The 38th International Symposium on Lattice Field Theory

Performance optimizations for porting the openQ*D package to GPUs

30 Jul 2021, 07:00

15m

Oral presentation Algorithms (including Machine Learning, Quantum Computing, Tensor Networks) Algorithms (including Machine Learning, Quantum Computing, Tensor Networks)

Speaker

Roman Gruber (ETH Zurich)

Description

OpenQ$^\ast$D code has been used by the RC$^\ast$ collaboration for the generation of fully dynamical QCD+QED gauge configurations with C$^\ast$ boundary conditions. In this talk, optimization of solvers provided with the openQ$^\ast$D package relevant for porting the code on GPU-accelerated supercomputing platforms is discussed. We present the analysis of the current implementations of the GCR solver preconditioned with Schwarz alternating procedure for ill-conditioned Dirac-operators. With the goal of enabling support for GPUs from various vendors, a novel method of adaptive CPU/GPU-hybrid implementation is proposed.

Presentation materials

Performance optimizations for porting the openQ*D package to GPUs

Slides.pdf