←back to thread

Faster Argmin on Floats

(algorithmiker.github.io)
19 points return_to_monke | 1 comments | 18 Sep 25 16:20 UTC | HN request time: 0.204s | source
1. why_only_15 ◴[20 Sep 25 04:25 UTC] No.45310355[source]▶
This trick is very useful on Nvidia GPUs for calculating mins and maxes in some cases, e.g. atomic mins (better u32 support than f32) or warp-wide mins with `redux.sync` (only supports u32, not f32).