Pull requests / #1411

#1411 OLDER_GPUS.md: the P100 row now has a measurement (2x P100, IQ3_S)

open · @lechlna000 · 0 评论 · 在 GitHub 查看

BenchmarksMulti-GPUModels & quants

描述

Follow-up to #1157. The Support matrix still said 'not measured' for the Tesla P100; #1157 supplied a 2x P100 IQ3_S report, so this fills the row in the same style as the P40 and V100 rows: prompt 526-532 tok/s, decode 28-32 tok/s at 128K prompt tokens, 3 of 3 needle checks at 122K, linking bench/results/2026-10-06-community-2x-p100/. Kept separate from the results PR so reviewers can assess the measurements directly, as COMMUNITY_BENCHMARKS.md asks.

站内延伸阅读

链到安装、模型与版本说明,便于 SEO/GEO,非官方 issue 正文。