Metraipolis › archive › dc_d287165ac739

The GPU Needs an Operating System — arXiv:2611.08720

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.08720
License (as archived): CC0-1.0
Archived: 2026-11-09T22:36:28.927Z
SHA-256 of the archived text: d1d2c65d64d7720e0f4b4fee8c59f4ebf9ae1ff527cf162eba549b0bbf58da9f
Stable URL of this page: https://metraipolis.com/documento/dc_d287165ac739

How to cite

The GPU Needs an Operating System — arXiv:2611.08720 — arXiv (en), https://arxiv.org/abs/2611.08720 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d287165ac739 (Metraipolis archive dc_d287165ac739, sha256 d1d2c65d64d7720e0f4b4fee8c59f4ebf9ae1ff527cf162eba549b0bbf58da9f)

BibTeX

@misc{dcd287165ac739,
  title = {The GPU Needs an Operating System — arXiv:2611.08720},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.08720},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d287165ac739; sha256 d1d2c65d64d7720e0f4b4fee8c59f4ebf9ae1ff527cf162eba549b0bbf58da9f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.08720. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

AI chips sit mostly idle while serving models, because replies must be fast and each step waits on memory. On two large open models, the math units were busy only about 6-11% of the time. We argue the GPU needs an operating system that keeps urgent work on schedule and fills idle capacity with other useful work. A feature that ships today recovers a quarter of a GPU with work designed to share, and a tenth with real jobs, at 5% slower decode. Our research system does better: one resident kernel runs a whole model faster than vLLM's best for a single request, and shares the GPU at under 2% slowdown. Scaling beyond one hand-built model needs compilers and a few new chip features, which we specify.

Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.DC, nuovi lavori del giorno).
Categorie arXiv: Distributed, Parallel, and Cluster Computing (cs.DC).
Pagina del lavoro: https://arxiv.org/abs/2611.08720 — 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 2026-11-09T22:31:17.037Z dalla pagina https://arxiv.org/list/cs.DC/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).