vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

CUDA: switch tile sizes based on binary version

Browse source
This commit is contained in:
JohannesGaessler 2024-02-17 20:09:13 +01:00
parent d250c9d61d
commit 44a80b4119
1 changed files with 13 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

16
ggml-cuda.cu
View file

@ -6941,11 +6941,12 @@ static void ggml_mul_mat_q4_0_q8_1_cuda(
const void * vx, const void * vy, float * dst, const int ncols_x, const int nrows_x, const void * vx, const void * vy, float * dst, const int ncols_x, const int nrows_x,
const int ncols_y, const int nrows_y, const int nrows_dst, cudaStream_t stream) { const int ncols_y, const int nrows_y, const int nrows_dst, cudaStream_t stream) {
int mmq_x, mmq_y, nwarps;
#if defined(GGML_USE_HIPBLAS) && defined(__HIP_PLATFORM_AMD__)
int id; int id;
CUDA_CHECK(cudaGetDevice(&id)); CUDA_CHECK(cudaGetDevice(&id));
const int compute_capability = g_device_caps[id].cc; const int compute_capability = g_device_caps[id].cc;
int mmq_x, mmq_y, nwarps;
if (compute_capability >= CC_RDNA2) { if (compute_capability >= CC_RDNA2) {
mmq_x = MMQ_X_Q4_0_RDNA2; mmq_x = MMQ_X_Q4_0_RDNA2;
mmq_y = MMQ_Y_Q4_0_RDNA2; mmq_y = MMQ_Y_Q4_0_RDNA2;
@ -6954,17 +6955,26 @@ static void ggml_mul_mat_q4_0_q8_1_cuda(
mmq_x = MMQ_X_Q4_0_RDNA1; mmq_x = MMQ_X_Q4_0_RDNA1;
mmq_y = MMQ_Y_Q4_0_RDNA1; mmq_y = MMQ_Y_Q4_0_RDNA1;
nwarps = NWARPS_Q4_0_RDNA1; nwarps = NWARPS_Q4_0_RDNA1;
} else if (compute_capability >= CC_VOLTA) { } else {
GGML_ASSERT(false);
}
#else
cudaFuncAttributes attributes;
CUDA_CHECK(cudaFuncGetAttributes(&attributes, mul_mat_q4_0<false>));
const int cc_binary = 10*attributes.binaryVersion;
if (cc_binary >= CC_VOLTA) {
mmq_x = MMQ_X_Q4_0_AMPERE; mmq_x = MMQ_X_Q4_0_AMPERE;
mmq_y = MMQ_Y_Q4_0_AMPERE; mmq_y = MMQ_Y_Q4_0_AMPERE;
nwarps = NWARPS_Q4_0_AMPERE; nwarps = NWARPS_Q4_0_AMPERE;
} else if (compute_capability >= MIN_CC_DP4A) { } else if (cc_binary >= MIN_CC_DP4A) {
mmq_x = MMQ_X_Q4_0_PASCAL; mmq_x = MMQ_X_Q4_0_PASCAL;
mmq_y = MMQ_Y_Q4_0_PASCAL; mmq_y = MMQ_Y_Q4_0_PASCAL;
nwarps = NWARPS_Q4_0_PASCAL; nwarps = NWARPS_Q4_0_PASCAL;
} else { } else {
GGML_ASSERT(false); GGML_ASSERT(false);
} }
#endif // defined(GGML_USE_HIPBLAS) && defined(__HIP_PLATFORM_AMD__)
const int block_num_x = (nrows_x + mmq_y - 1) / mmq_y; const int block_num_x = (nrows_x + mmq_y - 1) / mmq_y;
const int block_num_y = (ncols_y + mmq_x - 1) / mmq_x; const int block_num_y = (ncols_y + mmq_x - 1) / mmq_x;