{"id":543,"date":"2020-12-17T08:37:53","date_gmt":"2020-12-17T08:37:53","guid":{"rendered":"https:\/\/webroad-seisaku.com\/fujiko-test\/?p=543"},"modified":"2023-07-13T02:56:55","modified_gmt":"2023-07-13T02:56:55","slug":"demonstration-of-kudan-slam-with-40-image-processing-acceleration-by-synopsys-arc-ev-processor-ip","status":"publish","type":"post","link":"https:\/\/webroad-seisaku.com\/fujiko-test\/archives\/543","title":{"rendered":"Demonstration of Kudan SLAM with 40% image processing acceleration by Synopsys ARC EV Processor IP"},"content":{"rendered":"<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-408\" src=\"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/07\/Screenshot-2019-02-26-at-09.32.30.png\" alt=\"\" width=\"2754\" height=\"392\" srcset=\"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/07\/Screenshot-2019-02-26-at-09.32.30.png 2754w, https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/07\/Screenshot-2019-02-26-at-09.32.30-300x43.png 300w, https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/07\/Screenshot-2019-02-26-at-09.32.30-1024x146.png 1024w, https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/07\/Screenshot-2019-02-26-at-09.32.30-768x109.png 768w, https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/07\/Screenshot-2019-02-26-at-09.32.30-1536x219.png 1536w, https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/07\/Screenshot-2019-02-26-at-09.32.30-2048x292.png 2048w\" sizes=\"auto, (max-width: 2754px) 100vw, 2754px\" \/><\/p>\n<p>Running SLAM on a system still requires a lot of CPU cycles and memory, and while Kudan has been bringing down these system resource requirements, there is always a demand for more CPU and memory to be used by other applications on the system, especially as we continue to support lower-powered devices.<\/p>\n<p>This is where hardware acceleration really helps boost the capabilities of the system, and what we have been working to make a reality. As you may have read from past <a href=\"https:\/\/news.synopsys.com\/2019-05-20-Synopsys-and-Kudan-Collaborate-to-Accelerate-Development-of-Intelligent-Computer-Vision-Processing-SoCs\" target=\"_blank\" rel=\"noopener noreferrer\">press releases<\/a>, we have been working with Synopsys on accelerating Kudan\u2019s visual SLAM on the Synopsys <a href=\"https:\/\/www.synopsys.com\/designware-ip\/processor-solutions\/ev-processors.html\" target=\"_blank\" rel=\"noopener noreferrer\">DesignWare\u00ae ARC\u00ae EV6x Embedded Vision Processor IP<\/a>.<\/p>\n<p>It\u2019s time to unwrap some of the details of the collaboration and highlights. The processing of the raw video stream uses a significant proportion of resources in the visual SLAM pipeline. Being able to offload and accelerate this frontend part of the pipeline frees up valuable CPU resources on a system, and ultimately accelerates the SLAM process significantly.<\/p>\n<p>The demonstrations in the presentation below cover both indoor and outdoor use cases, as well as applications in automotive, consumer and industrial robotics, mobile and XR (AR\/VR). Based on this successful demonstration, we are planning to accelerate our collaboration across multiple applications.<\/p>\n<p><strong>Highlights<\/strong><\/p>\n<p>\u25cf40% performance improvement in the image processing pipeline, which accounts for roughly 50% of the entire SLAM process time<br \/>\n\u25cfAcceleration of frontend image processing on the ARC EV Processor, and application with SLAM running on a host, with support for multiple processor host, including x86, ARM, and RISC as needed for the target application<br \/>\n\u25cfEnables embedded systems using the ARC EV platform to run accurate visual SLAM in real-time for a wide range of applications<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"size-full wp-image-547 aligncenter\" src=\"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/12\/Pic1.png\" alt=\"\" width=\"1164\" height=\"646\" srcset=\"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/12\/Pic1.png 1164w, https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/12\/Pic1-300x166.png 300w, https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/12\/Pic1-1024x568.png 1024w, https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/12\/Pic1-768x426.png 768w\" sizes=\"auto, (max-width: 1164px) 100vw, 1164px\" \/><\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-548\" src=\"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/12\/Pic2.png\" alt=\"\" width=\"1162\" height=\"646\" srcset=\"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/12\/Pic2.png 1162w, https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/12\/Pic2-300x167.png 300w, https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/12\/Pic2-1024x569.png 1024w, https:\/\/webroad-seisaku.com\/fujiko-test\/wp-content\/uploads\/2020\/12\/Pic2-768x427.png 768w\" sizes=\"auto, (max-width: 1162px) 100vw, 1162px\" \/><\/p>\n<p>Watch Liliya Tazieva from Synopsys give a great summary of the collaboration and demonstration of Kudan SLAM accelerated through the ARC EV6x platform.