{"id":181639,"date":"2026-05-21T18:25:05","date_gmt":"2026-05-21T16:25:05","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/fluidx3d-v3-7"},"modified":"2026-05-21T18:25:05","modified_gmt":"2026-05-21T16:25:05","slug":"fluidx3d-v3-7","status":"publish","type":"post","link":"https:\/\/prohoster.info\/sv\/blog\/news\/fluidx3d-v3-7","title":{"rendered":"FluidX3D v3.7","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

<\/p>\n

FluidX3D 3.7<\/a> \u2013 paket f\u00f6r ber\u00e4kningshydrodynamik som anv\u00e4nder lattice Boltzmann-metoden (LBM) och OpenCL. Projektet positioneras som ett h\u00f6gpresterande och minnesekonomiskt CFD-paket som kan k\u00f6ras p\u00e5 GPU:er och CPU:er fr\u00e5n olika tillverkare via OpenCL.<\/p>\n

<\/noindex><\/noindex><\/p>\n

Den st\u00f6rsta \u00e4ndringen i version 3.7 \u00e4r optimeringen av k\u00e4rnan f\u00f6r rendering av isop surfaces Q-kriteriet (graphics_q()), som anv\u00e4nds f\u00f6r att visualisera virvelstrukturer i fl\u00f6det. I den tidigare implementationen laddade k\u00e4rnan ett tungt m\u00f6nster med 32 tre-dimensionella hastigheter f\u00f6r varje cell i n\u00e4tet fr\u00e5n videominnet. Nu laddar en arbetstr\u00e5dgrupp i storlek 8\u00d78\u00d78 ett block med 11\u00d711\u00d711 hastigheter i 16 KB lokal minne av f\u00f6rsta niv\u00e5n, vilket sedan g\u00f6r att n\u00e4rliggande tr\u00e5dar kan \u00e5teranv\u00e4nda dessa data. F\u00f6rfattaren uppskattar att detta minskar antalet laddningar fr\u00e5n VRAM med cirka 12 g\u00e5nger \u2014 till 2,6 hastigheter per cell \u2014 och flyttar k\u00e4rnan fr\u00e5n ett minnesbandbreddsbegr\u00e4nsat l\u00e4ge till ett ber\u00e4kningsbegr\u00e4nsat l\u00e4ge.<\/p>\n

Den praktiska vinsten beror p\u00e5 den specifika h\u00e5rdvaran. P\u00e5 nya GPU:er med avancerade cache-system kan manuell optimering av det lokala minnet n\u00e4stan ge ingen \u00f6kning eller till och med minska bildhastigheten n\u00e5got. P\u00e5 vissa \u00e4ldre GPU:er har f\u00f6rfattaren noterat en acceleration p\u00e5 upp till tv\u00e5 g\u00e5nger, vilket s\u00e4rskilt m\u00e4rks i den interaktiva grafiska l\u00e4get FluidX3D. Till exempel har offentliggjorda tester visat att NVIDIA Titan Xp accelererats fr\u00e5n 7 till 13 bilder\/s, Intel UHD 770 fr\u00e5n 8 till 14 bilder\/s i ett av l\u00e4gena, och Intel Arc A750 fr\u00e5n 22 till 26 bilder\/s.<\/p>\n

F\u00f6r \u00e4ldre integrerade GPU:er har en automatisk tillbakag\u00e5ng till den tidigare implementationen f\u00f6rsetts: vissa av dessa enheter st\u00f6djer inte arbetstr\u00e5dgrupper med en storlek av 512 str\u00f6mmar eller har mindre \u00e4n 16 KB lokal minne L1. I dessa fall kommer FluidX3D att forts\u00e4tta anv\u00e4nda den gamla varianten av k\u00e4rnan, vilket bevarar kompatibiliteten.<\/p>\n

Denna version har \u00e4ven f\u00f6renklat villkoren f\u00f6r VR-rendering, st\u00e4ngt av loop-expansion i rasteriseringen f\u00f6r att minska \"spaghettin\" i assemblerutmatningen, lagt till mikrooptimering i convert_triangle\/_interpolated(), och r\u00e4ttat ett fel i OpenCL-adaptern som uppstod vid tomt k\u00e4rnnamn och storlek p\u00e5 arbetstr\u00e5dgrupp.<\/p>\n

