{"id":20153,"date":"2010-08-15T09:34:29","date_gmt":"2010-08-15T08:34:29","guid":{"rendered":"http:\/\/alienbabeltech.com\/main\/?p=20153"},"modified":"2010-08-16T04:13:06","modified_gmt":"2010-08-16T03:13:06","slug":"nvidia-gtx470-bottleneck-investigation","status":"publish","type":"post","link":"https:\/\/alienbabeltech.com\/main\/nvidia-gtx470-bottleneck-investigation\/","title":{"rendered":"nVidia GTX470 Bottleneck Investigation"},"content":{"rendered":"<p><!--pagetitle:Introduction--><strong><span style=\"color: #99ccff;\"><span style=\"text-decoration: underline;\">Introduction<\/span><\/span><\/strong><\/p>\n<p>This article is a continuation of my ongoing investigation of bottlenecking in video cards. The crux of it is to individually underclock the various clock speeds of a video card to see which has the biggest performance impact. Here are the past articles I&#8217;ve written on the same subject:<\/p>\n<ul>\n<li><a href=\"http:\/\/alienbabeltech.com\/main\/?p=12474\" target=\"_blank\">ATi Radeon 5770<\/a>.<\/li>\n<li><a href=\"http:\/\/alienbabeltech.com\/main\/?p=9368\" target=\"_blank\">nVidia GTX285<\/a>.<\/li>\n<li><a href=\"http:\/\/alienbabeltech.com\/main\/?p=4116\" target=\"_blank\">nVidia GTX260<\/a>+.<\/li>\n<li><a href=\"http:\/\/alienbabeltech.com\/main\/?p=969\" target=\"_blank\">nVidia 8800 Ultra<\/a>.<\/li>\n<\/ul>\n<p><strong><\/strong><\/p>\n<p>From the 8xxx series onwards it was possible to control core, shader and memory clocks on nVidia\u2019s cards. With the 4xx series that has changed, and now the shader clock controls all non-memory clocks. More specifically, the core clock is always half of the shader clock.<\/p>\n<p>Since there\u2019s no real limit to how far we can underclock, I\u2019ve chosen a nice round figure of <span style=\"color: #99ccff;\">20%<\/span>. I\u2019ll individually underclock the shader and memory (leaving the other at stock) to see which has the most performance impact. This equates to a <span style=\"color: #99ccff;\">972 MHz shader<\/span> clock and a <span style=\"color: #99ccff;\">1339 MHz memory<\/span> clock.<\/p>\n<p>The tests will be done on a single <span style=\"color: #99ccff;\">GTX470<\/span>, and this card just happens to have about 16% less memory bandwidth than a GTX285, but has more pixel fillrate and shader performance. In light of this, it\u2019ll be interesting to see whether the card is castrated by the reduced bandwidth.<\/p>\n<p>1920&#215;1200 will be used with both 2xAA and 4xAA, because these are quite realistic settings for someone gaming on a single GTX470.<\/p>\n<p align=\"center\"><script type=\"text\/javascript\"><!--\ngoogle_ad_client = \"pub-7021221536180758\";\n\/* 468x60, created 11\/11\/09 *\/\ngoogle_ad_slot = \"1949837292\";\ngoogle_ad_width = 468;\ngoogle_ad_height = 60;\n\/\/-->\n<\/script><br \/>\n<script type=\"text\/javascript\"\nsrc=\"http:\/\/pagead2.googlesyndication.com\/pagead\/show_ads.js\">\n<\/script><\/p>\n<p><!--nextpage--><!--pagetitle:Setup--><strong><span style=\"color: #99ccff;\"><span style=\"text-decoration: underline;\">Hardware<\/span><\/span><\/strong><\/p>\n<ul>\n<li>Intel Core i5 750 (2.8 GHz, turbo on).<\/li>\n<li>4 GB DDR3-1333 RAM (2\u00d72 GB, dual-channel).<\/li>\n<li>Gigabyte GA-P55-UD3 (Intel P55 chipset, F6 BIOS).<\/li>\n<li>nVidia GeForce GTX470 (1.28 GB, reference clocks).<\/li>\n<li>Creative X-Fi XtremeMusic.<\/li>\n<li>30\u201d HP LP3065.<\/li>\n<\/ul>\n<p><strong><\/strong><\/p>\n<p><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">Software<\/span><\/strong><\/span><\/p>\n<ul>\n<li>Windows 7 (64 bit).<\/li>\n<li>nVidia driver 258.96, high quality filtering, all optimizations off, LOD clamp enabled.<\/li>\n<li>DirectX June 2010.<\/li>\n<li>All games patched to their latest versions.