{"id":32488,"date":"2012-11-12T14:00:49","date_gmt":"2012-11-12T14:00:49","guid":{"rendered":"http:\/\/alienbabeltech.com\/main\/?p=32488"},"modified":"2012-11-19T23:23:25","modified_gmt":"2012-11-19T23:23:25","slug":"announcing-nvidias-gk110-tesla-k20-and-k20x","status":"publish","type":"post","link":"http:\/\/alienbabeltech.com\/main\/announcing-nvidias-gk110-tesla-k20-and-k20x\/","title":{"rendered":"Announcing Nvidia&#8217;s GK110 Tesla K20 and K20x"},"content":{"rendered":"<p style=\"text-align: center;\" align=\"left\"><script type=\"text\/javascript\">\/\/ <![CDATA[\n                                                            google_ad_client = \"pub-7021221536180758\"; \/* 468x60, created 11\/11\/09 *\/ google_ad_slot = \"1949837292\"; google_ad_width = 468; google_ad_height = 60;\n\/\/ ]]><\/script><script type=\"text\/javascript\" src=\"http:\/\/pagead2.googlesyndication.com\/pagead\/show_ads.js\">\/\/ <![CDATA[\n\n\n\/\/ ]]><\/script><\/p>\n<p style=\"text-align: left;\">Two weeks ago, the U.S. Department of Energy\u2019s Oak Ridge National Laboratory (ORNL)\u00a0launched a <a href=\"http:\/\/alienbabeltech.com\/main\/?p=32384\">new era of scientific supercomputing with Titan<\/a>.\u00a0\u00a0Today it is announced that\u00a0Titan\u00a0has just\u00a0taken\u00a0the title as the fastest supercomputer\u00a0in the world.\u00a0\u00a0\u00a0Titan\u00a0was made possible by\u00a0using Nvidia&#8217;s new K20X Tesla accelerators to do ninety percent of the computing load.\u00a0<img loading=\"lazy\" decoding=\"async\" class=\"aligncenter\" src=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/10\/TitanCabs_10-3.jpg?resize=480%2C171\" alt=\"\" data-recalc-dims=\"1\" \/><\/p>\n<p>Using <span style=\"font-size: small;\">18,688 Nvidia Tesla K20X GPU accelerators, the Titan supercomputer took the world&#8217;s top spot with a performance record of 17.59 petaflops as measured by the LINPACK benchmark.\u00a0 Best of all, Tesla K20X accelerator\u00a0is energy efficient\u00a0and<\/span> Titan achieved 2,142.77 megaflops of performance per watt, which surpasses the energy efficiency of the\u00a0number\u00a0one system on the most recent Green500 list of the world\u2019s most energy-efficient supercomputers.<\/p>\n<p style=\"text-align: left;\">To coincide with <a href=\"http:\/\/sc12.supercomputing.org\/\" onclick=\"_gaq.push(['_trackEvent', 'outbound-article', 'http:\/\/sc12.supercomputing.org\/', 'SC12']);\" >SC12<\/a> in Salt Lake City\u00a0\u00a0&#8211; where it is today\u00a0confirmed that Titan is the world&#8217;s fastest\u00a0supercomputer on <a href=\"http:\/\/www.top500.org\/\" onclick=\"_gaq.push(['_trackEvent', 'outbound-article', 'http:\/\/www.top500.org\/', 'the top 500 list']);\" >the top 500 list<\/a> &#8211;\u00a0\u00a0Nvidia is announcing the Tesla K20 family of accelerators.\u00a0 Intel will be there also to unveil their brand new\u00a0Xeon Phi accelerators and although AMD is behind in supercomputing, AMD\u00a0launches its\u00a0dual-GPU FirePro S10000 server-card with\u00a06GB of memory.\u00a0\u00a0 <a href=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/NVIDIA_Tesla_K20X_K20_GPU_A.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter  wp-image-32489\" title=\"NVIDIA_Tesla_K20X_K20_GPU_A\" src=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/NVIDIA_Tesla_K20X_K20_GPU_A.jpg?resize=480%2C394\" alt=\"\" srcset=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/NVIDIA_Tesla_K20X_K20_GPU_A.jpg?resize=480%2C394 800w, http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/NVIDIA_Tesla_K20X_K20_GPU_A-300x246.jpg 300w\" sizes=\"auto, (max-width: 480px) 100vw, 480px\" data-recalc-dims=\"1\" \/><\/a><\/p>\n<p>Today we are going to take a closer look at Nvidia&#8217;s new K20 family of accelerators &#8211;\u00a0 the K20 and the K20X.\u00a0\u00a0The K20X provides\u00a0the highest computing performance ever available in a single processor, surpassing all other processors on two common measures of computational performance \u2013 3.95 teraflops single-precision and 1.31 teraflops double-precision peak floating point performance.