PI Name: Michel Schanen
PI Institution: argonne-national-laboratory
Collaborating ANL Division: mathematics-and-computer-science-mcs
Description: We have ported our software stack to V100 and A100 GPUs. As an ECP project (ExaSGD) we target the upcoming AMD and Intel architectures. However we lack a decent development platform. Our research involves porting Julia on AMD GPUs and using our GPU power flow solver that is used on a single node in our distributed large scale setup.
Testbed: AMD GPU