Home
last modified time | relevance | path

Searched defs:total_threads (Results 1 – 9 of 9) sorted by relevance

/third_party/mindspore/mindspore-src/source/mindspore/lite/src/extendrt/delegate/tensorrt/cuda_impl/
Dcuda_helper.h42 #define GET_BLOCKS(total_threads) CudaHelper::GetInstance().GetBlocksNum(total_threads) argument
44 #define GET_BLOCKS_CAL(total_threads, block_size) CudaHelper::GetInstance().GetBlocksNum(total_thre… argument
Dcuda_helper.cc32 int CudaHelper::GetBlocksNum(const int total_threads, const int block_size) const { in GetBlocksNum() argument
/third_party/mindspore/mindspore-src/source/mindspore/ccsrc/plugin/device/gpu/kernel/cuda_impl/cuda_ops/
Dcuda_device_info.h74 inline size_t blocks_num(const size_t total_threads) const { in blocks_num()
77 inline size_t blocks_num(const size_t total_threads, const size_t block_size) const { in blocks_num()
109 #define CUDA_BLOCKS(device_id, total_threads) \ argument
111 #define CUDA_BLOCKS_CAL(device_id, total_threads, block_size) \ argument
Dcuda_common.h32 …inline size_t blocks_num(const size_t total_threads) const { return CUDA_BLOCKS(device_id_, total_… in blocks_num()
50 #define GET_BLOCKS(total_threads) mindspore::device::gpu::CudaCommon::GetInstance().blocks_num(tota… argument
Dno_repeat_ngram_impl.cu73 int total_threads) { in reassign_probability_no_shared()
/third_party/mindspore/mindspore-src/source/mindspore/ccsrc/plugin/device/gpu/kernel/arrays/
Dreverse_sequence_gpu_kernel.cc66 size_t total_threads = GET_BLOCKS(input_size_) * GET_THREADS; in Resize() local
/third_party/mindspore/mindspore-src/source/mindspore/ccsrc/plugin/device/gpu/kernel/akg/
Dakg_gpu_kernel_mod.cc58 int total_threads = thread_info->at(3) * thread_info->at(4) * thread_info->at(5); in GetFunction() local
/third_party/mindspore/mindspore-src/source/mindspore/lite/src/extendrt/delegate/tensorrt/op/
Dakg_tensorrt.cc163 int total_threads = params_.tx * params_.ty * params_.tz; in GetFunction() local
/third_party/mesa3d/src/gallium/frontends/rusticl/core/
Dkernel.rs1243 let total_threads = block.iter().take(work_dim).product::<usize>(); in suggest_local_size() localVariable