llama.cpp/ggml
Gong-Mi 9b86eb6207 ggml-vulkan: optimize Mali-G720/Adreno tuning & fix stability
- Implement 4x4 warptile tuning for Mali-G720/Immortalis MC12.
- Optimize tuning parameters for ARM Mali and Qualcomm Adreno.
- Fix matrix multiplication out-of-bounds (OOB) access by moving restrictions to initialization.
- Ensure stability by removing risky subgroup size clamping on Qualcomm devices.
2026-02-02 13:06:33 +08:00
..
cmake ggml: Skip backend library linking code when GGML_BACKEND_DL=ON (#15094) 2025-08-07 13:45:41 +02:00
include ggml : add ggml_build_forward_select (#18550) 2026-01-19 20:03:19 +02:00
src ggml-vulkan: optimize Mali-G720/Adreno tuning & fix stability 2026-02-02 13:06:33 +08:00
.gitignore vulkan : cmake integration (#8119) 2024-07-13 18:12:39 +02:00
CMakeLists.txt ggml : bump version to 0.9.5 (ggml/1410) 2025-12-31 18:54:43 +02:00