{"id":108413,"date":"2023-05-08T12:48:42","date_gmt":"2023-05-08T10:48:43","guid":{"rendered":"https:\/\/prohoster.info\/?p=108413"},"modified":"2023-05-08T16:10:08","modified_gmt":"2023-05-08T14:10:08","slug":"sozdatel-llvm-razvivaet-novyj-yazyk-programmirovaniya-mojo","status":"publish","type":"post","link":"https:\/\/prohoster.info\/sq\/blog\/news\/sozdatel-llvm-razvivaet-novyj-yazyk-programmirovaniya-mojo","title":{"rendered":"Krijuesi i LLVM po zhvillon nj\u00eb gjuh\u00eb programimi t\u00eb re, Mojo","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

Chris Lattner, themelues dhe arkitekt kryesor i LLVM, si dhe krijuesi i gjuh\u00ebs s\u00eb programimit Swift, dhe Tim Davis, ish-drejtor i projekteve AI t\u00eb Google, si Tensorflow dhe JAX, prezantuan nj\u00eb gjuh\u00eb programimi t\u00eb re, Mojo, e cila kombinon thjesht\u00ebsin\u00eb e p\u00ebrdorimit p\u00ebr zhvillimet k\u00ebrkimore dhe krijimin e shpejt\u00eb t\u00eb prototipeve me aft\u00ebsin\u00eb p\u00ebr t\u00eb formuarprodukteve me performanc\u00eb t\u00eb lart\u00eb. Kjo arrihet fal\u00eb p\u00ebrdorimit t\u00eb sintaks\u00ebs s\u00eb njohur t\u00eb gjuh\u00ebs Python, dhe p\u00ebrmes mund\u00ebsis\u00eb p\u00ebr kompilim n\u00eb kodin e makin\u00ebs, mekanizmave t\u00eb pun\u00ebs s\u00eb sigurt me memorie dhe angazhimit t\u00eb mjeteve p\u00ebr akcelerimin harduerik t\u00eb llogaritjeve. <\/p>\n

Projekti \u00ebsht\u00eb i fokusuar n\u00eb p\u00ebrdorimin p\u00ebr zhvillimet n\u00eb fush\u00ebn e m\u00ebsimit makinerik, por paraqitet si nj\u00eb gjuh\u00eb e p\u00ebrgjithshme, q\u00eb zgjeron mund\u00ebsit\u00eb e gjuh\u00ebs Python me mjete t\u00eb programimit t\u00eb sistemit dhe \u00ebsht\u00eb e p\u00ebrshtatshme p\u00ebr nj\u00eb gam\u00eb t\u00eb gjer\u00eb detyrash. P\u00ebr shembull, gjuha \u00ebsht\u00eb e aplikueshme n\u00eb fusha si llogaritjet me performanc\u00eb t\u00eb lart\u00eb, p\u00ebrpunimi dhe transformimi i t\u00eb dh\u00ebnave. Nj\u00eb ve\u00e7ori interesante e Mojo \u00ebsht\u00eb mund\u00ebsia e caktimit t\u00eb simbolit emoji «🔥» si nj\u00eb zgjerim p\u00ebr skedar\u00ebt me kod (p\u00ebr shembull, «helloworld.🔥»), p\u00ebrpos zgjerimit tekstual «.mojo». <\/p>\n

Aktualisht, gjuha \u00ebsht\u00eb n\u00eb faz\u00ebn e zhvillimit intensiv dhe p\u00ebr testim ofrohet vet\u00ebm nj\u00eb nd\u00ebrfaqe online. Nd\u00ebrtesat e ve\u00e7anta p\u00ebr ekzekutimin n\u00eb sisteme lokale pritet t\u00eb publikohen m\u00eb von\u00eb, pas marrjes s\u00eb komenteve n\u00eb lidhje me pun\u00ebn eambientit interaktiv web. Plani \u00ebsht\u00eb q\u00eb t\u00eb hapen kodet burimore t\u00eb kompilatorit, JIT dhe punimeve t\u00eb tjera lidhur me projektin pas p\u00ebrfundimit t\u00eb projektimit t\u00eb arkitektur\u00ebs s\u00eb brendshme (modeli i zhvillimit t\u00eb prototipit funksional pas dyerve t\u00eb mbyllura i ngjason faz\u00ebs fillestare t\u00eb zhvillimit t\u00eb LLVM, Clang dhe Swift). Duke qen\u00eb se sintaksa Mojo bazohet n\u00eb gjuh\u00ebn Python, dhe sistemi i tipeve \u00ebsht\u00eb i af\u00ebrt me C\/C++, n\u00eb t\u00eb ardhmen planifikojn\u00eb t\u00eb zhvillojn\u00eb mjete p\u00ebr t\u00eb thjeshtuar transferimin n\u00eb Mojo t\u00eb projekteve ekzistuese, t\u00eb shkruara n\u00eb C\/C++ dhe Python, si dhe p\u00ebr zhvillimin e projekteve hibride, q\u00eb kombinojn\u00eb kodin n\u00eb Python dhe Mojo. <\/p>\n

