The Supercomputer in My Backyard Has Agents

Speaker: Dr. Michael E. Papka
Argonne National Laboratory and University of Illinois Chicago
Join ACM Chicago on Wednesday, December 2, 2026, for a talk exploring how supercomputers and AI agents are changing scientific discovery—from Aurora’s exascale computing to experiments guided by real-time data.
When and where

• Event: 5:30–7:30 PM Central Time (CST)
• Talk: begins around 6:00 PM Central Time
• In person: University of Illinois Chicago (UIC), Computer Design Research and Learning Center (CDRLC), Room 1413, 850 W Taylor St, Chicago, IL 60607
• Online: Zoom webinar, 6:00–7:30 PM Central Time. Please follow the online registration steps below.

How to register
Please RSVP to this Meetup event and select in-person or online attendance.
In-person attendees: Your Meetup RSVP is sufficient; no additional Zoom registration is needed.
Online attendees: After you RSVP on Meetup, open the online event link and register for the December 2 session on Zoom using your name and email address. Zoom will email you a personal joining link. Use that emailed link to join around 6:00 PM Central Time.
A Meetup RSVP alone does not register you for Zoom. Please complete Zoom registration separately for this event.
About the talk
Supercomputers used to wait for instructions. Increasingly, they don’t. This talk revisits our 2012 Chicago ACM presentation and traces ALCF’s path from Intrepid and Mira through Aurora, one of the world’s first exascale systems, and on toward Helios, the next-generation system planned for 2030.
The story is about more than faster hardware. Supercomputers are no longer only batch machines that run a job and hand back results days later. They are increasingly coupled in real time to experimental facilities such as Argonne’s Advanced Photon Source, steering experiments as the data arrives. And the newest users of these systems are not always people: AI agents are beginning to plan, launch, and analyze computational work alongside scientists, raising new questions about how facilities are designed, governed, and even measured. Is counting FLOPS still the right way to judge a supercomputer?
We conclude by examining what has changed since 2012, what has remained constant, and the implications for scientific computing as leadership-class systems take on a growing role in autonomous discovery.
About Dr. Michael E. Papka
Dr. Michael E. Papka is a Senior Scientist and an Argonne Distinguished Fellow. Since 2006, he has served as Argonne’s Deputy Associate Laboratory Director for Computing, Environment, and Life Sciences, and since 2010 as Director of the Argonne Leadership Computing Facility.
He is also the Warren S. McCulloch Professor of Computer Science at the University of Illinois Chicago. He serves as Director of UIC’s Electronic Visualization Laboratory and as Co-Director of the George Crabtree Institute for Discovery. Before joining UIC in 2022, he served as a Presidential Research, Scholarship and Artistry Professor of Computer Science at Northern Illinois University. Mike’s research spans scientific visualization, large-scale data analytics, and research infrastructure for scientific discovery. His current work focuses on analyzing data from supercomputers, large-scale scientific instruments, and sensor networks, with an emphasis on integrating edge and high-performance computing across the computing continuum.