You’ll Do
1	HPC Foundations	Explore the cluster, filesystems, software modules, and hardware basics.
2	Slurm Scheduling	Submit, monitor, and understand jobs running on compute nodes.
3	OpenMP	Use shared-memory parallelism to speed up CPU programs.
4	MPI	Coordinate distributed processes and message passing across ranks.
5	HIP GPU Programming	Write GPU kernels and reason about host/device memory movement.
6	AI on HPC	Run inference and fine-tune models using GPU-backed batch jobs.
7	AI Agents	Build a tool-using agent and pull the day’s HPC skills together.
Reference	Commands Reference	Find common shell, Slurm, OpenMP, MPI, HIP, Python, and agent commands.