Projekti \u00ebsht\u00eb i orientuar drejt shfryt\u00ebzimit t\u00eb burimeve t\u00eb disponueshme t\u00eb harduerit n\u00eb sisteme heterogjene p\u00ebr t\u00eb realizuar llogaritjet. P\u00ebr shembull, p\u00ebr t\u00eb ekzekutuar aplikacione n\u00eb gjuh\u00ebn Mojo dhe p\u00ebr t\u00eb paralelizarar llogaritjet, mund t\u00eb angazhohen GPU-t\u00eb, p\u00ebrshpejtuesit e specializuar p\u00ebr m\u00ebsimin makinerik dhe instrukcionet procesor\u00ebve vektor\u00eb (SIMD). Nj\u00eb nga arsyet p\u00ebr zhvillimin e nj\u00eb n\u00ebngrupi t\u00eb ve\u00e7ant\u00eb t\u00eb gjuh\u00ebs Python, n\u00eb vend q\u00eb t\u00eb p\u00ebrfshihet n\u00eb optimizimin e CPython ekzistues, \u00ebsht\u00eb orientimi drejt kompajlimit, integrimi i mund\u00ebsive t\u00eb programimit sistemor dhe p\u00ebrdorimi i nj\u00eb arkitekture t\u00eb brendshme themelore t\u00eb ndryshme q\u00eb lejon ekzekutimin e kodit n\u00eb GPU dhe p\u00ebrshpejtues t\u00eb ndrysh\u00ebm harduerik\u00eb. Nd\u00ebrkoh\u00eb, zhvilluesit e Mojo p\u00ebrpiqen t\u00eb ruajn\u00eb sa m\u00eb shum\u00eb t\u00eb jet\u00eb e mundur compatibilitetin me CPython. <\/p>\n

Mojo mund t\u00eb p\u00ebrdoret si n\u00eb modin e interpretimeve me p\u00ebrdorimin e JIT, ashtu edhe p\u00ebr kompajlim n\u00eb skedar\u00eb ekzekutiv\u00eb (AOT, ahead-of-time). Kompajleri p\u00ebrmban teknologji moderne p\u00ebr optimizimin automatizuar, caching dhe kompajlim t\u00eb shp\u00ebrndar\u00eb. Kodet burimor n\u00eb gjuh\u00ebn Mojo transformohen n\u00eb kod t\u00eb posht\u00ebm t\u00eb nd\u00ebrmjet\u00ebm MLIR (Multi-Level Intermediate Representation), e zhvilluar nga projekti LLVM dhe q\u00eb ofron mund\u00ebsi shtes\u00eb p\u00ebr optimizimin e p\u00ebrpunimit t\u00eb grafit t\u00eb rrjedh\u00ebs s\u00eb t\u00eb dh\u00ebnave. Kompajleri lejon p\u00ebrdorimin e backward\u00ebve t\u00eb ndrysh\u00ebm p\u00ebr gjenerimin e kodit makinerik, q\u00eb mb\u00ebshtesin MLIR. <\/p>\n

Angazhimi i mekanizmave t\u00eb tjer\u00eb harduerik\u00eb p\u00ebr p\u00ebrshpejtimin e llogaritjeve mund t\u00eb ofroj\u00eb performanc\u00eb q\u00eb gjat\u00eb llogaritjeve intensive kalon aplikacionet n\u00eb C\/C++. P\u00ebr shembull, gjat\u00eb testimit t\u00eb nj\u00eb aplikuese p\u00ebr gjenerimin e nj\u00eb morie Mandelbrot, aplikacioni i kompajluar n\u00eb gjuh\u00ebn Mojo, kur ekzekutohet n\u00eb AWS cloud (r7iz.metal-16xl), \u00ebsht\u00eb treguar 6 her\u00eb m\u00eb i shpejt\u00eb se implementimi n\u00eb C++ (0.03 sek. kundrejt 0.20 sek.), dhe gjithashtu 35 mij\u00eb her\u00eb m\u00eb i shpejt\u00eb se aplikacioni n\u00eb Python duke p\u00ebrdorur CPython standard 3.10.9 (0.03 sek. kundrejt 1027 sek.) dhe 1500 her\u00eb m\u00eb i shpejt\u00eb duke p\u00ebrdorur PYPY (0.03 sek. kundrejt 46.1 sek.). <\/p>\n