<\/li>\n<\/ul>\n<p><strong><\/strong><\/p>\n<p><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">Settings<\/span><\/strong><\/span><\/p>\n<ul>\n<li>16xAF forced in the driver, vsync forced off in the driver.<\/li>\n<li>AA forced either through the driver or enabled in-game, whichever works better.<\/li>\n<li>Highest quality sound (stereo) used in all games.<\/li>\n<li>All results show an average framerate.<\/li>\n<\/ul>\n<p align=\"center\"><script type=\"text\/javascript\"><!--\ngoogle_ad_client = \"pub-7021221536180758\";\n\/* 468x60, created 11\/11\/09 *\/\ngoogle_ad_slot = \"1949837292\";\ngoogle_ad_width = 468;\ngoogle_ad_height = 60;\n\/\/-->\n<\/script><br \/>\n<script type=\"text\/javascript\"\nsrc=\"http:\/\/pagead2.googlesyndication.com\/pagead\/show_ads.js\">\n<\/script><\/p>\n<p><!--nextpage--><!--pagetitle:2xAA--><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">2xAA<\/span><\/strong><\/span><\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-20155\" title=\"2xAA Text\" src=\"https:\/\/i2.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/2xAA-Text.png?resize=411%2C255\" alt=\"2xAA Text\" srcset=\"https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/2xAA-Text.png 411w, https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/2xAA-Text-300x186.png 300w\" sizes=\"auto, (max-width: 411px) 100vw, 411px\" data-recalc-dims=\"1\" \/><\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-20154\" title=\"2xAA Graph\" src=\"https:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/2xAA-Graph.png?resize=451%2C408\" alt=\"2xAA Graph\" srcset=\"https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/2xAA-Graph.png 451w, https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/2xAA-Graph-300x271.png 300w\" sizes=\"auto, (max-width: 451px) 100vw, 451px\" data-recalc-dims=\"1\" \/><\/p>\n<p>At 2xAA the shader clock impacts performance by <span style=\"color: #99ccff;\">14.63%<\/span> overall, while the memory clock only causes a loss of <span style=\"color: #99ccff;\">6.51%<\/span>. In particular, <em>Stalker<\/em> sees an almost non-existent performance drop from a 20% underclock to the memory.<\/p>\n<p align=\"center\"><script type=\"text\/javascript\"><!--\ngoogle_ad_client = \"pub-7021221536180758\";\n\/* 468x60, created 11\/11\/09 *\/\ngoogle_ad_slot = \"1949837292\";\ngoogle_ad_width = 468;\ngoogle_ad_height = 60;\n\/\/-->\n<\/script><br \/>\n<script type=\"text\/javascript\"\nsrc=\"http:\/\/pagead2.googlesyndication.com\/pagead\/show_ads.js\">\n<\/script><\/p>\n<p><!--nextpage--><!--pagetitle:4xAA--><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">4xAA<\/span><\/strong><\/span><\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-20157\" title=\"4xAA Text\" src=\"https:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/4xAA-Text.png?resize=412%2C255\" alt=\"4xAA Text\" srcset=\"https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/4xAA-Text.png 412w, https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/4xAA-Text-300x185.png 300w\" sizes=\"auto, (max-width: 412px) 100vw, 412px\" data-recalc-dims=\"1\" \/><\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-20156\" title=\"4xAA Graph\" src=\"https:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/4xAA-Graph.png?resize=452%2C408\" alt=\"4xAA Graph\" srcset=\"https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/4xAA-Graph.png 452w, https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/4xAA-Graph-300x270.png 300w\" sizes=\"auto, (max-width: 452px) 100vw, 452px\" data-recalc-dims=\"1\" \/><\/p>\n<p>As expected, moving to 4xAA makes the loss of memory bandwidth more significant, but the shader clock still impacts performance more overall (<span style=\"color: #99ccff;\">13.47%<\/span> vs <span style=\"color: #99ccff;\">8.09%<\/span>). Once again, <em>Stalker<\/em> sees little performance drop from a 20% loss in bandwidth.