From 4b90865ab985d571f70c60583cdfb8c7a65f1670 Mon Sep 17 00:00:00 2001 From: Gian Marco Iodice Date: Thu, 18 Oct 2018 10:21:02 +0100 Subject: COMPMID-1413 - Improve the performance of GEMMLowp with 8 bit dot product on OpenCL COMPMID-1424 - Add dot product support for CLDepthwise QASYMM8 3x3 NHWC non-unit stride With this patch we are able to improve the performance of MobileNet v1-qasymm8 by 37 % Tried to use the dot product instruction in CLDepthwise QASYMM8 3x3 NHWC non-unit stride but I have not seen any benefit (maybe because we have few arithemtic operation and we do not have more load instructions). However Depthwise convolution has been improved by 30% Change-Id: Id768a99c2e53a04276707e427af5d0ec93419ada Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/155082 Tested-by: bsgcomp Reviewed-by: Georgios Pinitas --- src/runtime/NEON/functions/NEFullyConnectedLayer.cpp | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'src/runtime/NEON/functions/NEFullyConnectedLayer.cpp') diff --git a/src/runtime/NEON/functions/NEFullyConnectedLayer.cpp b/src/runtime/NEON/functions/NEFullyConnectedLayer.cpp index 60f6294394..45e21b53d1 100644 --- a/src/runtime/NEON/functions/NEFullyConnectedLayer.cpp +++ b/src/runtime/NEON/functions/NEFullyConnectedLayer.cpp @@ -50,6 +50,7 @@ Status validate_mm(const ITensorInfo &input, const ITensorInfo &weights, const I // Validate gemmlowp function ARM_COMPUTE_RETURN_ON_ERROR(NEGEMMLowpMatrixMultiplyCore::validate(&input.clone()->set_quantization_info(input_quantization_info), &weights.clone()->set_quantization_info(weights_quantization_info), + nullptr, &output)); } else @@ -93,7 +94,7 @@ void NEFullyConnectedLayer::configure_mm(const ITensor *input, const ITensor *we weights->info()->set_quantization_info(QuantizationInfo(weights_quantization_info.scale, -weights_quantization_info.offset)); // Configure gemmlowp function - _mm_gemmlowp.configure(input, weights, output); + _mm_gemmlowp.configure(input, weights, nullptr, output); // Revert back QuantizatioInfo as input and weights could be used in other fully connected layers input->info()->set_quantization_info(input_quantization_info); -- cgit v1.2.1