mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2026-09-04 02:37:27 +02:00
* vulkan: RDNA3 static mat-vec rows above four columns On RDNA3 above four columns a static 4 rows for all types benches faster than the default. * vulkan: RDNA3 static mat-vec-id rows mul_mat_vec_id has no column dimension to switch on. On my Strix Halo machine, a static 4 is faster here than the defaults across types and batch sizes.