millaguie/llama.cpp
1
0 Fork 0
Code Issues Pull Requests Actions 30 Packages Projects Releases Wiki Activity

CUDA: revert part of the RDNA1 optimizations (#8309)

Browse Source 
The change on the launch_bounds was causing a small performance drop in perplexity of 25 t/s
This commit is contained in:
Daniele authored and GitHub committed 2024-07-05 09:06:09 +02:00
1 parent d12f781074
commit 0a423800ff
1 file changed
+2 -2
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-cuda/mmq.cuh
+2 -2
View File
@@ -2263,9 +2263,9 @@ static __device__ void mul_mat_q_process_tile(
template <ggml_type type, int mmq_x, int nwarps, bool need_check>
#if defined(GGML_USE_HIPBLAS) && defined(__HIP_PLATFORM_AMD__)
#if defined(RDNA3) || defined(RDNA2) || defined(RDNA1)
#if defined(RDNA3) || defined(RDNA2)
__launch_bounds__(WARP_SIZE*nwarps, 2)
#endif // defined(RDNA3) || defined(RDNA2) || defined(RDNA1)
#endif // defined(RDNA3) || defined(RDNA2)
#else
#if __CUDA_ARCH__ >= CC_VOLTA
__launch_bounds__(WARP_SIZE*nwarps, 1)