{"id":6307,"date":"2017-02-21T13:34:25","date_gmt":"2017-02-21T12:34:25","guid":{"rendered":"http:\/\/www.walkingrandomly.com\/?p=6307"},"modified":"2017-02-21T13:34:25","modified_gmt":"2017-02-21T12:34:25","slug":"dell-xps-15-9560-kaby-lake-cpu-pascal-gpu-performance-for-scientific-computing","status":"publish","type":"post","link":"https:\/\/walkingrandomly.com\/?p=6307","title":{"rendered":"Dell XPS 15 9560 (Kaby Lake CPU, Pascal GPU) performance for scientific computing"},"content":{"rendered":"<p>My new toy is a <a href=\"http:\/\/www.windowscentral.com\/dell-xps-15-9560\">2017 Dell XPS 15 9560 <\/a>laptop on which I am running Windows 10. Once I got over (and fixed) the <a href=\"https:\/\/www.walkingrandomly.com\/?p=6294\">annoyance of all the advertising in Windows Home<\/a>, I quickly starting loving this new device.<\/p>\n<p>To get a handle on its performance, I used <a href=\"https:\/\/uk.mathworks.com\/matlabcentral\/fileexchange\/34080-gpubench\">GPUBench<\/a> in MATLAB 2016b and got the following results (This was the best of 4 runs&#8230;I note that MTimes performance for the CPU (Host PC), for example, varied between 130 and 150 Glops).<\/p>\n<ul>\n<li>CPU: Intel Core I7-7700HQ (6M Cache, up to 3.8Ghz)<\/li>\n<li>GPU: NVIDIA GTX 1050 with 4GB GDDR5<\/li>\n<\/ul>\n<p><a href=\"https:\/\/www.walkingrandomly.com\/wp-content\/uploads\/2017\/02\/xps15_run4.png\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone  wp-image-6310\" src=\"https:\/\/www.walkingrandomly.com\/wp-content\/uploads\/2017\/02\/xps15_run4-1024x359.png\" alt=\"xps15_run4\" width=\"729\" height=\"256\" srcset=\"https:\/\/walkingrandomly.com\/wp-content\/uploads\/2017\/02\/xps15_run4-1024x359.png 1024w, https:\/\/walkingrandomly.com\/wp-content\/uploads\/2017\/02\/xps15_run4-300x105.png 300w, https:\/\/walkingrandomly.com\/wp-content\/uploads\/2017\/02\/xps15_run4-768x270.png 768w, https:\/\/walkingrandomly.com\/wp-content\/uploads\/2017\/02\/xps15_run4.png 1376w\" sizes=\"auto, (max-width: 729px) 100vw, 729px\" \/><\/a><\/p>\n<p>I <a href=\"https:\/\/www.walkingrandomly.com\/?p=5702\">last did this for my Retina MacBook Pro <\/a>and am happy to see that the numbers are better across the board. The standout figure for me is the 1206 Gflops (That&#8217;s 1.2 Teraflops!) of single precision performance for Matrix-Matrix Multiply.<\/p>\n<p>That figure of 1.2 Teraflops rang a bell for me and it took me a while to realise why&#8230;..<\/p>\n<p><strong>My laptop vs Manchester University&#8217;s old HPC system &#8211; Horace<\/strong><\/p>\n<p>Old timers like me (I&#8217;m almost 40) like to compare modern hardware with bygone supercomputers (<a href=\"https:\/\/www.walkingrandomly.com\/?p=2684\">1980s Crays vs mobile phones<\/a>\u00a0for example) and we\u00a0know we are truly old when the numbers coming out of laptop benchmarks match the peak theoretical performance of institutional HPC systems we actually used as part of our career.<\/p>\n<p>This has now finally happened to me! I was at the University of Manchester when it commissioned a HPC service called Horace and I was there when it was switched off in 2010 (only 6 and a bit years ago!). It was the University&#8217;s primary HPC service with\u00a0a support team, helpdesk, sysadmins&#8230;the lot.\u00a0 The specs are still available on <a href=\"http:\/\/www.rcs.manchester.ac.uk\/services\/computational\/Horace\">Manchester&#8217;s website<\/a>:<\/p>\n<ul>\n<li>24 nodes, each with 8 cores giving 192 cores in total.<\/li>\n<li>Each core had a theoretical peak compute performance of 6.4 double precision Gflop\/s<\/li>\n<li>So a node had a theoretical peak performance of 51.2 Gflop\/s<\/li>\n<li>The whole thing could theoretically manage 1.2 Teraflop\/s<\/li>\n<li>It had four special &#8216;high memory&#8217; nodes with 32Gb RAM each<\/li>\n<\/ul>\n<p>Good luck getting that 1.2 Teraflops out of it in practice!<\/p>\n<p>I get a big geek-kick out of the fact that my new laptop has the same amount of RAM as one of these &#8216;big memory&#8217; nodes and that\u00a0my laptop&#8217;s\u00a0double precision CPU performance is on par with the combined power of 3\u00a0of Horace&#8217;s nodes. Furthermore, my laptop&#8217;s GPU\u00a0can just about manage\u00a01.2 Teraflop\/s of <strong>single precision <\/strong>\u00a0performance in MATLAB &#8212; on par with the total combined power of the HPC system*.<\/p>\n<p>* (I know, I know&#8230;.Horace&#8217;s numbers are for\u00a0<strong>double precision <\/strong>and my GPU\u00a0numbers are <strong>single precision<\/strong> &#8212; apples to oranges &#8212; but it still astonishes me that the headline numbers are the same &#8212; 1.2 Teraflops).<\/p>\n<p>&nbsp;<\/p>\n","protected":false},"excerpt":{"rendered":"<p>My new toy is a 2017 Dell XPS 15 9560 laptop on which I am running Windows 10. Once I got over (and fixed) the annoyance of all the advertising in Windows Home, I quickly starting loving this new device. To get a handle on its performance, I used GPUBench in MATLAB 2016b and got [&hellip;]<\/p>\n","protected":false},"author":3,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"jetpack_post_was_ever_published":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":[51,68,11,41],"tags":[],"class_list":["post-6307","post","type-post","status-publish","format-standard","hentry","category-gpu","category-hpc","category-matlab","category-parallel-programming"],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_shortlink":"https:\/\/wp.me\/p3swhs-1DJ","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/walkingrandomly.com\/index.php?rest_route=\/wp\/v2\/posts\/6307","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/walkingrandomly.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/walkingrandomly.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/walkingrandomly.com\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/walkingrandomly.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=6307"}],"version-history":[{"count":3,"href":"https:\/\/walkingrandomly.com\/index.php?rest_route=\/wp\/v2\/posts\/6307\/revisions"}],"predecessor-version":[{"id":6312,"href":"https:\/\/walkingrandomly.com\/index.php?rest_route=\/wp\/v2\/posts\/6307\/revisions\/6312"}],"wp:attachment":[{"href":"https:\/\/walkingrandomly.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6307"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/walkingrandomly.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6307"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/walkingrandomly.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6307"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}