Portable GPU Acceleration of HPC Applications with Standard C++
Date:
I helped lead the tutorial, Portable GPU Acceleration of HPC Applications with Standard C++. This session focused on leveraging ISO C++23 parallelism and concurrency features to portably accelerate HPC applications on heterogeneous systems (AMD, Intel, NVIDIA GPUs) without vendor-specific extensions. We provided hands-on experience covering classic HPC patterns like multi-dimensional loops, reductions, and overlapping MPI communication with GPU computation.