Jack Dongarra Early Career Award Winner: Staircases and Trampoline: HPC in the Age of Quantum Computing and AI

HPC -- both the community and the research field -- has evolved dramatically over the last few decades. Some of these evolutions were primarily driven by technological advancements, and some were due more to social dynamics (e.g., how we share research progress, how we conduct peer review of our research findings, and how we measure success).

A few years back, many of us were debating the retirement age or death of Moore’s Law. Now, we find ourselves in a very exciting time -- quite far away from debating retirement and death -- we are observing the toddler signs of what quantum computing and AI can do today and will be able to do in the near future, albeit at a very rapid pace. All rapid advancements naturally hold a mix of promise and uncertainty.

If you are excited by the promise of this new future but, at the same time, uncertain about what it holds for HPC research and education, the audience size of this conference room will confirm that you are not alone. I will try hard not to make predictions about the future — after all, we have a very boring future ahead if it can easily and accurately be predicted by me. I will share a few examples of what we, my research group, have been doing to prepare for a post-Moore future, a few lessons we have learned the hard way, and how we can prepare ourselves better for quantum computing and AI-enabled HPC — an exciting future where all three pillars are likely to be disrupted: what problems we research, how we conduct our research, and how we communicate our research results.