{"id":13454,"date":"2009-11-24T14:59:57","date_gmt":"2009-11-24T14:59:57","guid":{"rendered":"http:\/\/alienbabeltech.com\/main\/?p=13454"},"modified":"2009-11-28T02:38:42","modified_gmt":"2009-11-28T02:38:42","slug":"intel-e6850-bottleneck-investigation","status":"publish","type":"post","link":"https:\/\/alienbabeltech.com\/main\/intel-e6850-bottleneck-investigation\/","title":{"rendered":"Intel E6850 Bottleneck Investigation"},"content":{"rendered":"<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><!--pagetitle:Introduction--><strong><span style=\"color: #99ccff;\"><span style=\"text-decoration: underline;\">Introduction<\/span><\/span><\/strong><\/p>\n<p>Some of you have commented on my use of an E6850 processor in my reviews, and have questioned whether it could be limiting my results in some way.<\/p>\n<p>Also historically many tech forums contain a proliferation of people extolling the virtues of quad-core over dual-core for gaming. Others go even further and imply that unless you have a quad-core i7 overclocked to 4 GHz, you\u2019ll be CPU limited.<\/p>\n<p>I believe this is simply untrue, and I also believe CPU requirements given in such forums are vastly overblown for real world situations. I\u2019ve long argued that any decent mainstream dual-core platform is capable of taxing a graphics system just as good as the fastest quad-core CPU, providing you always run your games at the highest playable settings.<\/p>\n<p>So today I\u2019ll test 16 games to see if my theories are correct. The settings I\u2019ll use are quite varied but rest assured, they haven\u2019t been cherry-picked just to prove an agenda. They\u2019re the actual settings that I use when playing these games, so you can\u2019t get more real-world than that. They\u2019re infinitely more realistic than some 1024&#215;768 benchmarks done without AA to artificially isolate the CPU.<\/p>\n<p>My methodology is quite simple: since my E6850 is multiplier unlocked down to three speed grades, it means I can drop down to <span style=\"color: #99ccff;\">2 GHz<\/span> without affecting anything else in the system. Since this is a <span style=\"color: #99ccff;\">33%<\/span> underclock, I\u2019ll also underclock my GTX285 by 33% too, which is core 433 MHz, shader 986 MHz, and memory 832 MHz.<\/p>\n<p>I\u2019ll show results with both the CPU and GPU at stock, then I\u2019ll underclock the CPU and GPU individually by 33% (leaving the other at stock) to see which has the bigger performance impact. Note that any negative percentages in the <span style=\"color: #99ccff;\">2 GHz<\/span> column are benchmarking noise (given 2 GHz cannot be faster than 3 GHz), and hence have been truncated to 0% on the graphs.<\/p>\n<p>In the results, <span style=\"color: #99ccff;\">MS<\/span> denotes transparency multi-sampling, while <span style=\"color: #99ccff;\">SS<\/span> denotes transparency super-sampling.<\/p>\n<p><!--nextpage--><!--pagetitle:Setup--><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">Hardware<\/span><\/strong><\/span><\/p>\n<ul>\n<li>nVidia GeForce GTX285 (1 GB, reference clocks, and a 33% underclock).<\/li>\n<li>Intel Core 2 Duo E6850 (reference 3 GHz clock, and underclocked to 2 GHz).<\/li>\n<li>4 GB DDR2-800 RAM (4\u00d71 GB, dual-channel).<\/li>\n<li>Gigabyte GA-G33M-DS2R (Intel G33 chipset, F7 BIOS).<\/li>\n<li>Creative X-Fi XtremeMusic.<\/li>\n<li>30\u201d HP LP3065 (maximum resolution 2560&#215;1600).<\/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 XP 32 bit SP3<\/li>\n<li>nVidia driver 191.07, high quality filtering, all optimizations off, LOD clamp enabled.<\/li>\n<li>DirectX August 2009.<\/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><!--nextpage--><!--pagetitle:Benchmarks, Part 1--><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">Benchmarks, Part 1<\/span><\/strong><\/span><\/p>\n<p style=\"TEXT-ALIGN: center\"> <img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-13459\" title=\"Text 2xAA\" src=\"https:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/text-2xaa.png?resize=348%2C221\" alt=\"Text 2xAA\" srcset=\"https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/text-2xaa.png 348w, https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/text-2xaa-300x190.png 300w\" sizes=\"auto, (max-width: 348px) 100vw, 348px\" data-recalc-dims=\"1\" \/><\/p>\n<p style=\"TEXT-ALIGN: center\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-13456\" title=\"Graph 2xAA\" src=\"https:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/graph-2xaa.png?resize=528%2C442\" alt=\"Graph 2xAA\" srcset=\"https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/graph-2xaa.png 528w, https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/graph-2xaa-300x251.png 300w\" sizes=\"auto, (max-width: 528px) 100vw, 528px\" data-recalc-dims=\"1\" \/><\/p>\n<p>In this first batch of benchmarks we see an average performance drop of <span style=\"color: #99ccff;\">2.55%<\/span> from the CPU, while we see <span style=\"color: #99ccff;\">30.12%<\/span> from the GPU.