This is an informal workshop mostly involving researchers at ALCF and CPS. The goal is to focus on HPC applications with GPUs and to focus on the nitty-gritty of HPC applications running on GPUs. Topics include programming models, performance bottlenecks, algorithmic issues, status of the GPU support toolchain, differences between the available GPUs, thoughts for the future, and anything else that might be relevant.
The workshop is organized around a quick introduction to the use of GPUs across different science domains (Overview) followed by very short talks on specific simulation codes (Presentations), and ending with a final session to assess the current situation and actionable items for the future (Discussion).
