Metraipolis › archive › dc_86d3b2cef130
Performance Analysis of Low-Order, GPU-accelerated Finite Element Kernels using Kokkos — arXiv:2609.21681
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.21681
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-10T01:40:22.929Z
- SHA-256 of the archived text
c852d15076a9940114cfb5e8a41483807285b207e359d58b23b7c84f06a7908d- Stable URL of this page
https://metraipolis.com/documento/dc_86d3b2cef130
How to cite
Performance Analysis of Low-Order, GPU-accelerated Finite Element Kernels using Kokkos — arXiv:2609.21681 — arXiv (en), https://arxiv.org/abs/2609.21681 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_86d3b2cef130 (Metraipolis archive dc_86d3b2cef130, sha256 c852d15076a9940114cfb5e8a41483807285b207e359d58b23b7c84f06a7908d)
BibTeX
@misc{dc86d3b2cef130,
title = {Performance Analysis of Low-Order, GPU-accelerated Finite Element Kernels using Kokkos — arXiv:2609.21681},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.21681},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_86d3b2cef130; sha256 c852d15076a9940114cfb5e8a41483807285b207e359d58b23b7c84f06a7908d},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.21681. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We study performance portability for low-order, matrix-free finite element kernels, using the example of a vectorial, variable-coefficient PDE operator originating in geophysical models. Written in Kokkos, the kernel is compared on NVIDIA H100, AMD MI250X, AMD MI300A and Intel PVC Max 1550 GPUs. Owing to its low order and to optimizations that reduce the arithmetic, the kernel has a low arithmetic intensity, so that its performance is determined by how the finite element assembly is mapped onto the memory hierarchy. This is a dimension in which the architectures differ even within one vendor family, causing different performance characteristics. We examine how Kokkos' hierarchical parallelism and shared scratch memory, which are used for the shared degrees of freedom of the conforming discretization, behave on each device. Finally, we show how portability gaps can be narrowed with tuning levers such as the size of the thread groups, the balance between occupancy and register use, and the atomic accumulation strategy at the end of the kernel. Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-09-21 in cs.PF). Categorie arXiv: Performance (cs.PF). Pagina del lavoro: https://arxiv.org/abs/2609.21681 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-10T01:08:03.188Z dalla pagina https://arxiv.org/catchup/cs.PF/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).