<\/p>\n<p>Considering the underclock is <span style=\"color: #99ccff;\">33%<\/span>, a 30.12% performance drop is nearly linear and confirms the GPU is bottlenecking the system by almost 100%, even with the processor clocked at 2 GHz.<\/p>\n<p>Surprisingly we see movement in <em>Crysis<\/em> from the CPU underclock (<span style=\"color: #99ccff;\">8.99%<\/span>), but it\u2019s relatively trivial given the GPU affects performance more than three times more. We also see movement in <em>Far Cry 2<\/em> (<span style=\"color: #99ccff;\">4.71%<\/span>), but the GPU again affects performance much more &#8211; almost five times more. In these two games, adding a faster GPU will help performance tremendously more than a faster CPU will, even at 2 GHz.<\/p>\n<p>In <em>Call of Duty 5<\/em>, <em>Stalker Clear Sky<\/em> and <em>Crysis Warhead<\/em> we see absolutely zero movement from the processor. This confirms the GTX285 is bottlenecking the system by 100%, even when my processor is running at 2 GHz. In fact, the latter two games show a perfectly linear performance drop of 33%. This is a text-book bottleneck.<\/p>\n<p>And remember, these games are only using 2xAA because I want to keep the framerate up, and <em>Stalker Clear Sky<\/em> is only running at 1680&#215;1050.<\/p>\n<p><p><!--nextpage--><!--pagetitle:Benchmarks, Part 2--><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">Benchmarks, Part 2<\/span><\/strong><\/span><\/p>\n<p style=\"TEXT-ALIGN: center\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-13460\" title=\"Text 1920\" src=\"https:\/\/i2.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/text-1920.png?resize=367%2C186\" alt=\"Text 1920\" srcset=\"https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/text-1920.png 367w, https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/text-1920-300x152.png 300w\" sizes=\"auto, (max-width: 367px) 100vw, 367px\" data-recalc-dims=\"1\" \/> <\/p>\n<p style=\"TEXT-ALIGN: center\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-13457\" title=\"Graph 1920\" src=\"https:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/Graph-19201.png?resize=529%2C356\" alt=\"Graph 1920\" srcset=\"https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/Graph-19201.png 529w, https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/Graph-19201-300x201.png 300w\" sizes=\"auto, (max-width: 529px) 100vw, 529px\" data-recalc-dims=\"1\" \/><\/p>\n<p>In this next set of benchmarks we see the E6850 down-clocked to 2 GHz affecting performance by just <span style=\"color: #99ccff;\">2.76%<\/span> overall, while doing the same percentage underclock to the GTX285 reduces performance by <span style=\"color: #99ccff;\">31.02%<\/span>. Again, this is almost a linear performance drop from the GPU and confirms it\u2019s the bottleneck by almost 100%.<\/p>\n<p>We see slight movement in <em>Jericho<\/em> and <em>Unreal Tournament 3<\/em> from running the CPU at 2 GHz, but this is again vastly overshadowed by the much bigger impact the GPU has.<\/p>\n<p><em>Call of Duty 4<\/em> and <em>Wolfenstein<\/em> show absolutely no movement from the CPU dropping to 2 GHz, and the latter game also shows completely linear scaling when underclocking the GPU.<\/p>\n<p><!--nextpage--><!--pagetitle:Benchmarks, Part 3--><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">Benchmarks, Part 3<\/span><\/strong><\/span><\/p>\n<p style=\"TEXT-ALIGN: center\"> <img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-13455\" title=\"Text 2560\" src=\"https:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/text-2560.png?resize=369%2C289\" alt=\"Text 2560\" srcset=\"https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/text-2560.png 369w, https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/text-2560-300x234.png 300w\" sizes=\"auto, (max-width: 369px) 100vw, 369px\" data-recalc-dims=\"1\" \/><\/p>\n<p style=\"TEXT-ALIGN: center\"> <img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-13458\" title=\"Graph 2560\" src=\"https:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/Graph-25601.png?resize=528%2C563\" alt=\"Graph 2560\" srcset=\"https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/Graph-25601.png 528w, https:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/Graph-25601-281x300.png 281w\" sizes=\"auto, (max-width: 528px) 100vw, 528px\" data-recalc-dims=\"1\" \/><\/p>\n<p>There isn\u2019t much to say about this next set of benchmarks, except to show the processor is making practically zero performance difference. Meanwhile, the average performance drop from the GPU being underclocked is <span style=\"color: #99ccff;\">33.11%<\/span>. This means the GTX285 is bottlenecking the system by 100% in all of these games, even with my E6850 running at 2 GHz.