<\/p>\n

FluidX3D inkluderar en inbyggd interaktiv 3D-renderare som g\u00f6r det m\u00f6jligt att visualisera hastighetsf\u00e4lt, fl\u00f6deslinjer, isopartitioner av Q-kriteriet och skivor direkt under ber\u00e4kningen, utan behov av att exportera till ett separat efterbehandlingsverktyg. FluidX3D<\/a> Projektet inkluderar ocks\u00e5 st\u00f6d f\u00f6r fria ytor, termisk konvektion, Smagorinsky-Lilly turbulensmodeller, partikelf\u00f6ljning, VTK-export och hantering av flera GPU:er.<\/p>\n

<\/noindex>
\n<\/p>\n

Det \u00e4r s\u00e4rskilt v\u00e4rt att notera licensstatusen: FluidX3D:s k\u00e4llkod \u00e4r \u00f6ppen och tillg\u00e4nglig p\u00e5 GitHub, men projektets licens<\/a> \u00e4r inte fri i strikt mening enligt OSI\/FSF. Den till\u00e5ter anv\u00e4ndning f\u00f6r offentliga forskningar, utbildning och personliga \u00e4ndam\u00e5l, men f\u00f6rbjuder kommersiell och milit\u00e4r anv\u00e4ndning samt tr\u00e4ning av AI-modeller p\u00e5 k\u00e4llkoden.<\/p>\n

<\/noindex><\/p>\n

K\u00e4lla: linux.org.ru<\/a><\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"

