{"id":10781,"date":"2013-10-24T00:17:26","date_gmt":"2013-10-23T21:17:26","guid":{"rendered":"http:\/\/hgpu.org\/?p=10781"},"modified":"2013-10-24T00:17:26","modified_gmt":"2013-10-23T21:17:26","slug":"a-multi-teraflop-constituency-parser-using-gpus","status":"publish","type":"post","link":"https:\/\/hgpu.org\/?p=10781","title":{"rendered":"A multi-Teraflop Constituency Parser using GPUs"},"content":{"rendered":"<p>Constituency parsing with rich grammars remains a computational challenge. Graphics Processing Units (GPUs) have previously been used to accelerate CKY chart evaluation, but gains over CPU parsers were modest. In this paper, we describe a collection of new techniques that enable chart evaluation at close to the GPU&#8217;s practical maximum speed (a Teraflop), or around a half-trillion rule evaluations per second. Net parser performance on a 4-GPU system is over 1 thousand length &#8211; 30 sentences\/second (1 trillion rules\/sec), and 400 general sentences\/second for the Berkeley Parser Grammar. The techniques we introduce include grammar compilation, recursive symbol blocking, and cache-sharing.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Constituency parsing with rich grammars remains a computational challenge. Graphics Processing Units (GPUs) have previously been used to accelerate CKY chart evaluation, but gains over CPU parsers were modest. In this paper, we describe a collection of new techniques that enable chart evaluation at close to the GPU&#8217;s practical maximum speed (a Teraflop), or around [&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":[11,89,3],"tags":[1782,14,95,20,1306,176,1347],"class_list":["post-10781","post","type-post","status-publish","format-standard","hentry","category-computer-science","category-nvidia-cuda","category-paper","tag-computer-science","tag-cuda","tag-high-level-languages","tag-nvidia","tag-nvidia-geforce-gtx-680","tag-package","tag-tesla-k10"],"views":3140,"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts\/10781","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=10781"}],"version-history":[{"count":0,"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts\/10781\/revisions"}],"wp:attachment":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=10781"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=10781"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=10781"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}