| /aosp_15_r20/external/pytorch/torch/csrc/distributed/c10d/quantization/ |
| H A D | quantization_gpu.cu | 79 constexpr size_t threads_per_block = 256; in _float_to_bfloat16_cuda() local
126 constexpr size_t threads_per_block = 256; in _bfloat16_to_float_cuda() local
|
| /aosp_15_r20/external/pytorch/aten/src/ATen/native/cuda/ |
| H A D | PersistentSoftmax.cuh | 320 constexpr int threads_per_block = 128; in dispatch_softmax_forward() local
370 constexpr int threads_per_block = 128; in dispatch_softmax_backward() local
|
| H A D | Shape.cu | 56 constexpr unsigned int threads_per_block = 256; in getCatGridRocm() local
78 constexpr unsigned int threads_per_block = 128; in getCatGridContig() local
|
| H A D | SegmentReduce.cu | 299 constexpr int threads_per_block = 256; in _segment_reduce_lengths_offsets_backward_cuda_kernel() local
427 constexpr int threads_per_block = 256; in _segment_reduce_lengths_offsets_cuda_kernel() local
|
| H A D | EmbeddingBag.cu | 530 const int threads_per_block = 512; in _embedding_bag_per_sample_weights_backward_cuda() local
|
| /aosp_15_r20/external/tensorflow/tensorflow/stream_executor/rocm/ |
| H A D | rocm_helpers.cu.cc | 36 const int threads_per_block = 256; in rocm_MakeBatchPointers() local
|
| H A D | rocm_driver.cc | 1339 GpuContext* context, hipFunction_t kernel, int threads_per_block, in GetMaxOccupiedBlocksPerCore()
|
| /aosp_15_r20/external/tensorflow/tensorflow/core/kernels/image/ |
| H A D | adjust_saturation_op_gpu.cu.cc | 36 const int threads_per_block = config.thread_per_block; in operator ()() local
|
| H A D | adjust_hue_op_gpu.cu.cc | 35 const int threads_per_block = config.thread_per_block; in operator ()() local
|
| /aosp_15_r20/external/tensorflow/tensorflow/compiler/xla/stream_executor/ |
| H A D | device_description.cc | 143 int64_t element_count, int64_t *threads_per_block, in CalculateDimensionality()
|
| /aosp_15_r20/external/tensorflow/tensorflow/compiler/xla/service/gpu/ |
| H A D | launch_dimensions.cc | 39 int64_t threads_per_block = gpu_device_info.threads_per_block_limit; in ThreadsPerBlockLimit() local
|
| H A D | elemental_ir_emitter.cc | 329 llvm::Value* threads_per_block = IntCast( in EmitThreadId() local
|
| H A D | ir_emitter_unnested.cc | 3666 int threads_per_block, int num_results_per_warp) { in EmitFullWarpShuffleDownLoopForReduce()
4055 llvm::Value* IrEmitterUnnested::EmitThreadId(int64_t threads_per_block, in EmitThreadId()
|
| /aosp_15_r20/external/tensorflow/tensorflow/compiler/xla/stream_executor/gpu/ |
| H A D | redzone_allocator.cc | 224 int64_t threads_per_block = std::min( in RunRedzoneChecker() local
|
| /aosp_15_r20/external/mesa3d/src/gallium/drivers/radeonsi/ |
| H A D | si_get.c | 1242 unsigned threads_per_block = get_max_threads_per_block(sscreen, ir_type); in si_get_compute_param() local
|
| /aosp_15_r20/external/pytorch/c10/macros/ |
| H A D | Macros.h | 304 #define C10_MIN_BLOCKS_PER_SM(threads_per_block, blocks_per_sm) \ argument
|
| /aosp_15_r20/external/mesa3d/src/gallium/drivers/r600/ |
| H A D | r600_pipe_common.c | 932 unsigned threads_per_block = get_max_threads_per_block(rscreen, ir_type); in r600_get_compute_param() local
|
| /aosp_15_r20/external/tensorflow/tensorflow/compiler/xla/stream_executor/cuda/ |
| H A D | cuda_driver.cc | 1637 GpuContext* context, CUfunction kernel, int threads_per_block, in GetMaxOccupiedBlocksPerCore()
|