{"id":3110,"date":"2009-02-06T05:36:00","date_gmt":"2009-02-06T05:36:00","guid":{"rendered":"http:\/\/alienbabeltech.com\/main\/?p=3110"},"modified":"2009-04-09T13:10:51","modified_gmt":"2009-04-09T13:10:51","slug":"ati-catalyst-91-performance-review","status":"publish","type":"post","link":"http:\/\/alienbabeltech.com\/main\/ati-catalyst-91-performance-review\/","title":{"rendered":"ATi Catalyst 9.1 Performance Review"},"content":{"rendered":"<p><!--pagetitle:Introduction--><strong><span style=\"color: #99ccff;\"><span style=\"text-decoration: underline;\">Introduction<\/span> <\/span> <\/strong><\/p>\n<p>Catalyst 9.1 is the first driver to be released by ATi in 2009. Aside from adding OpenGL 3.0 support, it also contains the usual plethora of fixes. The last ATi driver I extensively tested was 8.10 and since I\u2019ll use it as a comparison to 9.1, any performance improvements in 8.12 and 8.11 will be relevant.<\/p>\n<p>ATi claims the following performance gains with <span style=\"color: #99ccff;\">8.12<\/span>:<\/p>\n<ul>\n<li><em>3DMark Vantage<\/em> DX10 \u2013 performance gains of up to 5% for Single and Crossfire mode. Performance gains were noticed on HD46xx, HD4550 and HD4350.<\/li>\n<li><em>Call of Duty: World at War<\/em> DX9 \u2013 performance gains of up to 21% for Single and Crossfire mode. Performance gains were noticed on all Radeon HD48xx.<\/li>\n<li><em>Crysis<\/em> DX10 \u2013 performance gains of up to 25% for Single and Crossfire mode. Performance gains were noticed on all Radeon HD4xxx.<\/li>\n<li><em>Crysis Warhead<\/em> DX10 \u2013 performance gains of up to 13% for Single card mode and up to 16% for Crossfire mode. Performance gains were noticed on all Radeon HD4xxx.<\/li>\n<li><em>Devil May Cry<\/em> 4 DX10\u2013 performance gains of up to 6% for Single and Crossfire mode. Performance gains were noticed on all Radeon HD4xxx and HD38xx.<\/li>\n<li><em>Fallout 3<\/em> \u2013 performance gains of up to 15% for Single card mode.<\/li>\n<li><em>Far Cry 2<\/em> DX10 \u2013 performance gains of up to 10% for Single card mode and up to 57% in Crossfire mode. Performance gains were noticed on all Radeon HD4xxx and HD38xx.<\/li>\n<li><em>FEAR<\/em> DX9 \u2013 performance gains of up to 6% for Single and Crossfire mode. Performance gains were noticed on Radeon HD4870X2 and HD4870.<\/li>\n<li><em>Hellgate: London<\/em> DX10 \u2013 performance gains of up to 6% for Single card mode and up to 10% for Crossfire mode. Performance gains were noticed on all Radeon HD4xxx.<\/li>\n<li><em>Left 4 Dead<\/em> DX9\u2013 performance gains of up to 10% for Single card mode and up to 5% for Crossfire mode. Performance gains were noticed on all Radeon HD4870 series.<\/li>\n<li><em>Lost Planet Colonies<\/em> DX10 \u2013 performance gains of up to 10% for Single and Crossfire mode. Performance gains were noticed on all Radeon HD4xxx and HD38xx.<\/li>\n<li><em>Prey<\/em> OGL \u2013 performance gains of up to 15% for Crossfire mode. Performance gains were noticed on Radeon HD4870 1GB products, HD46xx, HD45xx.<\/li>\n<li><em>S.T.A.L.K.E.R.: Clear Sky<\/em> \u2013 performance gains of up to 10% for Crossfire mode. Performance gains were noticed on all Radeon HD4xxx and HD38xx.<\/li>\n<li><em>Unreal Tournament 3<\/em> DX9 \u2013 performance gains of up to 18% for Single card mode and up to 15% in Crossfire mode, especially in cases where AA is enabled. Performance gains were noticed on all Radeon HD4xxx and HD38xx.<\/li>\n<\/ul>\n<p><strong><\/strong><\/p>\n<p>With <span style=\"color: #99ccff;\">8.11<\/span>, they claim the following:<\/p>\n<ul>\n<li><em>FarCry 2<\/em> \u2013 performance gains of 3 to 10% for Single card mode and 8 to 14% in Crossfire mode, with some specific configurations showing even greater improvements. Performance gains were noticed on all Radeon HD4xxx and Radeon HD3xxx products.<\/li>\n<li><em>S.T.A.L.K.E.R.: Clear Sky<\/em> \u2013 performance gains of 6 to 18% in single card settings and 10 to 30% in CrossfireX mode. Performance gains were noticed on all Radeon HD4xxx and Radeon HD3xxx products.