{"id":1067,"date":"2016-11-15T18:08:02","date_gmt":"2016-11-15T09:08:02","guid":{"rendered":"http:\/\/rageworx.info\/?p=1067"},"modified":"2016-11-15T18:28:01","modified_gmt":"2016-11-15T09:28:01","slug":"raw-scale-process-speed-at-intel-avx-instruction","status":"publish","type":"post","link":"https:\/\/rageworx.info\/?p=1067","title":{"rendered":"RAW scale process speed at Intel AVX instruction."},"content":{"rendered":"<p><a href=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/code_thumb_rawscale.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-1079\" src=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/code_thumb_rawscale.jpg\" alt=\"code_thumb_rawscale\" width=\"400\" height=\"344\" \/><\/a><\/p>\n<h3>Testbase<\/h3>\n<p>My medical RAW image scale processor benchmarked for Intel AVX instruction set. It was fully reprogrammed for be affected by optimization algorithm by compiler option.<\/p>\n<h3><\/h3>\n<h3>Specification of benchmark<\/h3>\n<ul>\n<li>CPU: Intel Core i7-6700 @ 4GHz (Skylake-S)<\/li>\n<li>Memory: DDR4-1066 16GB (Single channel)<\/li>\n<li>OS: Windows 7 professional 64bit<\/li>\n<\/ul>\n<h3><\/h3>\n<h3>Compiler<\/h3>\n<ul>\n<li>MinGW-W64 64bit version 4.8.0<\/li>\n<\/ul>\n<h3><\/h3>\n<h3>Compiler options<\/h3>\n<ul>\n<li>-fomit-frame-pointer<\/li>\n<li>-fexpensive-optimizations<\/li>\n<li>-O3<\/li>\n<li>-s<\/li>\n<li>-m64<\/li>\n<li>-march=corei7-avx<\/li>\n<\/ul>\n<h3><\/h3>\n<h3>RAW image source<\/h3>\n<ul>\n<li>3072&#215;3072, 16bit grayscale X-Ray image<\/li>\n<\/ul>\n<p><a href=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/rawviewer_sample.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-1073\" src=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/rawviewer_sample.jpg\" alt=\"rawviewer_sample\" width=\"400\" height=\"464\" \/><\/a><\/p>\n<h3>Benchmark rule<\/h3>\n<ul>\n<li>Scale each different filter for double size.<\/li>\n<\/ul>\n<p>&nbsp;<\/p>\n<h3>Results<\/h3>\n<ul>\n<li>Nearest (box) scale : 0.597 seconds.<br \/>\n<h3><a href=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/bench_result_nearest.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-1077\" src=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/bench_result_nearest.jpg\" alt=\"bench_result_nearest\" width=\"432\" height=\"168\" \/><\/a><\/h3>\n<\/li>\n<\/ul>\n<ul>\n<li>Bilinear scale : 0.597 seconds.<br \/>\n<a href=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/bench_result_bilinear.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-1075\" src=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/bench_result_bilinear.jpg\" alt=\"bench_result_bilinear\" width=\"433\" height=\"171\" \/><\/a><\/li>\n<li>Bicubic scale : 0.693 seconds.<br \/>\n<a href=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/bench_result_bicubic.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-1074\" src=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/bench_result_bicubic.jpg\" alt=\"bench_result_bicubic\" width=\"435\" height=\"171\" \/><\/a><\/li>\n<li>Lanzcos 3 taps : 0.884 seconds.<br \/>\n<a href=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/bench_result_lanczos3.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-1076\" src=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/bench_result_lanczos3.jpg\" alt=\"bench_result_lanczos3\" width=\"430\" height=\"171\" \/><\/a><\/li>\n<\/ul>\n<p>Specially lanzcos 3 scale took 30% more speed at under 1 second. So I looking fore more faster way how it make faster than now. And it is almost same with B-Spline filter, or sometimes B-Spline faster than lanczos3, Hmmm &#8230;<\/p>\n<h3>Download test console program<\/h3>\n<ul>\n<li><a href=\"http:\/\/rageworx.info\/wp-content\/uploads\/2016\/11\/raw_resize_win64_console_20161115a.7z\">raw_resize_win64_console_20161115a<\/a><\/li>\n<li>You may sure run this program on Windows 64bit.<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>Testbase My medical RAW image scale processor benchmarked for Intel AVX instruction set. It was fully reprogrammed for be affected by optimization algorithm by compiler option. Specification of benchmark CPU: Intel Core i7-6700 @ 4GHz (Skylake-S) Memory: DDR4-1066 16GB (Single channel) OS: Windows 7 professional 64bit Compiler MinGW-W64 64bit version&#8230; <a href=\"https:\/\/rageworx.info\/?p=1067\">Read more &raquo;<\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[5,8,3],"tags":[411,414,195,413,412],"class_list":["post-1067","post","type-post","status-publish","format-standard","hentry","category-development","category-diy","category-raphs","tag-dental","tag-fast","tag-image","tag-scaling","tag-x-ray"],"_links":{"self":[{"href":"https:\/\/rageworx.info\/index.php?rest_route=\/wp\/v2\/posts\/1067","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/rageworx.info\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/rageworx.info\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/rageworx.info\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/rageworx.info\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1067"}],"version-history":[{"count":0,"href":"https:\/\/rageworx.info\/index.php?rest_route=\/wp\/v2\/posts\/1067\/revisions"}],"wp:attachment":[{"href":"https:\/\/rageworx.info\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1067"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/rageworx.info\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=1067"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/rageworx.info\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=1067"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}