<\/p>\n<p>The new K20\u00a0family also includes the Tesla K20 accelerator, which provides 3.52 teraflops of single-precision and 1.17 teraflops of double-precision peak performance.\u00a0\u00a0\u00a0The K20X is\u00a0clocked at 732MHz for the core clock and 5.2GHz for the memory clock while the K20 is clocked slightly lower\u00a0at 706MHz with\u00a0the same\u00a0memory clock.<img loading=\"lazy\" decoding=\"async\" class=\"aligncenter\" src=\"http:\/\/i2.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/05\/TeslaKeplerGK110_FNL_800_PR.jpg?resize=287%2C300\" alt=\"\" data-recalc-dims=\"1\" \/>Above\u00a0is pictured Nvidia&#8217;s new GK110 Kepler GPU which at 7.1 billion transistors is the most complex piece\u00a0of silicon anywhere.\u00a0 Although we are primarily gamers, we realize that these very same GPUs that power supercomputers as Tesla are used in our GeForce video cards.\u00a0 ABT has been following\u00a0GK110 closely since we covered <a href=\"http:\/\/alienbabeltech.com\/main\/?p=29692&amp;all=1\">Nvidia&#8217;s GTC 2012<\/a>\u00a0where the new architecture was unveiled.<\/p>\n<p>The GPU\u00a0Technology Conference 2012\u00a0was not about gaming GPUs and it\u00a0wasn\u2019t mentioned in the Whitepaper, but look very carefully at the die shot above. It\u00a0is obvious that there are 5 Graphics Processing Clusters (GPCs)\u00a0and 3 SMX modules per GPC.\u00a0 A GPC is constructed like a \u201cmini GPU\u201d which\u00a0contain SMXs; two in the case of GK104 and\u00a0three in the case of GK110.<\/p>\n<p>A completely functional GK110 GPU would be made up of 15 SMXes for a total of 2880 CUDA cores (192 x 15).\u00a0 However, for yield purposes and to differentiate faster, more complex,\u00a0and more expensive processors from less expensive, slower and less complex products, parts are often disabled and clockspeeds lowered.<\/p>\n<p>In the case of K20X, 14 SMXes are enabled for a total of 2688 CUDA cores.\u00a0 The K20 features 13 SMXes for a total of 2496 CUDA cores.\u00a0 Although Nvidia absolutely will not comment on unreleased products, we can probably expect gaming GPUs based on this same GK110 GPU and perhaps the K20X may correspond to a future GTX 780 and\u00a0the K20, to the future\u00a0GTX 770 just as the K10 corresponds to the GTX 690.\u00a0 So the specifications of these professional cards\u00a0do\u00a0interest gamers.<\/p>\n<p><strong>The Specifications<\/strong><\/p>\n<p>Here are the specifications as released in Nvidia&#8217;s chart comparing the new single GPU K20\/K20X to the dual GPU K10.\u00a0 TDP isn&#8217;t mentioned but both varieties of KX20 fit within the 225W TDP specification.<\/p>\n<p style=\"text-align: center;\"><a href=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/SPECIFICATIONS.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter  wp-image-32490\" title=\"SPECIFICATIONS\" src=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/SPECIFICATIONS.jpg?resize=560%2C288\" alt=\"\" srcset=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/SPECIFICATIONS.jpg?resize=560%2C288 800w, http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/SPECIFICATIONS-300x154.jpg 300w\" sizes=\"auto, (max-width: 560px) 100vw, 560px\" data-recalc-dims=\"1\" \/><\/a><\/p>\n<p><strong>Features<\/strong><\/p>\n<p>Here are the\u00a0K20 features and benefits\u00a0as released in Nvidia&#8217;s chart.<\/p>\n<p style=\"text-align: center;\"><a href=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/features.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter  wp-image-32491\" title=\"features\" src=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/features.jpg?resize=560%2C219\" alt=\"\" srcset=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/features.jpg?resize=560%2C219 800w, http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/features-300x117.jpg 300w\" sizes=\"auto, (max-width: 560px) 100vw, 560px\" data-recalc-dims=\"1\" \/><\/a><\/p>\n<p>As we learned at the <a href=\"http:\/\/alienbabeltech.com\/main\/?p=29692&amp;all=1\">GTC 2012<\/a> this Spring, the new Kepler architecture is far more\u00a0efficient than Fermi and it also offers new features which speed up computing such as Dynamic Parallism and HyperQ.