<\/li>\n<\/ul>\n<p><strong><\/strong><\/p>\n<p>I\u2019m about to add several new titles to my benchmarking suite, but for now I\u2019ll stick to the same tests I ran with nVidia\u2019s <a href=\"http:\/\/alienbabeltech.com\/main\/?p=1703\" target=\"_blank\">Big Bang driver<\/a>. The 8.10 results will be recycled from my <a href=\"http:\/\/alienbabeltech.com\/main\/?p=1261\" target=\"_blank\">GTX260+ vs 4850 comparison<\/a>, and the system setup and configuration will be identical. I\u2019ll benchmark my <span style=\"color: #99ccff;\">single Radeon 4850 <\/span> with Catalyst 9.1 and compare the results to those garnered from 8.10.<\/p>\n<p>I\u2019m particularly looking forward to testing <em>Fear<\/em> and <em>Prey<\/em> because I like it when hardware vendors optimize drivers for older games, as it shows they care about legacy gaming.<\/p>\n<p>The color-coding in the text charts is as follows: <span style=\"color: #ffcc99;\">orange<\/span> means 8.10 wins while <span style=\"color: #ffff99;\">yellow<\/span> means Catalyst 9.1 wins.<\/p>\n<p><!--nextpage--> <!--pagetitle:Direct3D (Current)--> <strong><span style=\"color: #99ccff;\"><span style=\"text-decoration: underline;\">Direct3D (Current)<\/span> <\/span> <\/strong><\/p>\n<p style=\"TEXT-ALIGN: center\"><img loading=\"lazy\" decoding=\"async\" class=\"size-full wp-image-3120 aligncenter\" title=\"d3d-new-chart\" src=\"http:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/d3d-new-chart.png?resize=279%2C323\" alt=\"d3d-new-chart\" title=\"d3d-new-chart\" srcset=\"http:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/d3d-new-chart.png?resize=279%2C323 279w, http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/d3d-new-chart-259x300.png 259w\" sizes=\"auto, (max-width: 279px) 100vw, 279px\" data-recalc-dims=\"1\" \/>  <img loading=\"lazy\" decoding=\"async\" class=\"size-full wp-image-3121 aligncenter\" title=\"d3d-new-graph\" src=\"http:\/\/i2.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/d3d-new-graph.png?resize=635%2C561\" alt=\"d3d-new-graph\" title=\"d3d-new-graph\" srcset=\"http:\/\/i2.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/d3d-new-graph.png?resize=635%2C561 635w, http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/d3d-new-graph-300x265.png 300w\" sizes=\"auto, (max-width: 635px) 100vw, 635px\" data-recalc-dims=\"1\" \/><\/p>\n<p>The massive performance drop in <em>Bioshock<\/em> with 4xAA is actually a driver bug, but it\u2019s not related to performance. It appears that Catalyst 9.1 applies 8xAA when 4xAA is forced in that particular game. Both CCC and ATi Tray Tools had the same problem so the issue is likely at driver back-end which is responsible for applying the AA, rather than front-end utilities setting the wrong AA value.<\/p>\n<p>It\u2019s also disappointing to see <em>Crysis<\/em> being much slower without AA because that\u2019s a situation where you would really need the extra performance. And while 4xAA is <span style=\"color: #99ccff;\">15.84%<\/span> faster in said game, the victory is questionable since both results are a slide-show. Don\u2019t dismiss it completely though, as this gain could be more relevant for a 4870 because it would improve on a higher starting score.<\/p>\n<p><em>Call of Duty 4<\/em> does however run a bit better with 8xAA.<\/p>\n<p><!--nextpage--> <!--pagetitle:Direct3D (Legacy)--><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">Direct3D (Legacy)<\/span> <\/strong> <\/span><\/p>\n<p style=\"TEXT-ALIGN: center\"><img loading=\"lazy\" decoding=\"async\" class=\"size-full wp-image-3118 aligncenter\" title=\"d3d-legacy-chart\" src=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/d3d-legacy-chart.png?resize=267%2C187\" alt=\"d3d-legacy-chart\" title=\"d3d-legacy-chart\" data-recalc-dims=\"1\" \/><\/p>\n<p style=\"TEXT-ALIGN: center\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-3119\" title=\"d3d-legacy-graph\" src=\"http:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/d3d-legacy-graph.png?resize=636%2C320\" alt=\"d3d-legacy-graph\" title=\"d3d-legacy-graph\" srcset=\"http:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/d3d-legacy-graph.png?resize=636%2C320 636w, http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/d3d-legacy-graph-300x150.png 300w\" sizes=\"auto, (max-width: 636px) 100vw, 636px\" data-recalc-dims=\"1\" \/> <em><\/em><\/p>\n<p style=\"TEXT-ALIGN: left\"><em>Fear<\/em> does indeed show a small but relevant performance gain, as does <em>Serious Sam 2<\/em>. Free performance in older games is always nice.