Deep learning for new alloys

Supercomputer simulations are helping scientists discover new high-entropy alloys. XSEDE allocations on TACC’s Stampede2 supercomputer supported density function theory calculations for largest database yet of high-entropy alloy properties. Deep Sets architecture generated predictive models on Stampede2 for the properties of new high-entropy alloys. Study of high-entropy alloys represents an effort of materials scientists to develop new materials for a more sustainable future.