<\/p>\n<p>If I dropped a Fermi or 5870 into my system right now and re-ran these tests, I\u2019d see a massive performance gain, even with my E6850 running at 2 GHz. But I\u2019d get nothing from throwing a 4 GHz quad-core at my GTX285.<\/p>\n<p><!--nextpage--><!--pagetitle:Frequent Responses, Part 1--><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">Frequent Responses, Part 1<\/span><\/strong><\/span><\/p>\n<p><span style=\"color: #99ccff;\">Overclocking is not the same as underclocking.<\/span><\/p>\n<p>This is untrue. A clock rate is simply a number with an associated amount of work tied to it. The hardware doesn\u2019t care about English terms like \u201cover\u201d and \u201cunder\u201d. All it cares about is the frequency it\u2019s running at, because that determines how fast it is.<\/p>\n<p>Or to put it another way, you can simply treat 2 GHz as \u201cstock\u201d and 3 GHz as a 50% \u201coverclock\u201d. Then invert the percentages in the graphs, and you\u2019ll come out with exactly the same results. Then do the same for the GPU, and you\u2019ll again come out with exactly the same results.<\/p>\n<p><span style=\"color: #99ccff;\">If you had used lower settings, you would have seen a bigger impact from the CPU.<\/span><\/p>\n<p>I used the highest playable settings for my setup (i.e. the actual settings I use when I play these games), and I expect everyone else to do the same for their respective GPU. If you aren\u2019t doing that then you\u2019re underutilizing your GPU.<\/p>\n<p>I most certainly don\u2019t expect someone with a 4 GHz quad-core to be playing games at 1024&#215;768 with no AA. The whole purpose of a powerful system is to crank the settings, otherwise it\u2019s worthless over a weaker system.<\/p>\n<p>Benchmarks done at the likes of 1024&#215;768 to show CPU \u201cdifferences\u201d are totally meaningless given they create artificially contrived scenarios that mean nothing in the real world.<\/p>\n<p><span style=\"color: #99ccff;\">If you had used game X, you would have seen more of a CPU difference.<\/span><\/p>\n<p>There are certainly some games that depend on the CPU more, but they\u2019re a drop in the bucket. Also with every such game I\u2019ve been presented in the past, I\u2019ve been able to demonstrate that any reasonable detail level mostly nullifies the difference between mainstream and high-end CPUs, and the GPU makes the biggest difference by far.<\/p>\n<p>If you think you\u2019ve found a good game, feel free to drop by our forums and let me know about it.<\/p>\n<p><span style=\"color: #99ccff;\">You\u2019re using a dinosaur OS (XP), so no wonder you found no difference.<\/span><\/p>\n<p>Actually XP is easier on the GPU because DX10\/DX11 render paths are typically more demanding than DX9 render paths. Also DX11 allows multi-threaded rendering, so it\u2019ll load the graphics card more. So if anything, Vista\/7 makes the system even more GPU bound than XP does.<\/p>\n<p><!--nextpage--><!--pagetitle:Frequent Responses, Part 2--><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">Frequent Responses, Part 2<\/span><\/strong><\/span><\/p>\n<p><span style=\"color: #99ccff;\">Not everyone has a 2560&#215;1600 monitor, or even a 1920&#215;1200 monitor.<\/span><\/p>\n<p>If you have a low resolution display like 1680&#215;1050 and you\u2019re sinking money into an expensive quad-core CPU\/platform, then your money is being wasted. You\u2019ll get far more benefit to gaming by upgrading to a larger high resolution display.<\/p>\n<p>Gaming on a 30\u201d display at 2560&#215;1600 with a mainstream dual-core is far more immersive than gaming on a 22\u201d display at 1680&#215;1050 with a high-end quad-core. Even the difference between 2560&#215;1600 and 1920&#215;1200 is like night and day.<\/p>\n<p>Anyway, there\u2019s nothing stopping you from using higher AA levels if your monitor is limited to low resolutions. Try 8xMSAA, or any of the more taxing vendor-specific AA modes.<\/p>\n<p><span style=\"color: #99ccff;\">You test old games.<\/span><\/p>\n<p>All titles tested today shipped in 2006 or later, and three of them came out this year (2009).<\/p>\n<p><span style=\"color: #99ccff;\">Your findings don\u2019t apply to multi-GPU setups.<\/span><\/p>\n<p>This is true to some extent, until you realize that it\u2019s always about the highest playable settings. Since multi-GPU allows higher settings than a single GPU, using said settings moves the bottleneck back onto the graphics system. If I had a quad GPU system for example, I\u2019d be aiming to use super-sampling in every game.