Communications of the ACM
Digital Twin for Intense Weather Gives Scientists 'Control Loop'
Computer maker Cerebras used its AI computer on a non-AI problem: simulating "buoyancy-driven Navier-Stokes flows" that capture dynamics of many systems in nature and the built environment.
Credit Cerebras/DoE NETL 2023
Scientists at artificial intelligence computing developer Cerebras and the U.S. Department of Energy's National Energy Technology Laboratory (NETL) said they can model extreme weather by accelerating field equations.
Said Cerebras' Andrew Feldman, "This is a real-time simulation of the behavior of fluids with different volumes in a dynamic environment."
He explained this digital twin of real-world conditions basically enables a "control loop" for manipulating reality.
Cerebras' CS-2 supercomputer can model the Rayleigh-Bénard convection process caused by fluids being heated from the bottom and cooled from the top, while the Cerebras-NETL Wafer Scale Engine field equation application programming interface describes scientific equations.
The partners said, "The simulation is expected to run several hundred times faster than what is possible on traditional distributed computers, as has been previously demonstrated with similar workloads."
From ZDNet
View Full Article
Abstracts Copyright © 2023 SmithBucklin, Washington, D.C., USA
No entries found