diff options
Diffstat (limited to 'src/core/NEON/kernels/arm_conv/depthwise/depthwise_fp32.cpp')
| -rw-r--r-- | src/core/NEON/kernels/arm_conv/depthwise/depthwise_fp32.cpp | 14 |
1 files changed, 13 insertions, 1 deletions
diff --git a/src/core/NEON/kernels/arm_conv/depthwise/depthwise_fp32.cpp b/src/core/NEON/kernels/arm_conv/depthwise/depthwise_fp32.cpp index 09ee983907..382ccd3c62 100644 --- a/src/core/NEON/kernels/arm_conv/depthwise/depthwise_fp32.cpp +++ b/src/core/NEON/kernels/arm_conv/depthwise/depthwise_fp32.cpp @@ -1,5 +1,5 @@ /* - * Copyright (c) 2021-2022 Arm Limited. + * Copyright (c) 2021-2023 Arm Limited. * * SPDX-License-Identifier: MIT * @@ -103,6 +103,18 @@ namespace ); } + template <class Strategy> + unsigned int fast_mode_cycle_estimate(const DepthwiseArgs &args, const Nothing &) + { + // First-pass: compute the number of output pixels which will be computed. + return arm_gemm::roundup(args.output_rows, Strategy::output_rows) * + arm_gemm::roundup(args.output_cols, Strategy::output_cols) * + arm_gemm::iceildiv( + (long unsigned) args.input_channels * args.channel_multiplier, + arm_gemm::utils::get_vector_length<typename Strategy::return_type>(Strategy::vl_type) + ) * 2 / 3; + } + #if defined(__aarch64__) unsigned int not_preferred(const DepthwiseArgs &, const Nothing &) { |