Metraipolis › archive › dc_717d1da734af

FSGen: Agile Fused and Sparse Accelerator Generator with Accurate Power Model for LLM Applications — arXiv:2608.09252

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.09252
License (as archived)
CC0-1.0
Archived
2026-10-19T08:53:21.521Z
SHA-256 of the archived text
a7a99c16885c1e6ce0853ce544e4b67b36d909f2dd09dd332d50c328a7df6ba7
Stable URL of this page
https://metraipolis.com/documento/dc_717d1da734af

How to cite

FSGen: Agile Fused and Sparse Accelerator Generator with Accurate Power Model for LLM Applications — arXiv:2608.09252 — arXiv (en), https://arxiv.org/abs/2608.09252 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_717d1da734af (Metraipolis archive dc_717d1da734af, sha256 a7a99c16885c1e6ce0853ce544e4b67b36d909f2dd09dd332d50c328a7df6ba7)
BibTeX
@misc{dc717d1da734af,
  title = {FSGen: Agile Fused and Sparse Accelerator Generator with Accurate Power Model for LLM Applications — arXiv:2608.09252},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.09252},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_717d1da734af; sha256 a7a99c16885c1e6ce0853ce544e4b67b36d909f2dd09dd332d50c328a7df6ba7},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.09252. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

With the growing demand of artificial intelligence (AI) applications, large language models (LLMs) have become important workloads in many domains. The question of how to efficiently generate optimal AI chip accelerator designs remains unresolved and challenging. Currently, there is a lack of end-to-end design methodologies for efficient design space exploration (DSE). We propose FSGen, an agile framework for attention-based LLM accelerator generation with an early-stage PPA estimator. FSGen supports fused operator dataflows and sparsity with a diverse design space and finds designs with 1.4x better power efficiency or 10x speedup with similar PPA metrics compared to prior work. Pareto-optimal designs have much better performance over a wide range of LLM benchmarks and have 58x better figures of merit (FoM). Design exploration is also faster due to our PPA estimators, which have better accuracy than prior art and reduce DSE runtime drastically.

Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-08-11 in cs.AR).
Categorie arXiv: Hardware Architecture (cs.AR).
Pagina del lavoro: https://arxiv.org/abs/2608.09252 — 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 2026-10-19T08:50:52.503Z dalla pagina https://arxiv.org/catchup/cs.AR/2026-08-11?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).