diff options
| author | Onur Gündüz <ogunduz@gmail.com> | 2026-06-27 00:58:05 +0200 |
|---|---|---|
| committer | Onur Gündüz <ogunduz@gmail.com> | 2026-06-27 01:03:41 +0200 |
| commit | af55803d0dff9eb692c760c8d486e736c945061d (patch) | |
| tree | 0875a1112bf98304e445e0544c4309670d6f4769 /libffwd/avx.c | |
| parent | 0f861b7f3098750ddee4f6570f3be228d1da5320 (diff) | |
| download | ffwd-af55803d0dff9eb692c760c8d486e736c945061d.tar.gz ffwd-af55803d0dff9eb692c760c8d486e736c945061d.zip | |
chore: clean up comment formatting and wording
Diffstat (limited to 'libffwd/avx.c')
| -rw-r--r-- | libffwd/avx.c | 20 |
1 files changed, 10 insertions, 10 deletions
diff --git a/libffwd/avx.c b/libffwd/avx.c index 41259a1..7cbbd5a 100644 --- a/libffwd/avx.c +++ b/libffwd/avx.c @@ -1,12 +1,14 @@ -/* - * avx.c - x86 SIMD hot kernels (AVX2+FMA, with AVX-512 when available) +/* x86 SIMD hot kernels: AVX2+FMA, plus AVX-512 when available. * - * bf16 to f32 conversion: load 16 uint16, zero-extend to 32-bit, shift left 16. - * Uses AVX-512F+BW for 16-wide bf16 matvec (dominant for seq_len==1), and - * AVX2+FMA / AVX-512F for f32 attention helpers (which operate on cache-resident data). + * BF16 -> F32 conversion: + * load 16 uint16 values, zero-extend to 32-bit, then shift left 16. * - * The bf16 matvec processes 4 output rows simultaneously to reduce instruction - * overhead and improve out-of-order execution on memory-bound workloads. + * Uses AVX-512F+BW for 16-wide BF16 matvec, dominant when seq_len == 1. + * Uses AVX2+FMA / AVX-512F for F32 attention helpers, which operate on + * cache-resident data. + * + * BF16 matvec processes 4 output rows at once to reduce instruction overhead + * and improve out-of-order execution on memory-bound workloads. */ #include "impl.h" @@ -16,9 +18,7 @@ # include <immintrin.h> # include <string.h> -/* ===================================================================== - * BF16 matvec - uses AVX-512F+BW when available, AVX2 fallback otherwise - * ===================================================================== */ +/* BF16 matvec: uses AVX-512F+BW when available, AVX2 fallback otherwise */ # if defined(__AVX512F__) && defined(__AVX512BW__) |
