In view of upcoming exascale supercomputing facilities, programming efficiently high-performance GPU hardware is of increasing importance for each scientific field that relies on compute-heavy simulations, lattice quantum chromodynamics (lattice QCD) being a prominent example. What kind of changes in the popular codebases are required in order to leverage the computational power and throughput of the GPU thoroughly? This workshop aims to discuss recent progress in lattice QCD and HPC community in developing simulation software for various GPU architectures. At the same time, we plan to facilitate exchange between the lattice QCD community and HPC experts from adjacent disciplines and introduce students to challenges in these rapidly developing fields. The workshop includes a dedicated session on machine learning applications as well as talks aimed at a broader HPC audience.
Youtube recording of the first day: https://www.youtube.com/watch?v=Kfy4FSsFtSw
Invited speakers:
- Luca Benini (ETHZ)
- Tal Ben-Nun (ETHZ)
- Denis Boyda (ANL)
- Peter Boyle (BNL/University of Edinburgh)
- Kate Clark (NVIDIA)
- Henry Ann Day-Hall (Czech Technical University in Prague)
- Heng-Tong Ding (Central China Normal University)
- Steven Gottlieb (Indiana University)
- Dan Graur (ETHZ)
- Balint Joo (Oak Ridge National Laboratory)
- Javad Komijani (ETHZ)
- Piotr Korcyl (Jagiellonian University)
- Bartosz Kostrzewa (University of Bonn)
- Bill McColl (Huawei)
- Kari Rummukainen (University of Helsinki)
- Andrea Shindler (Michigan State)
- Daniel Spitz (Heidelberg University)
- Foteini Strati (ETHZ)
- Mathias Wagner (NVIDIA)
- Felix Ziegler (University of Edinburgh)
Booths:
- Axelera AI
Organizers (ETHZ):
- Fritz Bayer (D-BSSE)
- Luca Benini (D-ITET)
- Roman Gruber (D-PHYS)
- Ana Klimovic (D-INFK)
- Marina Krstic Marinkovic (D-PHYS)
- Wayne Zeng (AI Center)