<\/p>\n<p>If you use the same settings that were already playable on a single GPU, then you\u2019re simply squandering the performance advantage that multi-GPU provides over a single GPU.<\/p>\n<p>Also each time a new single GPU flagship comes out, it competes with (or even surpasses) the previous generation multi-GPU. Yet after two years of using my E6850 with single-GPU flagship cards, it\u2019s showing absolutely no signs of slowing down, even when underclocked to 2 GHz.<\/p>\n<p><span style=\"color: #99ccff;\">Your findings don&#8217;t apply to heavy multi-tasking.<\/span><\/p>\n<p>This is outside the scope of this article, which focuses entirely on the myth of CPU bottlenecking pertaining to normal gaming. If you regularly encode multiple 1080p videos in the background while playing games then by all means, buy the fastest quad-core you can afford.<\/p>\n<p><!--nextpage--><!--pagetitle:Conclusion--><span style=\"color: #99ccff;\"><strong><span style=\"text-decoration: underline;\">Conclusion<\/span><\/strong><\/span><\/p>\n<p>After all of this, it should be obvious why I\u2019m still using my E6850: <span style=\"color: #99ccff;\">because there\u2019s no need for anything faster<\/span>. At reasonably high detail levels the GPU is by far the most important equation to gaming, and hence my GTX285 is the primary bottleneck in every gaming situation I use it in.<\/p>\n<p>Even when underclocked to 2 GHz, my E6850 still pushes my GTX285 hard enough so that it\u2019s a 100% bottleneck in most situations I game at. The fact that stock speed is actually 3 GHz demonstrates why I don\u2019t bother overclocking. The last time I was CPU limited to any meaningful degree was in the Pentium 4 days.<\/p>\n<p>Don\u2019t believe the \u201cyou\u2019re CPU limited under 4 GHz\u201d and the \u201cyou need quad-core to game\u201d myths. If you always configure your games to use the highest playable settings, any decent mainstream dual-core CPU is enough to push the graphics system into being the primary bottleneck, thereby erasing most practical differences the fastest quad-core setups potentially offer.<\/p>\n<p>If you have any kind of limited budget, sink as much money as you can into the graphics card, and also buy the biggest monitor with the biggest resolution you can afford. Then configure all of your games to run at the highest playable settings to fully take advantage of your purchase. You\u2019ll utilize your hardware far better this way, instead of using settings that are too low and squander GPU performance.<\/p>\n<p>Of course, if you already have the best monitor and GPU setup available and you still have money left over then by all means, sink it into the fastest CPU\/platform available. The same might also apply if you employ heavy multi-tasking.<\/p>\n<p><strong><\/strong><\/p>\n<blockquote>\n<p><span style=\"color: #99ccff;\">Please join us in our <a href=\"http:\/\/alienbabeltech.com\/abt\/index.php\" target=\"_blank\">Forums<\/a><\/span><\/p>\n<p><span style=\"color: #99ccff;\">Follow us on <a href=\"http:\/\/twitter.com\/alienbabeltech\" onclick=\"_gaq.push(['_trackEvent', 'outbound-article', 'http:\/\/twitter.com\/alienbabeltech', 'Twitter']);\" target=\"_blank\">Twitter<\/a><\/span><\/p>\n<p><span style=\"color: #99ccff;\">For the latest updates from ABT, please <a href=\"http:\/\/alienbabeltech.com\/main\/?feed=rss2\" target=\"_blank\">join our RSS News Feed<\/a><\/span><\/p>\n<\/blockquote>\n","protected":false},"excerpt":{"rendered":"<p><a href=\"http:\/\/alienbabeltech.com\/main\/?p=13454\"><img decoding=\"async\" align=top title=\"Author: BFG10K\" src=\"http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2009\/11\/Article-Image1.PNG\"><\/a><\/p>\n","protected":false},"author":5,"featured_media":13534,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2533,7814,7],"tags":[2763,293,294,842,2764,527,900],"class_list":["post-13454","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-abt-news","category-articles","category-technology","tag-190-07","tag-bottleneck","tag-bottleneck-investigation","tag-cpu","tag-cpu-limited","tag-e6850","tag-gtx285"],"_links":{"self":[{"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/posts\/13454","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=13454"}],"version-history":[{"count":0,"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/posts\/13454\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/media\/13534"}],"wp:attachment":[{"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/media?parent=13454"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/categories?post=13454"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/tags?post=13454"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}