FluidX3D 3.7 – \u043f\u0430\u043a\u0435\u0442 \u0434\u043b\u044f \u0432\u044b\u0447\u0438\u0441\u043b\u0438\u0442\u0435\u043b\u044c\u043d\u043e\u0439 \u0433\u0438\u0434\u0440\u043e\u0434\u0438\u043d\u0430\u043c\u0438\u043a\u0438, \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u0443\u044e\u0449\u0438\u0439 \u043c\u0435\u0442\u043e\u0434 \u0440\u0435\u0448\u0451\u0442\u043e\u0447\u043d\u044b\u0445 \u0443\u0440\u0430\u0432\u043d\u0435\u043d\u0438\u0439 \u0411\u043e\u043b\u044c\u0446\u043c\u0430\u043d\u0430 (LBM) \u0438 OpenCL. \u041f\u0440\u043e\u0435\u043a\u0442 \u043f\u043e\u0437\u0438\u0446\u0438\u043e\u043d\u0438\u0440\u0443\u0435\u0442\u0441\u044f \u043a\u0430\u043a \u0432\u044b\u0441\u043e\u043a\u043e\u043f\u0440\u043e\u0438\u0437\u0432\u043e\u0434\u0438\u0442\u0435\u043b\u044c\u043d\u044b\u0439 \u0438 \u044d\u043a\u043e\u043d\u043e\u043c\u043d\u044b\u0439 \u043f\u043e \u043f\u0430\u043c\u044f\u0442\u0438 CFD-\u043f\u0430\u043a\u0435\u0442, \u0441\u043f\u043e\u0441\u043e\u0431\u043d\u044b\u0439 \u0440\u0430\u0431\u043e\u0442\u0430\u0442\u044c \u043d\u0430 GPU \u0438 CPU \u0440\u0430\u0437\u043d\u044b\u0445 \u043f\u0440\u043e\u0438\u0437\u0432\u043e\u0434\u0438\u0442\u0435\u043b\u0435\u0439 \u0447\u0435\u0440\u0435\u0437 OpenCL. \u0413\u043b\u0430\u0432\u043d\u044b\u043c \u0438\u0437\u043c\u0435\u043d\u0435\u043d\u0438\u0435\u043c \u0432\u0435\u0440\u0441\u0438\u0438 3.7 \u0441\u0442\u0430\u043b\u0430 \u043e\u043f\u0442\u0438\u043c\u0438\u0437\u0430\u0446\u0438\u044f \u044f\u0434\u0440\u0430 \u043e\u0442\u0440\u0438\u0441\u043e\u0432\u043a\u0438 \u0438\u0437\u043e\u043f\u043e\u0432\u0435\u0440\u0445\u043d\u043e\u0441\u0442\u0438 Q-\u043a\u0440\u0438\u0442\u0435\u0440\u0438\u044f (graphics_q()), \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u0443\u0435\u043c\u043e\u0439 \u0434\u043b\u044f \u0432\u0438\u0437\u0443\u0430\u043b\u0438\u0437\u0430\u0446\u0438\u0438 \u0432\u0438\u0445\u0440\u0435\u0432\u044b\u0445 \u0441\u0442\u0440\u0443\u043a\u0442\u0443\u0440 \u0432 \u043f\u043e\u0442\u043e\u043a\u0435. \u0412 \u043f\u0440\u0435\u0436\u043d\u0435\u0439 \u0440\u0435\u0430\u043b\u0438\u0437\u0430\u0446\u0438\u0438 […]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":8,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[702],"tags":[],"class_list":["post-181639","post","type-post","status-publish","format-standard","hentry","category-news"],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\t\n\t\n\t\n\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\n","aioseo_head_json":{"title":"\ud83e\udd47FluidX3D v3.7 | ProHoster","description":"FluidX3D 3.7 \u2013 en mjukvarupaket f\u00f6r numerisk hydrodynamik, som anv\u00e4nder metod f\u00f6r Boltzmann-latticeekvationer (LBM) och OpenCL.","canonical_url":"https:\/\/prohoster.info\/sv\/blog\/news\/fluidx3d-v3-7","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":null,"og:locale":"sv_SE","og:site_name":"ProHoster | \u041a\u0443\u043f\u0438\u0442\u044c \u043d\u0430\u0434\u0435\u0436\u043d\u044b\u0439 \u0445\u043e\u0441\u0442\u0438\u043d\u0433 \u0434\u043b\u044f \u0441\u0430\u0439\u0442\u043e\u0432 \u0441 \u0437\u0430\u0449\u0438\u0442\u043e\u0439 \u043e\u0442 DDoS, VPS VDS \u0441\u0435\u0440\u0432\u0435\u0440\u044b","og:type":"article","og:title":"\ud83e\udd47FluidX3D v3.7 | ProHoster","og:description":"FluidX3D 3.7 \u2013 \u043f\u0430\u043a\u0435\u0442 \u0434\u043b\u044f \u0432\u044b\u0447\u0438\u0441\u043b\u0438\u0442\u0435\u043b\u044c\u043d\u043e\u0439 \u0433\u0438\u0434\u0440\u043e\u0434\u0438\u043d\u0430\u043c\u0438\u043a\u0438, \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u0443\u044e\u0449\u0438\u0439 \u043c\u0435\u0442\u043e\u0434 \u0440\u0435\u0448\u0451\u0442\u043e\u0447\u043d\u044b\u0445 \u0443\u0440\u0430\u0432\u043d\u0435\u043d\u0438\u0439 \u0411\u043e\u043b\u044c\u0446\u043c\u0430\u043d\u0430 (LBM) \u0438 OpenCL.","og:url":"https:\/\/prohoster.info\/sv\/blog\/news\/fluidx3d-v3-7","og:image":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:secure_url":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:width":350,"og:image:height":350,"article:published_time":"2026-05-21T16:25:05+00:00","article:modified_time":"2026-05-21T16:25:05+00:00","article:publisher":"https:\/\/www.facebook.com\/prohoster","article:author":"https:\/\/www.facebook.com\/prohoster"},"aioseo_meta_data":[],"gt_translate_keys":[{"key":"link","format":"url"}],"_links":{"self":[{"href":"https:\/\/prohoster.info\/sv\/wp-json\/wp\/v2\/posts\/181639","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/prohoster.info\/sv\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/prohoster.info\/sv\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/prohoster.info\/sv\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/prohoster.info\/sv\/wp-json\/wp\/v2\/comments?post=181639"}],"version-history":[{"count":0,"href":"https:\/\/prohoster.info\/sv\/wp-json\/wp\/v2\/posts\/181639\/revisions"}],"wp:attachment":[{"href":"https:\/\/prohoster.info\/sv\/wp-json\/wp\/v2\/media?parent=181639"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/prohoster.info\/sv\/wp-json\/wp\/v2\/categories?post=181639"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/prohoster.info\/sv\/wp-json\/wp\/v2\/tags?post=181639"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}