Metraipolis › archive › dc_b9933ca84db5
Stencil Computation at the Intersection of AI and HPC — arXiv:2609.10368
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.10368
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T07:54:02.792Z
- SHA-256 of the archived text
f2d51dbdb1f26b381743563f15199982ce81c4ecf20303df86d28d0656c3565b- Stable URL of this page
https://metraipolis.com/documento/dc_b9933ca84db5
How to cite
Stencil Computation at the Intersection of AI and HPC — arXiv:2609.10368 — arXiv (en), https://arxiv.org/abs/2609.10368 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b9933ca84db5 (Metraipolis archive dc_b9933ca84db5, sha256 f2d51dbdb1f26b381743563f15199982ce81c4ecf20303df86d28d0656c3565b)
BibTeX
@misc{dcb9933ca84db5,
title = {Stencil Computation at the Intersection of AI and HPC — arXiv:2609.10368},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.10368},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b9933ca84db5; sha256 f2d51dbdb1f26b381743563f15199982ce81c4ecf20303df86d28d0656c3565b},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.10368. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Tensor compilers such as TinyTC and OpenAI Triton were originally developed for AI workloads, but the same tiling and memory abstractions can be applied to implement efficient high-order stencils for scientific and industrial applications. We demonstrate this for an 8th-order, 25-point acoustic stencil with boundary conditions over an a demanding-sized grid, targeting GPGPUs, where we compare the hardware-specialized TinyTC implementation with a portable PyTorch/Triton implementation. The target platforms for evaluation include Intel B70, B580, GPU MAX 1550, NVIDIA A100/RTX6000 Blackwell/H100, and AMD MI325x. For instance, on Battlemage B580 TinyTC reaches 15.6 Gpts/s versus 13.5 Gpts/s for PT/Triton under random initialization, while zero-initialized runs reach up to 35.8 Gpts/s due to hardware memory compression. Using roofline and memory-hierarchy profiling, we show that -as expected- performance is predominantly bandwidth-limited and that compiler-managed L1/LSC caching can effectively replace programmer-managed shared-memory staging for this stencil class. Overall, the results position TinyTC as the performance-oriented path on Intel hardware and PyTorch/Triton as a strong portability/productivity baseline for cross-vendor HPC stencil development. Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-09-10 in cs.DC). Categorie arXiv: Distributed, Parallel, and Cluster Computing (cs.DC). Pagina del lavoro: https://arxiv.org/abs/2609.10368 — 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 2026-10-16T07:25:19.018Z dalla pagina https://arxiv.org/catchup/cs.DC/2026-09-10?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).