Reproduce “Performance Analysis of Low-Order, GPU-accelerated Finite Element Kernels using Kokkos” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Fabian Böhm, Nils Kohl, Harald Köstler, Ulrich Rüde
The authors present a portable Kokkos matrix-free low-order finite-element viscous operator for geophysical simulation. It groups neighboring cells to reuse input data in GPU scratch memory and tests local output accumulation before atomic updates. On a 4 × 4 × 32 tile, MI300A throughput rises from 10.90 to 13.87 GDoF/s, while MI250X falls from 7.54 to 6.50, showing the change has different effects on these GPUs.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.