Samuel Omlin
Computational Scientist and Responsible for Julia computing, at the Swiss National Supercomputing Centre (CSCS), ETH Zurich
Session
We present an approach for building a Reactant backend for ParallelStencil, a Julia package for high-performance stencil computations. The approach includes the generation of kernel code and data structures that are pre-optimized to serve as optimal input for Reactant to generate efficient and correct GPU, TPU, and CPU code. We report performance of representative stencil mini-apps on recent hardware platforms, including NVIDIA H100 GPUs, evaluate it in absolute terms, and compare it with performance obtained with straightforward implementations using CUDA.jl, KernelAbstractions.jl, and other Julia packages that enable explicit GPU kernel programming.