<\/p>\n<p align=\"center\"><script type=\"text\/javascript\"><!--\ngoogle_ad_client = \"pub-7021221536180758\";\n\/* 468x60, created 11\/11\/09 *\/\ngoogle_ad_slot = \"1949837292\";\ngoogle_ad_width = 468;\ngoogle_ad_height = 60;\n\/\/-->\n<\/script><br \/>\n<script type=\"text\/javascript\"\nsrc=\"http:\/\/pagead2.googlesyndication.com\/pagead\/show_ads.js\">\n<\/script><\/p>\n<p><!--nextpage--><!--pagetitle:Conclusion--><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">Conclusion<\/span><\/strong><\/span><\/p>\n<p>The GTX470 does not appear to be primarily limited by its memory bandwidth, even with the 16% reduction over the GTX285\u2019s bandwidth. Even at 4xAA the shader clock is making a much bigger difference overall, and this suggests nVidia has equipped the GTX470 with the bandwidth it needs.<\/p>\n<p>Needless to say, if you\u2019re overclocking a GTX470, raise the shader clock as high as possible.<\/p>\n<p><strong><\/strong><\/p>\n<blockquote>\n<p>Please join us in our <a href=\"http:\/\/alienbabeltech.com\/abt\/index.php\" target=\"_blank\">Forums<\/a><\/p>\n<p>Follow us on <a href=\"http:\/\/twitter.com\/alienbabeltech\" onclick=\"_gaq.push(['_trackEvent', 'outbound-article', 'http:\/\/twitter.com\/alienbabeltech', 'Twitter']);\" target=\"_blank\">Twitter<\/a><\/p>\n<p>For the latest updates from ABT, please <a href=\"http:\/\/alienbabeltech.com\/main\/?feed=rss2\" target=\"_blank\">join our RSS News Feed<\/a><\/p>\n<p>Join our Distributed Computing teams<\/p>\n<ul>\n<li>Folding@Home &#8211; Team AlienBabelTech &#8211; 164304 <\/li>\n<li>SETI@Home &#8211; Team AlienBabelTech &#8211; 138705 <\/li>\n<li><a href=\"http:\/\/www.worldcommunitygrid.org\/reg\/viewRegister.do?teamID=3P39R3SRV1\" onclick=\"_gaq.push(['_trackEvent', 'outbound-article', 'http:\/\/www.worldcommunitygrid.org\/reg\/viewRegister.do?teamID=3P39R3SRV1', 'World Community Grid &#8211; Team AlienBabelTech']);\" target=\"_blank\">World Community Grid &#8211; Team AlienBabelTech<\/a><\/li>\n<\/ul>\n<\/blockquote>\n<p align=\"center\"><script type=\"text\/javascript\"><!--\ngoogle_ad_client = \"pub-7021221536180758\";\n\/* 468x60, created 11\/11\/09 *\/\ngoogle_ad_slot = \"1949837292\";\ngoogle_ad_width = 468;\ngoogle_ad_height = 60;\n\/\/-->\n<\/script><br \/>\n<script type=\"text\/javascript\"\nsrc=\"http:\/\/pagead2.googlesyndication.com\/pagead\/show_ads.js\">\n<\/script><\/p>\n","protected":false},"excerpt":{"rendered":"<p><a href=\"http:\/\/alienbabeltech.com\/main\/?p=20153 \"><img decoding=\"async\" align=top title=\"Author: BFG10K\" src=\"http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2010\/08\/Article-Image1.png\"><\/a><\/p>\n","protected":false},"author":5,"featured_media":20174,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2533,7814,7],"tags":[4029,293,3170,943,292,4034,291],"class_list":["post-20153","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-abt-news","category-articles","category-technology","tag-258-96","tag-bottleneck","tag-gtx470","tag-memory","tag-overclocking","tag-shader","tag-underclocking"],"_links":{"self":[{"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/posts\/20153","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/users\/5"}],"replies":[{"embeddable":true,"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/comments?post=20153"}],"version-history":[{"count":0,"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/posts\/20153\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/media\/20174"}],"wp:attachment":[{"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/media?parent=20153"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/categories?post=20153"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/tags?post=20153"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}