diff options
author | Gian Marco <gianmarco.iodice@arm.com> | 2018-02-01 16:57:48 +0000 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:45:42 +0000 |
commit | c799ed85bcf0b18269e65a64f34382073a68bd93 (patch) | |
tree | bda7b56321638968a24032ff4e685a2a1d58acd7 /src/core/CL/CLKernelLibrary.cpp | |
parent | 604afc059c546b3c060a814f15798019cfb7a3de (diff) | |
download | ComputeLibrary-c799ed85bcf0b18269e65a64f34382073a68bd93.tar.gz |
COMPMID-895 - Optimizing CLDepthwiseConvolution3x3Kernel
This patch brings the MACs utilisation up to 25 % when both stride_x and stride_y are equal to 1
Performance reported in the following confluence page:
https://confluence.arm.com/display/MLENG/Depthwise+convolution+3x3+FP32+performance%3A+ACL+18.02
Change-Id: Ida1b64be9a88805902a3d90194559b58eb1224a3
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/119068
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/core/CL/CLKernelLibrary.cpp')
-rw-r--r-- | src/core/CL/CLKernelLibrary.cpp | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/src/core/CL/CLKernelLibrary.cpp b/src/core/CL/CLKernelLibrary.cpp index d7ae6e3127..c26d8d80a6 100644 --- a/src/core/CL/CLKernelLibrary.cpp +++ b/src/core/CL/CLKernelLibrary.cpp @@ -189,6 +189,8 @@ const std::map<std::string, std::string> CLKernelLibrary::_kernel_program_map = { "deconvolution_upsample", "deconvolution_layer.cl" }, { "depthwise_convolution_3x3", "depthwise_convolution.cl" }, { "depthwise_convolution_3x3_quantized", "depthwise_convolution_quantized.cl" }, + { "depthwise_convolution_3x3_stridex1_stridey1_bifrost", "depthwise_convolution.cl" }, + { "depthwise_convolution_3x3_stridex2_stridey2_bifrost", "depthwise_convolution.cl" }, { "depthwise_im2col", "depthwise_convolution.cl" }, { "depthwise_vector_to_tensor", "depthwise_convolution.cl" }, { "depthwise_weights_reshape", "depthwise_convolution.cl" }, |