Portable GPU Acceleration of HPC Applications with Standard C++

This hands-on tutorial teaches how to parallelize HPC applications on GPUs using the portable parallelism and concurrency features of the C++23 standard, without any language or vendor extensions, such that a single version of the code is portable to both multi-core CPUs and heterogeneous accelerated systems with GPUs. We further show how to integrate this approach with MPI to target homogeneous and heterogeneous HPC systems. The tutorial exercises follow classical HPC themes like a PDE solver mini-application for the 2D unsteady heat equation. The exercises provide attendees hands-on experience applying C++ parallel algorithms and execution policies to parallelize and accelerate HPC programs using only standard C++. The attendees are presented with problem-solving strategies for common tasks like computing reductions or running iterative solvers for multi-dimensional problems. Furthermore, the tutorial and exercises give attendees hands-on experience integrating C++ parallel algorithms into pre-existing MPI applications. All exercises can be carried out by the attendees on a web-based multi-GPU system provided as a part of the tutorial. Finally, we conclude with a summary of our professional experience applying the ISO C++ parallel programming model to accelerate large real-world HPC applications, an outlook of future topics in C++ standard parallelism, references, and further learning resources.