\u00a0 HyperQ (below left) provide a significant speedup for legacy MPI codes while Dynamic Parallelism (below right) enables the GPU to generate code for itself instead of waiting on the CPU.<\/p>\n<p style=\"text-align: center;\"><a href=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/FERMIvKEPLER.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter  wp-image-32492\" title=\"FERMIvKEPLER\" src=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/FERMIvKEPLER.jpg?resize=560%2C132\" alt=\"\" srcset=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/FERMIvKEPLER.jpg?resize=560%2C132 800w, http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/FERMIvKEPLER-300x70.jpg 300w\" sizes=\"auto, (max-width: 560px) 100vw, 560px\" data-recalc-dims=\"1\" \/><\/a><\/p>\n<p>\u00a0ABT has watched Nvidia venture officially\u00a0into supercomputing with <a href=\"http:\/\/alienbabeltech.com\/main\/?p=643&amp;all=1\">Nvision08,<\/a> just 4 years ago.\u00a0 And during this time, it has carved out and pioneered GP-GPU computing &#8211; first\u00a0taking a simple graphics accelerator for PC gaming\u00a0and making it programmable back in 1999\u00a0thus allowing it to be used for calculations other than graphics.\u00a0 There is very little difference in a GPU making scientific calculations to calculating geometry for a game.<\/p>\n<p>To support GP-GPU, Nvidia released CUDA in 2007 and it has gone through 5 iterations.\u00a0 \u00a0From humble beginnings,\u00a0CUDA has grown significantly\u00a0in just four years.\u00a0 CUDA is Nvidia\u2019s own proprietary GPU language which can be considered similar to x86 for CPU. From 150,000 CUDA downloads and 1 Supercomputer in 2008 to 1,500,000 CUDA downloads and 36 supercomputers today; and from 60 universities and 4,000 academic papers to\u00a0629 universities teaching CUDA and over\u00a022,500 academic papers &#8211; all in 4 years!<\/p>\n<p>The reason that GPGPU computing has exploded onto the supercomputing scene is because it is very disruptive to the CPU technology that dominates.\u00a0 The GPU is much more efficient at certain tasks &#8211; often there is a ten times speedup or far more in many programs over using the CPU by itself.\u00a0 Also, the energy efficiency of the GPU is superior.\u00a0 What Jaguar could accomplish in 42 days, can now be done in less than ten with Titan using about the same amount of energy per day!<\/p>\n<p>&nbsp;<\/p>\n<p><strong>K20 Availability<\/strong><\/p>\n<p>The Nvidia Tesla K20 family of GPU accelerators is shipping today and available for order from leading server manufacturers, including Appro, ASUS, Cray, Eurotech, Fujitsu, HP, IBM, Quanta Computer, SGI, Supermicro, T-Platforms and Tyan, as well as from Nvidia&#8217;s reseller partners.<\/p>\n<p><strong>The Future<img loading=\"lazy\" decoding=\"async\" class=\"aligncenter\" src=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/10\/Future.jpg?resize=480%2C283\" alt=\"\" data-recalc-dims=\"1\" \/><\/strong><\/p>\n<p style=\"text-align: left;\">To achieve exascale computing, GPUs have to become much faster as well as more energy efficient.\u00a0 Here is Nvidia&#8217;s latest roadmap which shows Maxwell due to arrive in 2014 on the new 20nm process.<a href=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/Future.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter  wp-image-32494\" title=\"Future\" src=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/Future.jpg?resize=480%2C230\" alt=\"\" srcset=\"http:\/\/i1.