400 - Specialized Workloads & Staying Current

Choose a topic after you are comfortable logging into ICE and submitting a Slurm job.

In this module

• 401 - Running LLMs with Ollama - request GPU resources, run a small model, and manage model storage.
• 402 - Staying Current - check PACE updates, inspect available software, and keep a reproducible Python environment.
• 403 - Distributed Training - learn PyTorch DDP and inspect a two-GPU Slurm example.
• 404 - Bioinformatics - learn common biological data formats and when HPC helps. This is an overview, not a step-by-step analysis.
• 405 - Run an ML Notebook on CPU and GPU - launch Jupyter on ICE and run the same small model on a CPU session and a GPU session.

Check the official PACE documentation for current cluster policies and software before trying a new workload.