Metraipolis › archive › dc_fd344eed1827
Vectorization Of Narrow Matrix Multiplication for Ascend AI Inference Acceleration — arXiv:2609.16009
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.16009
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T19:29:34.308Z
- SHA-256 of the archived text
58baff67bd0483ca762e7df90372b191935e99bd15171675b4e1ce75cdf28d81- Stable URL of this page
https://metraipolis.com/documento/dc_fd344eed1827
How to cite
Vectorization Of Narrow Matrix Multiplication for Ascend AI Inference Acceleration — arXiv:2609.16009 — arXiv (en), https://arxiv.org/abs/2609.16009 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fd344eed1827 (Metraipolis archive dc_fd344eed1827, sha256 58baff67bd0483ca762e7df90372b191935e99bd15171675b4e1ce75cdf28d81)
BibTeX
@misc{dcfd344eed1827,
title = {Vectorization Of Narrow Matrix Multiplication for Ascend AI Inference Acceleration — arXiv:2609.16009},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.16009},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fd344eed1827; sha256 58baff67bd0483ca762e7df90372b191935e99bd15171675b4e1ce75cdf28d81},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.16009. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
This research proposes and evaluates a novel approach to optimizing matrix multiplication (MatMul) on Huawei Ascend NPUs, motivated by a key insight: during matrix-vector multiplication (narrow MatMul), the Cube Unit (AIC) is often underutilized, while the Vector Unit (AIV) remains idle for most of the operator runtime. In this paper, we introduce the MatMul algorithm, which uses vector instructions of AscendC to effectively offload computations from the Cube Unit to the Vector Unit. The algorithm was tested and applied to accelerating the inference of MLA DeepSeek-V3 operator. By successfully overlapping AIV and AIC computations, our optimization showed a mean performance gain of 20% for a single token processing scenario. Our work addresses a significant gap in the literature on practical optimization techniques for AscendC, despite the availability of documentation and the active CANN community. Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-09-16 in cs.DC). Categorie arXiv: Distributed, Parallel, and Cluster Computing (cs.DC). Pagina del lavoro: https://arxiv.org/abs/2609.16009 — 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 2026-10-15T18:35:05.414Z dalla pagina https://arxiv.org/catchup/cs.DC/2026-09-16?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).