Adam Caldwell 6d4cdef511 Optimize mul_mat_q8_1_r8_q8_2 with AVX-512 for faster Q4_K/Q5_K prompt processing (#1578) ... Add HAVE_FANCY_SIMD path that processes 16 rows at a time using 512-bit operations, combining two R8 groups via _mm512_inserti32x8. Reuses the existing qx_r8_q8_dot_product 512-bit overload for the inner dot product. Also updates num_rows for Q8_1 to 16 under HAVE_FANCY_SIMD. Co-authored-by: Adam Caldwell <accaldwell@users.noreply.github.com> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>		2026-04-06 09:07:23 +02:00
..
cmake	Merge mainline llama.cpp (#3)	2024-07-27 07:55:01 +02:00
include	Bonsai support (AVX2, generic) (#1570)	2026-04-02 16:54:08 +02:00
src	Optimize mul_mat_q8_1_r8_q8_2 with AVX-512 for faster Q4_K/Q5_K prompt processing (#1578)	2026-04-06 09:07:23 +02:00
.gitignore	Merge mainline llama.cpp (#3)	2024-07-27 07:55:01 +02:00
CMakeLists.txt	Enable all CPU-backend FA supported quants by default (#1549)	2026-03-29 14:36:09 +02:00