git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit

]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit

overview / pkg / ggml / sources / llama.cpp / commit

author	Hongqiang Wang <redacted>
	Fri, 17 Jul 2026 19:02:27 +0000 (12:02 -0700)
committer	GitHub <redacted>
	Fri, 17 Jul 2026 19:02:27 +0000 (12:02 -0700)
commit	6bdd77f13cf11b264b4231d320afc404f48d576e
tree	f926dfb4c3315e288b83bc1c3b167d4f7ca451d6	tree
parent	86d86ed4396b4130922f7b9af26e3d9fc11a591b	commit \| diff

opencl: read/write MoE dp4a activation tiles to local memory as 128-bit (vectorized LD/ST perf opt) for Adreno GPUs (#25810)

* opencl: read MoE dp4a activation tile as 128-bit local loads

* opencl: vectorize MoE dp4a activation staging as 128-bit loads

Packaging of ggml-org/llama.cpp

RSS Atom

ggml/src/ggml-opencl/kernels/gemm_moe_mxfp4_q8_1_dp4a.cl		diff \| blob \| history
ggml/src/ggml-opencl/kernels/gemm_moe_q4_0_q8_1_dp4a.cl		diff \| blob \| history
ggml/src/ggml-opencl/kernels/gemm_moe_q4_k_q8_1_dp4a.cl		diff \| blob \| history
ggml/src/ggml-opencl/kernels/gemm_moe_q6_k_q8_1_dp4a.cl		diff \| blob \| history
ggml/src/ggml-opencl/kernels/gemm_moe_q8_1_dp4a.cl		diff \| blob \| history