diff options
| author | Pablo Marquez Tello <pablo.tello@arm.com> | 2024-08-27 13:44:40 +0100 |
|---|---|---|
| committer | Pablo Marquez Tello <pablo.tello@arm.com> | 2024-08-28 08:22:16 +0000 |
| commit | e76048c33e588447347677959e4f2cb74e94b836 (patch) | |
| tree | 866b4dca79c3afffea9d2ade343148b859ed986d /src/cpu/kernels/maxunpool/list.h | |
| parent | 2bd38f38e76b2ba3dc8c221744387f81b93edde5 (diff) | |
| download | ComputeLibrary-release_candidate.tar.gz | |
perf: enable FP16 winograd in CpuConv2d for v8a multi_isa builds.rc_24_08_29_0release_candidate
* Replaced __ARM_FEATURE_FP16_VECTOR_ARITHMETIC by ARM_COMPUTE_ENABLE_FP16 in CpuConv2d to enable winograd convolution method on multi_isa builds
* Improved NEDeconvolutionLayer performance on cpus that support FP16 when using fast_math
* Resolves ARMCL-1160
Change-Id: I63e48c22f4fc226908a1b145c0298d9f43ae96ea
Signed-off-by: Pablo Marquez Tello <pablo.tello@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/12234
Reviewed-by: Gunes Bayir <gunes.bayir@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/cpu/kernels/maxunpool/list.h')
0 files changed, 0 insertions, 0 deletions