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/Future.jpg?resize=480%2C230 800w, http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/Future-300x144.jpg 300w\" sizes=\"auto, (max-width: 480px) 100vw, 480px\" data-recalc-dims=\"1\" \/><\/a><\/p>\n<p><strong>Nvidia&#8217;s Competition<\/strong><\/p>\n<p>Of course, Nvidia is competing against the traditional CPU supercomputers which include IBM.\u00a0 And there is also Intel and AMD.\u00a0 AMD is brand new to supercomputing and they are working quickly to build their own ecosystem of partners using OpenCL as their GPU language.\u00a0 Intel is launching Xeon Phi at SC12 and is no doubt hoping their relationship with x86 will bring them success.\u00a0\u00a0From the\u00a0slide below, Nvidia wishes to remind us that there is no advantage\u00a0in x86\u00a0whatsoever.<a href=\"http:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/xeon-phi.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter  wp-image-32495\" title=\"xeon-phi\" src=\"http:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/xeon-phi.jpg?resize=480%2C247\" alt=\"\" srcset=\"http:\/\/i0.wp.com\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/xeon-phi.jpg?resize=480%2C247 800w, http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/xeon-phi-300x154.jpg 300w\" sizes=\"auto, (max-width: 480px) 100vw, 480px\" data-recalc-dims=\"1\" \/><\/a><\/p>\n<p>What\u00a0gamers are most looking forward to most\u00a0are to the new video cards that will be based on GK110.\u00a0 Since Spring,\u00a0ABT has\u00a0been predicting their arrival after the demand for GK110 Tesla and Quadro are filled and as yields continue to\u00a0improve on the 28nm process.<\/p>\n<p>Nvidia is over two months ahead of schedule in delivering over 18,000 Tesla K20X GK110 accelerators in filling Titan&#8217;s order.\u00a0 We are hoping that they are stockpiling GPUs for their next GeForce and we will keep our readers up-to-date with the very latest news in graphics.\u00a0 Make sure you check out ABT&#8217;s forum where some of the best tech discussions anywhere\u00a0are taking place.<\/p>\n<p>Happy <del>Gaming<\/del> Computing!<\/p>\n<p>&nbsp;<\/p>\n<blockquote style=\"padding-left: 30px;\"><p>Please join us in our <a href=\"http:\/\/alienbabeltech.com\/abt\/index.php\" target=\"_blank\">Forums<\/a><\/p>\n<p>Become a Fan on <a href=\"http:\/\/www.facebook.com\/pages\/AlienBabelTech\/123268467142\" onclick=\"_gaq.push(['_trackEvent', 'outbound-article', 'http:\/\/www.facebook.com\/pages\/AlienBabelTech\/123268467142', 'Facebook']);\" target=\"_blank\">Facebook<\/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","protected":false},"excerpt":{"rendered":"<p><a href=\"http:\/\/alienbabeltech.com\/main\/?p=32488\"><img decoding=\"async\" align=top title=\"Author: Mark Poppin\" src=\"http:\/\/alienbabeltech.com\/main\/wp-content\/uploads\/2012\/11\/TESLAimageLINK.jpg\" \/><\/a><\/p>\n","protected":false},"author":6,"featured_media":32602,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2533,7814,4453],"tags":[248,5887,5219,5888,5121,121,5889,1722,494],"class_list":["post-32488","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-abt-news","category-articles","category-software-2","tag-cuda","tag-gk110","tag-k20","tag-k20x","tag-kepler","tag-nvidia","tag-sc12","tag-supercomputer","tag-titan"],"_links":{"self":[{"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/posts\/32488","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\/6"}],"replies":[{"embeddable":true,"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/comments?post=32488"}],"version-history":[{"count":0,"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/posts\/32488\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/media\/32602"}],"wp:attachment":[{"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/media?parent=32488"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/categories?post=32488"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/alienbabeltech.com\/main\/wp-json\/wp\/v2\/tags?post=32488"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}