Home
last modified time | relevance | path

Searched refs:maxThreadsDim (Results 1 – 5 of 5) sorted by relevance

/external/pytorch/aten/src/ATen/native/cuda/
DAdaptiveAveragePooling.cu492 int* maxThreadsDim = at::cuda::getCurrentDeviceProperties()->maxThreadsDim; in adaptive_avg_pool2d_out_cuda_template() local
506 maxThreadsDim[0], std::min<int>(lastPow2(sizeC), at::cuda::warp_size())); in adaptive_avg_pool2d_out_cuda_template()
508 maxThreadsDim[1], std::min<int>(lastPow2(osizeW), max_threads / block_x)); in adaptive_avg_pool2d_out_cuda_template()
510 maxThreadsDim[2], std::min<int>(lastPow2(osizeH), max_threads / block_x / block_y)); in adaptive_avg_pool2d_out_cuda_template()
512 maxThreadsDim[0], std::min<int>(lastPow2(sizeC), max_threads / block_y / block_z)); in adaptive_avg_pool2d_out_cuda_template()
644 int* maxThreadsDim = at::cuda::getCurrentDeviceProperties()->maxThreadsDim; in adaptive_avg_pool2d_backward_out_cuda_template() local
660 maxThreadsDim[0], std::min<int>(lastPow2(sizeC), at::cuda::warp_size())), 1); in adaptive_avg_pool2d_backward_out_cuda_template()
662 maxThreadsDim[1], std::min<int>(lastPow2(isizeW), max_threads / block_x)), 1); in adaptive_avg_pool2d_backward_out_cuda_template()
664 … maxThreadsDim[2], std::min<int>(lastPow2(isizeH), max_threads / block_x / block_y)), 1); in adaptive_avg_pool2d_backward_out_cuda_template()
666 … maxThreadsDim[0], std::min<int>(lastPow2(sizeC), max_threads / block_y / block_z)), 1); in adaptive_avg_pool2d_backward_out_cuda_template()
DDilatedMaxPool2d.cu359 int* maxThreadsDim = at::cuda::getCurrentDeviceProperties()->maxThreadsDim; in TORCH_IMPL_FUNC() local
361 maxThreadsDim[0], std::min<int>(lastPow2(nInputPlane), at::cuda::warp_size())); in TORCH_IMPL_FUNC()
363 maxThreadsDim[1], std::min<int>(lastPow2(outputWidth), max_threads / block_x)); in TORCH_IMPL_FUNC()
365 … maxThreadsDim[2], std::min<int>(lastPow2(outputHeight), max_threads / block_x / block_y)); in TORCH_IMPL_FUNC()
367 … maxThreadsDim[0], std::min<int>(lastPow2(nInputPlane), max_threads / block_y / block_z)); in TORCH_IMPL_FUNC()
494 int* maxThreadsDim = at::cuda::getCurrentDeviceProperties()->maxThreadsDim; in TORCH_IMPL_FUNC() local
496 maxThreadsDim[0], std::min<int>(lastPow2(nInputPlane), at::cuda::warp_size())); in TORCH_IMPL_FUNC()
498 maxThreadsDim[1], std::min<int>(lastPow2(inputWidth), max_threads / block_x)); in TORCH_IMPL_FUNC()
500 … maxThreadsDim[2], std::min<int>(lastPow2(inputHeight), max_threads / block_x / block_y)); in TORCH_IMPL_FUNC()
502 … maxThreadsDim[0], std::min<int>(lastPow2(nInputPlane), max_threads / block_y / block_z)); in TORCH_IMPL_FUNC()
DUpSampleNearest2d.cu280 int* maxThreadsDim = at::cuda::getCurrentDeviceProperties()->maxThreadsDim; in upsample_nearest2d_out_cuda_template() local
285 maxThreadsDim[0], std::min<int>(lastPow2(output_width), max_threads)); in upsample_nearest2d_out_cuda_template()
287 maxThreadsDim[1], in upsample_nearest2d_out_cuda_template()
290 maxThreadsDim[2], std::min<int>(nc, max_threads / block_x / block_y)); in upsample_nearest2d_out_cuda_template()
DUpSampleBilinear2d.cu698 int* maxThreadsDim = at::cuda::getCurrentDeviceProperties()->maxThreadsDim; in upsample_gen2d_aa_out_cuda_template() local
701 int block_x = std::min<int>(maxThreadsDim[0], at::cuda::warp_size()); in upsample_gen2d_aa_out_cuda_template()
797 int* maxThreadsDim = at::cuda::getCurrentDeviceProperties()->maxThreadsDim; in upsample_gen2d_aa_backward_out_cuda_template() local
798 int block_x = std::min<int>(maxThreadsDim[0], at::cuda::warp_size()); in upsample_gen2d_aa_backward_out_cuda_template()
799 int block_y = std::min<int>(maxThreadsDim[1], num_threads / block_x); in upsample_gen2d_aa_backward_out_cuda_template()
/external/tensorflow/tensorflow/stream_executor/rocm/
Drocm_gpu_executor.cc911 thread_dim_limit.x = prop.maxThreadsDim[0]; in CreateDeviceDescription()
912 thread_dim_limit.y = prop.maxThreadsDim[1]; in CreateDeviceDescription()
913 thread_dim_limit.z = prop.maxThreadsDim[2]; in CreateDeviceDescription()