Files
whisper.cpp/ggml
WakeUpMorty dca2df4211 vulkan: add dedicated iq4_xs mat-vec shader (llama/28426)
* vulkan: add dedicated iq4_xs mat-vec shader

Dedicated mul_mat_vec_iq4_xs for the dmmv path, replacing the generic fallback. ~+6-17% token generation on RDNA4 depending on model.

Assisted-by: Pi agent with Qwen3.8 27B

* vulkan iq4_xs: remove dead n_it unroll branch

Remove the n_it <= 8 experimental branch that attempted to fully unroll
the block loop. Since n_it is a runtime value, [[unroll]] is ignored by
the compiler, making both branches equivalent. Kept the simple loop
matching mul_mat_vec_iq3_s.comp.
2026-09-14 20:45:06 +03:00
..