ALCF Developer Session: Inside the NVIDIA Ampere A100 GPU in ThetaGPU and Perlmutter

On Wednesday, July 28 at 11 am CT, the Argonne Leadership Computing Facility will present a developer session focused on utilization of the Nvidia Ampere A100 GPU in ALCF’s ThetaGPU and NERSC’s Perlmutter supercomputers. Registration is here.

The Ampere A100 GPU builds on the performance of the NVIDIA V100 GPU and includes new features for both HPC and deep learning. Particularly, tensor cores have been improved (including FP64 compute and efficient evaluation of sparse computation); the L2 cache is expanded and includes new residency controls; the SMs have enhanced support for asynchronous operations involving shared memory; and the new multi-instance GPU feature allows true isolation of partitions of the GPU.

NVIDIA has also invested significant effort in ensuring that both HPC applications and DL frameworks achieve high performance on A100. This talk will explain how these (and other) aspects of A100 enable computational scientists to fully leverage this architecture.

The speaker is Max Katz, a Senior Solutions Architect at NVIDIA who works with the US Department of Energy on the deployment and use of their GPU-powered supercomputers. Katz is a computational astrophysicist who researches explosions of stars using fluid dynamics simulations. He holds a PhD in Physics from Stony Brook University and a BS in Physics from Rensselaer Polytechnic Institute.