{"id":18743,"date":"2019-02-10T11:49:05","date_gmt":"2019-02-10T09:49:05","guid":{"rendered":"https:\/\/hgpu.org\/?p=18743"},"modified":"2019-02-10T11:49:05","modified_gmt":"2019-02-10T09:49:05","slug":"optimising-convolutional-neural-networks-inference-on-low-powered-gpus","status":"publish","type":"post","link":"https:\/\/hgpu.org\/?p=18743","title":{"rendered":"Optimising Convolutional Neural Networks Inference on Low-Powered GPUs"},"content":{"rendered":"<p>In this paper we present effective optimisation techniques for accelerating convolutional neural networks inference on low-powered heterogeneous devices with OpenCL. Using LeNet and VGG-16 as test networks, we implement a custom neural network system in OpenCL and optimise it to minimise their inference times. Our baseline system shows a speedup of 17x for LeNet. We also outline two methods for fast convolution: an iterative vectorised approach and a Morton GEMM based approach. The two approaches demonstrate VGG-16 inference speeds up to 3x faster than current state-of-the-art systems and outperform other custom neural network systems by speedup factors of up to 1.82x.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>In this paper we present effective optimisation techniques for accelerating convolutional neural networks inference on low-powered heterogeneous devices with OpenCL. Using LeNet and VGG-16 as test networks, we implement a custom neural network system in OpenCL and optimise it to minimise their inference times. Our baseline system shows a speedup of 17x for LeNet. We [&hellip;]<\/p>\n","protected":false},"author":351,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2}},"categories":[11,90,3],"tags":[1782,1673,452,34,1793],"class_list":["post-18743","post","type-post","status-publish","format-standard","hentry","category-computer-science","category-opencl","category-paper","tag-computer-science","tag-deep-learning","tag-heterogeneous-systems","tag-neural-networks","tag-opencl"],"views":2543,"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts\/18743","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/users\/351"}],"replies":[{"embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=18743"}],"version-history":[{"count":0,"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts\/18743\/revisions"}],"wp:attachment":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=18743"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=18743"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=18743"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}