Training Seminar    

GenAI on the HPCC  

Michigan State University’s Institute for Cyber-Enabled Research (ICER), in partnership with the Institute for Biodiversity, Ecology, Evolution and Macrosystems (IBEEM), are excited to announce the Fall-2026 GenAI on the High-Performance Computing Center (HPCC) seminar. This non-credit, two-part series is designed to provide researchers from a wide variety of disciplines with a hands-on introduction to tools for integrating GenAI into research computing pipelines on ICER’s HPCC.*  This is a hybrid seminar with both an in-person and virtual, synchronous option.  

*Participants need an active HPCC account to access the HPCC. If you do not have an existing account, please see our documentation on Obtaining an HPCC Account.  

The prerequisites for meaningful participation in the hands-on exercises are listed below.  

• Active ICER HPCC account  
• Basic understanding of the HPCC resource environment and the SLURM scheduler (see HPCC Foundations Webinar Series)
• Basic competency with Python and GNU/Linux Operating Systems (necessary code will be provided) (see Python on the HPCC)
• Basic understanding of fundamental machine learning concepts e.g., regression, classification, training, testing/inference, supervised vs unsupervised learning  

Fall 2026 GenAI on the HPCC Workshop Schedule

Part 1: GenAI overview and getting started with local LLMs

• Wednesday, November 4th, 2:00 pm - 5:00 pm    
• Learning objectives:  
  • Understand core concepts of Generative AI (GenAI) and Large Language Models (LLMs)  
  • Understand differences between cloud-hosted and locally run LLM workflows  
  • Understand basics of running and managing LLMs locally  
  • Use locally served LLMs with a Python environment for simple programmatic inference  
  • Recognize resource considerations (memory, CPU, GPU, etc.) when running LLMs locally  
  • Identify common use-cases and how to integrate local GenAI models into workflows (e.g., domain specific documentation chatbot, literature review, entity-relation extraction)  
  • Understand and apply lightweight fine-tuning on local LLMs    

Part 2:  RAG systems and fine-tuning

• Wednesday, November 11th, 2:00 pm - 5:00 pm  
• Learning objectives:  
  • Construct a minimal RAG system with use-case data  
  • Understand concepts for assessing model performance with quantitative and qualitative criteria  
  • Understand resource requirements for fine-tuning and RAG workflows  
  • Manage model artifacts, embeddings, and checkpoints  
  • Understand fundamental concepts of agentic GenAI systems  

Questions? Please contact Julian Venegas at venegas5@msu.edu