GPU Kernel Information Aggregated by Name
| kernel_name | kernel_count | kernel_duration (us) | model_duration_percentage | kernel_flops | kernel_dram_read_bytes | kernel_dram_write_bytes | kernel_achieved_occupancy (%) | kernel_arithmetic_intensity (flops/byte) | kernel_arithmetic_throughput (GFlops) | kernel_memory_bound |
|---|
kernel_name | kernel_count | kernel_duration (us) | model_duration_percentage | kernel_flops | kernel_dram_read_bytes | kernel_dram_write_bytes | kernel_achieved_occupancy (%) | kernel_arithmetic_intensity (flops/byte) | kernel_arithmetic_throughput (GFlops) | kernel_memory_bound |
|---|---|---|---|---|---|---|---|---|---|---|
| cudnn::maxwell::gemm::computeOffsetsKernel(cudnn::maxwell::gemm::ComputeOffsetsParams) | 123 | 257.20 | 0.32 | 0 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | true |
| maxwell_scudnn_128x128_relu_interior_nn | 7 | 519.33 | 0.65 | 229834752 | 0.00 | 0.00 | 0.00 | 0.00 | 442.56 | true |
| maxwell_scudnn_128x32_relu_interior_nn | 0 | 11.67 | 0.01 | 13647872 | 0.00 | 0.00 | 0.00 | 0.00 | 1169.78 | true |
| maxwell_scudnn_128x32_relu_small_nn | 112 | 1135.21 | 1.42 | 323174400 | 0.00 | 0.00 | 0.00 | 0.00 | 284.68 | true |
| maxwell_scudnn_128x64_relu_interior_nn | 1 | 43.48 | 0.05 | 40140800 | 0.00 | 0.00 | 0.00 | 0.00 | 923.29 | true |
| maxwell_scudnn_winograd_128x128_ldg1_ldg4_tile148t_nt | 1983 | 19840.00 | 24.86 | 1832910848 | 0.00 | 0.00 | 0.00 | 0.00 | 92.38 | true |
| void cudnn::detail::bn_fw_inf_1C11_kernel_NCHW<float, float, true, 1>(float, float, cudnnTensorStruct, float const*, cudnnTensorStruct, float*, cudnnTensorStruct, float const*, float const*, float const*, float const*, float) | 26 | 129.37 | 0.16 | 15666432 | 0.00 | 0.00 | 0.00 | 0.00 | 121.10 | true |
| void cudnn::detail::explicit_convolve_sgemm<float, int, 1024, 5, 5, 3, 3, 3, 0, true>(int, int, int, float const*, int, float const*, int, float*, kernel_conv_params, int, int, float, float, int, float*, float*) | 385 | 2473.89 | 3.10 | 96556800 | 0.00 | 0.00 | 0.00 | 0.00 | 39.03 | true |
| void cudnn::detail::pooling_fw_4d_kernel<float, float, cudnn::detail::averpooling_func<float>, 1, false>(cudnnTensorStruct, float const*, cudnnTensorStruct, float*, cudnnPoolingStruct, float, float, int, cudnn::reduced_divisor, cudnn::reduced_divisor) | 0 | 10.75 | 0.01 | 32287 | 0.00 | 0.00 | 0.00 | 0.00 | 3.00 | true |
| void cudnn::winograd::generateWinogradTilesKernel<0, float, float>(cudnn::winograd::GenerateWinogradTilesParams<float, float>) | 1983 | 7936.00 | 9.94 | 29458432 | 0.00 | 0.00 | 0.00 | 0.00 | 3.71 | true |
| void gemv2T_kernel_val<int, int, float, float, float, 128, 16, 2, 2, false, cublasGemvParams<cublasGemvTensor<float const>, cublasGemvTensor<float>, float> >(cublasGemvParams<cublasGemvTensor<float const>, cublasGemvTensor<float>, float>, float, float) | 0 | 10.00 | 0.01 | 1135000 | 0.00 | 0.00 | 0.00 | 0.00 | 113.50 | true |
| void im2col4d_kernel<float, int>(im2col4d_params, cudnnConvolutionStruct, cudnnTensor4dStruct, float const*, float*, int) | 385 | 2130.00 | 2.67 | 0 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | true |
| void mshadow::cuda::MapPlanKernel<mshadow::sv::plusto, 8, mshadow::expr::Plan<mshadow::Tensor<mshadow::gpu, 2, float>, float>, mshadow::expr::Plan<mshadow::expr::Broadcast1DExp<mshadow::Tensor<mshadow::gpu, 1, float>, float, 2, 1>, float> >(mshadow::expr::Plan<mshadow::Tensor<mshadow::gpu, 2, float>, float>, int, mshadow::Shape<2>, mshadow::expr::Plan<mshadow::expr::Broadcast1DExp<mshadow::Tensor<mshadow::gpu, 1, float>, float, 2, 1>, float>) | 0 | 3.00 | 0.00 | 1000 | 0.00 | 0.00 | 0.00 | 0.00 | 0.33 | true |
| void op_generic_tensor_kernel<2, float, float, float, 256, (cudnnGenericOp_t)8, (cudnnNanPropagation_t)0, (cudnnDimOrder_t)0, 1>(cudnnTensorStruct, float*, cudnnTensorStruct, float const*, cudnnTensorStruct, float const*, float, float, float, float, dimArray, reducedDivisorArray, bool) | 26 | 100.43 | 0.13 | 5042688 | 0.00 | 0.00 | 0.00 | 0.00 | 50.21 | true |
Showing 1 to 14 of 14 entries