Carnegie Mellon-led team generates atomic-scale simulations at Argonne in search of the best new solar cell materials.
Company embraces supercomputing in quest for viable fusion energy.
The current class of early-career honorees applies machine-learning tools to science problems.
Princeton researchers apply deep learning to a new code, Fusion Recurrent Neural Network (FRNN), to forecast events that disrupt fusion reactions.
Bagel-shaped fusion reactors called tokamaks produce energy of the kind that powers the sun. But massive disruptions can halt fusion reactions and damage the reactors. Princeton University researchers’ Fusion Recurrent Neural Network (FRNN) code harnesses an artificial-intelligence tool called deep-learning to predict disruptive events. Researchers can also use the code to make predictions that could open avenues for active reactor control and optimization. The method, reported in Nature, holds promise for enabling steady-state operation of tokamaks.View full highlight »