Impact Area: Science and Engineering Applications
This is not a destination as much as it’s a journey, and exascale is the next way on this journey towards ever increasingly more powerful computers.
[What can exascale computing do?]
It’s about a thousand times faster than the machines we had 10 years ago. With a billion things computing at a billion times per second, we’re going to have smarter AI [artificial intelligence]. We’ll have higher-resolution simulations. This is going to impact pretty much every technology-oriented problem—whether it’s climate, fluid dynamics, biology, chemistry, materials, combustion engine design, nuclear reactor design. All of that is going to get both boost from the core compute in exascale but then an extra boost from applying machine learning. It’s going to open up many possible paths of future development.
Climate simulations that are pushing towards the kilometer resolution will give us unprecedented insight as to how climate change is happening and what we can do about it. For drug development and understanding the system of biology in diseases like cancer and COVID, for example, we’ve built an AI-driven modeling pipeline that combines physics-based models with AI techniques for finding matches to any viral targets. And we need this large-scale computing so we can search much deeper into the space of possible therapeutics.
It’s that very opportunity of doing something new because you’ve got more capability, more capacity—that’s the creative spark. It gives us this kind of new environment in which we can imagine how we might solve new problems. And that’s the magic of exascale computing.