<\/p>\n<p><!--nextpage--> <!--pagetitle:OpenGL--><strong><span style=\"color: #99ccff;\"><span style=\"text-decoration: underline;\">OpenGL<\/span> <\/span> <\/strong><\/p>\n<p style=\"TEXT-ALIGN: center\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-3122\" title=\"ogl-chart\" src=\"http:\/\/i2.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/ogl-chart.png?resize=265%2C223\" alt=\"ogl-chart\" title=\"ogl-chart\" data-recalc-dims=\"1\" \/> <\/p>\n<p style=\"TEXT-ALIGN: center\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-3117\" title=\"ogl-graph\" src=\"http:\/\/i2.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/ogl-graph.png?resize=636%2C403\" alt=\"ogl-graph\" title=\"ogl-graph\" srcset=\"http:\/\/i2.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/ogl-graph.png?resize=636%2C403 636w, http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/ogl-graph-300x190.png 300w\" sizes=\"auto, (max-width: 636px) 100vw, 636px\" data-recalc-dims=\"1\" \/> <\/p>\n<p>Sadly <em>Prey\u2019s<\/em> performance doesn\u2019t change at all on my single 4850; perhaps the gains are only visible with Crossfire configurations.<\/p>\n<p>Performance elsewhere largely stays the same, except <em>Quake Wars<\/em> which sees a large <span style=\"color: #99ccff;\">17.44%<\/span> performance gain. This victory is again questionable on a 4850 due to the low scores in both cases, but it could be important to a 4870.<\/p>\n<p><!--nextpage--> <!--pagetitle:Conclusion--><strong><span style=\"color: #99ccff;\"><span style=\"text-decoration: underline;\">Conclusion<\/span> <\/span> <\/strong><\/p>\n<p>Catalyst 9.1 is quite a mixed bag, and there are some very large performance gains that could be pertinent depending on your particular ATi configuration, and the settings that you use.<\/p>\n<p>On a 4850 however, the performance gains that are relevant are usually quite small, but the performance loss in Crysis with 0xAA and Bioshock (due to the 4xAA bug) is severe enough to affect gameplay.<\/p>\n<p>In the specific case of a single 4850, I\u2019d personally skip these drivers unless there are specific bug fixes you need and\/or you\u2019re only concerned with the games that showed improvements.<\/p>\n","protected":false},"excerpt":{"rendered":"<p><a href=\"http:\/\/alienbabeltech.com\/main\/?p=3110\"><img decoding=\"async\" align=top title=\"Author: BFG10K\" src=\"http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/02\/article-image.png\" alt=\"\"><\/a><\/p>\n","protected":false},"author":5,"featured_media":3129,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[7814,7],"tags":[237,235,236,73,147,238],"class_list":["post-3110","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-articles","category-technology","tag-237","tag-235","tag-236","tag-ati","tag-catalyst","tag-radeon"],"_links":{"self":[{"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/posts\/3110","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/users\/5"}],"replies":[{"embeddable":true,"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/comments?post=3110"}],"version-history":[{"count":0,"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/posts\/3110\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/media\/3129"}],"wp:attachment":[{"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/media?parent=3110"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/categories?post=3110"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/tags?post=3110"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}