How Can Lessons Learned in the Past Forty Years Guide Future HPC Research Strategies?

Moderator: Paul Messina (Argonne National Laboratory)

Panelists: Arthur Bland (Oak Ridge National Laboratory), Jackie Chen (Sandia National Laboratories), Phil Colella (Lawrence Berkeley National Laboratory), Jack Dongarra (University of Tennessee, Oak Ridge National Laboratory), Thom Dunning (University of Washington, Pacific Northwest National Laboratory), Wendy Huntoon (Keystone Initiative for Network Based Education and Research), Dan Reed (University of Iowa), Eli Dart (Lawrence Berkeley National Laboratory)

Abstract: The HPC, computer science, and mathematics communities face disruptive changes due to the end of Dennard scaling and the emergence of quantum, neuromorphic, and other technologies. Having weathered previous disruptions, are there insights into effective investment strategies and guidance in research investments in the coming decades? Drawing on the experience of the Advanced Scientific Computing Research program of the U.S. Department of Energy, as well as other agencies, the panelists will address questions including:

Which programmatic techniques, including partnering with other agencies or industry, were most successful and might translate to the future?

What are examples of investment failures that might have been avoided?

What lessons have been learned to help adapt to the end of Moore's Law and the likely shift in software architecture that will be needed?

What has been the impact of supercomputing facilities and computer networks and will such facilities continue to be relevant?

Presentation: File


Back to the Panel Archive Listing