commit ec0113dd7749991959ae351934eea0c0d8077dcb
author Gunes Bayir <gunes.bayir@arm.com> Wed Nov 09 09:26:27 2022 +0000
committer Gunes Bayir <gunes.bayir@arm.com> Mon Nov 14 15:01:27 2022 +0000
tree 792f38b3a05fa85436339d54abddab096e20cc77
parent 46ede332c7064b85dc538891cad165b3bb427320

Optimize Transposed Convolution for CL backend (FP32/16)

This patch optimizes transposed convolution for CL backend by rewriting it in a single kernel instead of three (flip_kernel + upsample + conv). The new kernel skips the upsampling step which reduces the input space of convolution by stride_x * stride_y, resulting in significant performance improvement. It also skips the kernel flipping by traversing the weights accordingly, thus reduces the memory footprint.

Resolves: COMPMID-5676
Signed-off-by: Gunes Bayir <gunes.bayir@arm.com>
Change-Id: I8a333212dc7c5f7f0597aa58b0d56d44814baa14
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/8588
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>

SConscript

diff --git a/SConscript b/SConscript
index c05c1fe..42a03f0 100644
--- a/SConscript
+++ b/SConscript
@@ -460,6 +460,7 @@
                     'src/core/CL/cl_kernels/nhwc/scale.cl',
                     'src/core/CL/cl_kernels/nhwc/space_to_batch.cl',
                     'src/core/CL/cl_kernels/nhwc/space_to_depth.cl',
+                    'src/core/CL/cl_kernels/nhwc/transposed_convolution.cl',
                     'src/core/CL/cl_kernels/nhwc/upsample_layer.cl',
                     'src/core/CL/cl_kernels/nhwc/winograd_filter_transform.cl',
                     'src/core/CL/cl_kernels/nhwc/winograd_input_transform.cl',