Kur vler\u00ebsohet performanca n\u00eb fush\u00ebn e zgjidhjes s\u00eb problemeve t\u00eb m\u00ebsimit makinerik, AI-stack Modular Inference Engine, i shkruar n\u00eb gjuh\u00ebn Mojo, p\u00ebr krahasim me zgjidhjen e bazuar n\u00eb bibliotek\u00ebn TensorFlow u tregua 3 her\u00eb m\u00eb i shpejt\u00eb n\u00eb nj\u00eb sistem me procesor Intel gjat\u00eb p\u00ebrpunimit t\u00eb modelit gjuh\u00ebsor, 6.4 her\u00eb m\u00eb i shpejt\u00eb gjat\u00eb ekzekutimit t\u00eb modelit t\u00eb rekomandimeve dhe 2.1 her\u00eb m\u00eb i shpejt\u00eb gjat\u00eb pun\u00ebs me modelet p\u00ebr p\u00ebrpunimin e informacionit vizual. P\u00ebrdorimi i procesor\u00ebve AMD rezultoi n\u00eb nj\u00eb fitim prej 3.2, 5 dhe 2.2 her\u00eb, nd\u00ebrsa p\u00ebrdorimi i procesor\u00ebve ARM ndihmoi p\u00ebr nj\u00eb p\u00ebrmir\u00ebsim prej 5.3, 7.5 dhe 1.7 her\u00eb, p\u00ebrkat\u00ebsisht. Zgjidhja e bazuar n\u00eb PyTorch mbeti pas Mojo me 1.4, 1.1 dhe 1.5 her\u00eb n\u00eb CPU Intel, me 2.1, 1.2 dhe 1.5 her\u00eb n\u00eb CPU AMD dhe me 4, 4.3 dhe 1.3 her\u00eb n\u00eb CPU ARM.

\"Krijuesi<\/center> <\/p>\n

Gjuha mb\u00ebshtet tipizimin statik dhe mjete p\u00ebr pun\u00eb t\u00eb sigurta me nivel t\u00eb ul\u00ebt t\u00eb memories, t\u00eb ngjashme me mund\u00ebsit\u00eb e gjuh\u00ebs Rust, si ndjekja e koh\u00ebs s\u00eb jet\u00ebs s\u00eb referencave dhe kontrolli i huazimit t\u00eb variablave (borrow checker). P\u00ebrve\u00e7 mjeteve p\u00ebr pun\u00eb t\u00eb sigurta me tregues, gjuha ofron edhe mund\u00ebsi p\u00ebr pun\u00eb t\u00eb nivelit t\u00eb ul\u00ebt, p\u00ebr shembull, \u00ebsht\u00eb e mundur t\u00eb aksesoni drejtp\u00ebrdrejt memorien n\u00eb modalitetin unsafe duke p\u00ebrdorur llojin Pointer, thirrjen e instrukcioneve t\u00eb ve\u00e7anta SIMD ose aksesin n\u00eb zgjerimet harduerike, si TensorCores dhe AMX.

\"Krijuesi<\/center> <\/p>\n

P\u00ebr t\u00eb thjeshtuar ndarjen e kodit klasik dhe t\u00eb optimizuar Python p\u00ebr funksionet me nj\u00eb p\u00ebrcaktim t\u00eb qart\u00eb t\u00eb tipeve p\u00ebr t\u00eb gjith\u00eb variablat, \u00ebsht\u00eb propozuar p\u00ebrdorimi i nj\u00eb fjale ky\u00e7e t\u00eb ve\u00e7ant\u00eb «fn» n\u00eb vend t\u00eb «def». Po ashtu p\u00ebr klasa, kur \u00ebsht\u00eb e nevojshme paketimi static i t\u00eb dh\u00ebnave n\u00eb memorie gjat\u00eb kompilimit (si n\u00eb C), n\u00eb vend t\u00eb «class» mund t\u00eb p\u00ebrdoret tipi «struct». Gjithashtu \u00ebsht\u00eb e mundur importi i thjesht\u00eb i moduleve n\u00eb gjuh\u00ebt C\/C++, p\u00ebr shembull, p\u00ebr t\u00eb importuar funksionin cos nga biblioteka math, mund t\u00eb ceket «from «math.h» import cos».
\n
Burimi: opennet.ru<\/a> <\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"

