Sparse-ternary-fma – 23x faster AVX-512 kernel for FHE&AI, from my ePrint paper

(github.com) by HyperFoldUK | Dec 23, 2025 | 1 comments on HN