diff options
author | Michele Di Giorgio <michele.digiorgio@arm.com> | 2021-01-22 09:47:04 +0000 |
---|---|---|
committer | Michele Di Giorgio <michele.digiorgio@arm.com> | 2021-06-18 10:33:48 +0000 |
commit | d02d5edfa15ba6c04a9986a8a362a945cb38ac31 (patch) | |
tree | ced4f49691d6c7038e347a8709b315bff59c64cf /examples/gemm_tuner | |
parent | b014c27ba6db9840e4a72519760d51a87a2af7e7 (diff) | |
download | ComputeLibrary-d02d5edfa15ba6c04a9986a8a362a945cb38ac31.tar.gz |
Integrate improved CPU depthwise convolution kernels
* Replace assembly kernels for depthwise convolution with more optimized
ones.
* Add int8 assembly kernels.
* Fix implicit padding on optimized kernels
Resolves: COMPMID-3867, COMPMID-4361
Change-Id: I0b0867e05f61be4f368f62190d55e14d0ab3ebf2
Signed-off-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5622
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Diffstat (limited to 'examples/gemm_tuner')
-rw-r--r-- | examples/gemm_tuner/cl_gemmlowp_reshaped_rhs_only_fused_output_stage_fixedpoint.cpp | 2 |
1 files changed, 0 insertions, 2 deletions
diff --git a/examples/gemm_tuner/cl_gemmlowp_reshaped_rhs_only_fused_output_stage_fixedpoint.cpp b/examples/gemm_tuner/cl_gemmlowp_reshaped_rhs_only_fused_output_stage_fixedpoint.cpp index d8f8f1498a..ca7b7a5f04 100644 --- a/examples/gemm_tuner/cl_gemmlowp_reshaped_rhs_only_fused_output_stage_fixedpoint.cpp +++ b/examples/gemm_tuner/cl_gemmlowp_reshaped_rhs_only_fused_output_stage_fixedpoint.cpp @@ -236,7 +236,6 @@ public: gemmlowp_output_stage.output_data_type = dst.info()->data_type(); gemmlowp_output_stage.gemmlowp_offset = 0; { - const int idx_kernels = get_data_layout_dimension_index(lhs.info()->data_layout(), DataLayoutDimension::BATCHES); gemmlowp_output_stage.is_quantized_per_channel = false; // Num_filters is 1 unless quantized type is of per_channel type. Could be extended in the future to support per-channel quantization. const unsigned int num_filters = 1; @@ -249,7 +248,6 @@ public: quantization::compute_quantized_multipliers_and_shifts(lhs.info(), rhs.info(), dst.info(), - idx_kernels, gemmlowp_output_stage.gemmlowp_multipliers.data(), gemmlowp_output_stage.gemmlowp_shifts.data()); gemmlowp_output_stage.gemmlowp_multiplier = gemmlowp_output_stage.gemmlowp_multipliers[0]; |