<\/p>\n<p><iframe loading=\"lazy\" title=\"Demo: Combining SLAM and Object Detection with DesignWare ARC EV Processor IP | Synopsys\" width=\"1140\" height=\"641\" src=\"https:\/\/www.youtube.com\/embed\/9MmYy9auAmg?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe><\/p>\n<p><strong>About Synopsys<\/strong><br \/>\nSynopsys, Inc. (Nasdaq: SNPS) is the Silicon to Software\u2122 partner for innovative companies developing the electronic products and software applications we rely on every day. As the world\u2019s 15th largest software company, Synopsys has a long history of being a global leader in electronic design automation (EDA) and semiconductor IP and is also growing its leadership in software security and quality solutions. Whether you\u2019re a system-on-chip (SoC) designer creating advanced semiconductors, or a software developer writing applications that require the highest security and quality, Synopsys has the solutions needed to deliver innovative, high-quality, secure products.<br \/>\nLearn more at\u00a0<a href=\"http:\/\/www.synopsys.com\/\" target=\"_blank\" rel=\"noopener noreferrer\">www.synopsys.com<\/a>.<\/p>\n<p><strong>About Kudan<\/strong><br \/>\nKudan (TYO: 4425) is a leading Deep Tech firm developing Artificial Perception (AP) algorithms, providing the eyes for the machines (robots, computers), complementary working with Artificial Intelligence (AI) as the brain for the machines. On top of its proprietary Artificial Perception, Kudan is developing integration of AP with AI and IoT, based on its unique milestone approach to measure the maturity of technology and development specialised for Deep Tech. As a leading Deep Tech firm, Kudan spans across many technological areas and impacts diverse applications through its tangible engineering innovations.<br \/>\nLearn more at\u00a0<a href=\"https:\/\/webroad-seisaku.com\/fujiko-test\/\" target=\"_blank\" rel=\"noopener noreferrer\">https:\/\/webroad-seisaku.com\/fujiko-test\/<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Running SLAM on a system still requires a lot of CPU cycles and memory, and while Kudan has been bringing down these system resource requirements, there is always a demand for more CPU and memory to be used by other applications on the system, especially as we continue to support lower-powered devices. This is where [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":409,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"sns_share_botton_hide":"","vkExUnit_sns_title":"","_vk_print_noindex":"","sitemap_hide":"","_veu_custom_css":"","veu_display_promotion_alert":"","vkexunit_cta_each_option":"","_lightning_design_setting":[],"header_top_description":"","footnotes":""},"categories":[2],"tags":[36,5,37,35],"class_list":["post-543","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-blog","tag-arc-ev-processor-ip","tag-kudan","tag-kudan-slam","tag-synopsys"],"veu_head_title_object":{"title":"","add_site_title":""},"_links":{"self":[{"href":"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-json\/wp\/v2\/posts\/543","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-json\/wp\/v2\/comments?post=543"}],"version-history":[{"count":5,"href":"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-json\/wp\/v2\/posts\/543\/revisions"}],"predecessor-version":[{"id":551,"href":"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-json\/wp\/v2\/posts\/543\/revisions\/551"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-json\/wp\/v2\/media\/409"}],"wp:attachment":[{"href":"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-json\/wp\/v2\/media?parent=543"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-json\/wp\/v2\/categories?post=543"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/webroad-seisaku.com\/fujiko-test\/wp-json\/wp\/v2\/tags?post=543"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}