EUVIMEDEUVIMED entdecken
Crossref

Reducing LLM Inference Memory Bandwidth via Frequent Exponent Value Encoding

Maxwell Michalec, Swamit Tannu, Gurindar Sohi

IEEE Computer Architecture Letters · 2026-01 · 25 (1) · 85-88

DOI: 10.1109/lca.2026.3671166 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Computer Architecture Letters, 2026-01. Verfasst von Maxwell Michalec und 2 weiteren Autor:innen. Die Arbeit nennt 9 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Kein Abstract verfügbar.

Zitieren

APA

Michalec, M., Tannu, S., & Sohi, G. (2026). Reducing LLM Inference Memory Bandwidth via Frequent Exponent Value Encoding. IEEE Computer Architecture Letters, 25(1), 85-88. https://doi.org/10.1109/lca.2026.3671166

Vancouver

Michalec M, Tannu S, Sohi G. Reducing LLM Inference Memory Bandwidth via Frequent Exponent Value Encoding. IEEE Computer Architecture Letters. 2026;25(1):85-88. doi:10.1109/lca.2026.3671166

BibTeX
@article{Michalec2026,
  author = {Michalec, Maxwell and Tannu, Swamit and Sohi, Gurindar},
  title = {Reducing LLM Inference Memory Bandwidth via Frequent Exponent Value Encoding},
  journal = {IEEE Computer Architecture Letters},
  year = {2026},
  volume = {25},
  number = {1},
  pages = {85--88},
  doi = {10.1109/lca.2026.3671166}
}
Kostenlos starten →