Skip to content

High Performance Computing in the Age of LLMs #73

Description

@fathimamaha

Talk Request

Topic(s):
High-Performance Computing in the Age of LLMs & What It Takes to Slice Large Problems for Microprocessors

Additional Notes:

While in school at UChicago, I worked at Argonne National Lab on frontier physics problems, I want to give people a real look at high-performance computing in the age of machine learning and LLMs. I’d love to bring an actual physical piece of the Mira supercomputing node (a prized possession of mine haha) to show what the hardware looks like up close. Using weather prediction/ training LLMs as an example, i'll dive into what it actually takes to slice complex physical systems so they can run across massive arrays of microprocessors.

What would the ideal length be?

  • 5-15 minutes (lightning talk)
  • 20-30 minutes
  • 30-45 minutes
  • 60 minutes or more

On Behalf Of

Requesting people or Meetups:

The best way to reach out to them:

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions