diff options
author | Kawrakow <48489457+ikawrakow@users.noreply.github.com> | 2023-06-09 10:39:59 +0300 |
---|---|---|
committer | GitHub <noreply@github.com> | 2023-06-09 10:39:59 +0300 |
commit | 245fc3c37da5ac5963f9f11a9f4f2ac08d96afc6 (patch) | |
tree | b2312b5b19a6887526d9e25d41b29eb4fdbcd49e /ggml-metal.m | |
parent | 72ff5282bf0388c60821f504c4c8cc2b1f491aa6 (diff) |
metal : faster q4_0 (#1775)
* metal : 8% faster q4_0
Avoid copying into local uchar4 anf float4.
* metal : 17% faster Q4_0
Use 64 threads in a thread group.
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
Diffstat (limited to 'ggml-metal.m')
-rw-r--r-- | ggml-metal.m | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/ggml-metal.m b/ggml-metal.m index ac4f134..54cbaf8 100644 --- a/ggml-metal.m +++ b/ggml-metal.m @@ -526,7 +526,7 @@ void ggml_metal_graph_compute( GGML_ASSERT(ne12 == 1); nth0 = 8; - nth1 = 4; + nth1 = 8; [encoder setComputePipelineState:ctx->pipeline_mul_mat_q4_0_f32]; } break; case GGML_TYPE_Q2_K: |