Exascale Computing

Throughout the multiple phases of the Exascale Initiative, when people asked me what I was working on, I kept it simple and crisp: Exascale Computing. I learned to laser-focus on the meaningful challenges and develop simplicity elsewhere in life.

Today I am very proud to share with you one sweet fruit of our endeavour: the world’s fastest supercomputer, the Frontier, the next generation of supercomputer, expected to deliver more than 1.5 exaflops of peak processing power.

EPYC infinity fabric infographic

I am very grateful to people who gave me opportunity to be part of this exascale effort. In particular, it was clear that my technical domain at the time was Human Brain Mapping, and not the other, more directly relevant, HBM (high bandwidth memory); in-memory computing was about the development of novel biomarkers and approaches for early diagnosis and treatment of Alzheimer’s disease, not to bridge the gap between the memory and compute subsystems and to co-design both with a wide range of applications. I cannot thank enough the amazing people who supported me throughout this journey!

This again validated a belief I have always carried: I may not have a clue now but I can figure it out. The Frontier system is the proof that we together have figured something out, among many other current and future products. So, if you are like me, coming from a domain other than the “right” or “most desirable” one, do not let anyone tell you that you are an expert on X, hence you cannot possibly know or achieve Y. You can. Get on with the right challenges, simplify everything else. Be open-minded and give it all. You can do it!

For more about Exascale Project and the Frontier, check out the followings: AMD: Powering the Exascale Era and Exascale Computing Project.