Skip to content

UMD at the Supercomputing 2024 (SC24) Conference

The University of Maryland will be returning to the annual ACM/IEEE Supercomputing conference, SC24 brings together scientists, engineers, and researchers from the international community with a technical program full of presentations and tutorials. It is an excellent setting to meet other like-minded individuals with whom you can learn and share information.

SC24 Conference

  • Sunday, Nov 17 - Friday, Nov 22, 2024
  • Atlanta, GA
  • UMD Booth number 4015

The UMD booth will serve to showcase research and projects currently being performed at the University. Located in the exhibit hall, the UMD booth is number 4015, just look for the giant terrapin suspended from the ceiling.

Booth presentations will be held on Tuesday through Thursday during the conference. The schedule of speakers is included below. Be sure to check back here for updates, as this list is subject to change until the end of the conference.

Speaker Schedule

Tuesday, Nov 19
Day Time Topic Speaker
Tue 10:10 AM Taking GPU Programming Models to Task for Performance Portability Joshua Davis
(University of Maryland)
Tue 1:10 PM Eve: Pruning Adam's State for Scalable Deep Learning Aditya Tomar
(UC Berkeley)
Tue 3:10 PM Strategies for Parallelizing an Agent-Based Model of Infectious Disease Spread Joy Kitson
(University of Maryland)
Wednesday, Nov 20
Day Time Topic Speaker
Wed 10:10 AM Loki: Low-Rank Keys for Efficient Sparse Attention Prajwal Singhania
(University of Maryland)
Wed 1:10 PM Improving Build Likelihood in Package Managers with Probabilistic Constraints Daniel Nichols
(University of Maryland)
Wed 2:00 PM Quantum Supremacy? What can be accomplished with NISQ computers? Franz Klein
(University of Maryland))
Wed 3:10 PM Insights from Longitudinal GPU Workload Monitoring on Perlmutter Onur Cankur
(University of Maryland)
Thursday, Nov 21
Day Time Topic Speaker
Thu 10:10 AM Creating Code LLMs for HPC: It's LLMs All the Way Down Aman Chaturvedi
(University of Maryland)
Thu 1:45 PM A hybrid tensor-expert-data parallelism approach to optimize mixture-of-experts training Siddharth Singh
(University of Maryland)