Metraipolis › archive › dc_b7bc2a4dd5d0

The Rise of Low Precision Floating Point — arXiv:2611.00301

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.00301
License (as archived)
CC0-1.0
Archived
2026-11-03T18:28:34.445Z
SHA-256 of the archived text
fc6d12572918510e7fca80127e6f2dc36924a4711b835b0a399d81439addfcaf
Stable URL of this page
https://metraipolis.com/documento/dc_b7bc2a4dd5d0

How to cite

The Rise of Low Precision Floating Point — arXiv:2611.00301 — arXiv (en), https://arxiv.org/abs/2611.00301 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b7bc2a4dd5d0 (Metraipolis archive dc_b7bc2a4dd5d0, sha256 fc6d12572918510e7fca80127e6f2dc36924a4711b835b0a399d81439addfcaf)
BibTeX
@misc{dcb7bc2a4dd5d0,
  title = {The Rise of Low Precision Floating Point — arXiv:2611.00301},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00301},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b7bc2a4dd5d0; sha256 fc6d12572918510e7fca80127e6f2dc36924a4711b835b0a399d81439addfcaf},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00301. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Floating point computing has dominated scientific computations since the 1950s because of its versatile ability to approximate real numbers with both sufficiently high precision and sufficiently great range. Since 1985, two specific formats have become standard across the industry, one with 32 bits which is adequate for many purposes and one with 64 bits which allows numbers with both much higher precision and much greater range for applications where these are needed. However, during the past decade, deep learning has come to dominate computation, and narrower floating point formats with lower precision and reduced range have steadily become more common. In this article we discuss this phenomenon, why it is happening, how various 16-bit, 8-bit and even 4-bit floating point formats have been developed, and efforts to standardize these formats.

Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.AR, nuovi lavori del giorno).
Categorie arXiv: Mathematical Software (cs.MS) ; Hardware Architecture (cs.AR); Machine Learning (cs.LG); Numerical Analysis (math.NA).
Pagina del lavoro: https://arxiv.org/abs/2611.00301 — 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 2026-11-03T18:10:37.672Z dalla pagina https://arxiv.org/list/cs.AR/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).