\u041a\u0440\u0438\u0441 \u041b\u0430\u0442\u043d\u0435\u0440 (Chris Lattner), \u043e\u0441\u043d\u043e\u0432\u0430\u0442\u0435\u043b\u044c \u0438 \u0433\u043b\u0430\u0432\u043d\u044b\u0439 \u0430\u0440\u0445\u0438\u0442\u0435\u043a\u0442\u043e\u0440 LLVM, \u0430 \u0442\u0430\u043a\u0436\u0435 \u0441\u043e\u0437\u0434\u0430\u0442\u0435\u043b\u044c \u044f\u0437\u044b\u043a\u0430 \u043f\u0440\u043e\u0433\u0440\u0430\u043c\u043c\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f Swift, \u0438 \u0422\u0438\u043c \u0414\u044d\u0432\u0438\u0441 (Tim Davis), \u0431\u044b\u0432\u0448\u0438\u0439 \u0440\u0443\u043a\u043e\u0432\u043e\u0434\u0438\u0442\u0435\u043b\u044c AI-\u043f\u0440\u043e\u0435\u043a\u0442\u043e\u0432 Google, \u0442\u0430\u043a\u0438\u0445 \u043a\u0430\u043a Tensorflow \u0438 JAX, \u043f\u0440\u0435\u0434\u0441\u0442\u0430\u0432\u0438\u043b\u0438 \u043d\u043e\u0432\u044b\u0439 \u044f\u0437\u044b\u043a \u043f\u0440\u043e\u0433\u0440\u0430\u043c\u043c\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f Mojo, \u0441\u043e\u0447\u0435\u0442\u0430\u044e\u0449\u0438\u0439 \u043f\u0440\u043e\u0441\u0442\u043e\u0442\u0443 \u043f\u0440\u0438\u043c\u0435\u043d\u0435\u043d\u0438\u044f \u0434\u043b\u044f \u0438\u0441\u0441\u043b\u0435\u0434\u043e\u0432\u0430\u0442\u0435\u043b\u044c\u0441\u043a\u0438\u0445 \u0440\u0430\u0437\u0440\u0430\u0431\u043e\u0442\u043e\u043a \u0438 \u0431\u044b\u0441\u0442\u0440\u043e\u0433\u043e \u0441\u043e\u0437\u0434\u0430\u043d\u0438\u044f \u043f\u0440\u043e\u0442\u043e\u0442\u0438\u043f\u043e\u0432 \u0441 \u043f\u0440\u0438\u0433\u043e\u0434\u043d\u043e\u0441\u0442\u044c\u044e \u0434\u043b\u044f \u0444\u043e\u0440\u043c\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f \u0432\u044b\u0441\u043e\u043a\u043e\u043f\u0440\u043e\u0438\u0437\u0432\u043e\u0434\u0438\u0442\u0435\u043b\u044c\u043d\u044b\u0445 \u043a\u043e\u043d\u0435\u0447\u043d\u044b\u0445 \u043f\u0440\u043e\u0434\u0443\u043a\u0442\u043e\u0432. \u041f\u0435\u0440\u0432\u043e\u0435 \u0434\u043e\u0441\u0442\u0438\u0433\u0430\u0435\u0442\u0441\u044f \u0431\u043b\u0430\u0433\u043e\u0434\u0430\u0440\u044f \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u043e\u0432\u0430\u043d\u0438\u044e […]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":1,"featured_media":108414,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[702],"tags":[],"class_list":["post-108413","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news"],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\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\n","aioseo_head_json":{"title":"\ud83e\udd47Krijuesi i LLVM po zhvillon nj\u00eb gjuh\u00eb t\u00eb re programimi Mojo | ProHoster","description":"Chris Lattner, krijuesi dhe arkitekti kryesor i LLVM, si dhe krijuesi i gjuh\u00ebs s\u00eb programimit Swift, dhe Tim Davis, ish-drejtori i projekteve AI n\u00eb Google, si Tensorflow dhe JAX.","canonical_url":"https:\/\/prohoster.info\/sq\/blog\/news\/sozdatel-llvm-razvivaet-novyj-yazyk-programmirovaniya-mojo","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":null,"og:locale":"sq_AL","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\udd47\u0421\u043e\u0437\u0434\u0430\u0442\u0435\u043b\u044c LLVM \u0440\u0430\u0437\u0432\u0438\u0432\u0430\u0435\u0442 \u043d\u043e\u0432\u044b\u0439 \u044f\u0437\u044b\u043a \u043f\u0440\u043e\u0433\u0440\u0430\u043c\u043c\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f Mojo | ProHoster","og:description":"\u041a\u0440\u0438\u0441 \u041b\u0430\u0442\u043d\u0435\u0440 (Chris Lattner), \u043e\u0441\u043d\u043e\u0432\u0430\u0442\u0435\u043b\u044c \u0438 \u0433\u043b\u0430\u0432\u043d\u044b\u0439 \u0430\u0440\u0445\u0438\u0442\u0435\u043a\u0442\u043e\u0440 LLVM, \u0430 \u0442\u0430\u043a\u0436\u0435 \u0441\u043e\u0437\u0434\u0430\u0442\u0435\u043b\u044c \u044f\u0437\u044b\u043a\u0430 \u043f\u0440\u043e\u0433\u0440\u0430\u043c\u043c\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f Swift, \u0438 \u0422\u0438\u043c \u0414\u044d\u0432\u0438\u0441 (Tim Davis), \u0431\u044b\u0432\u0448\u0438\u0439 \u0440\u0443\u043a\u043e\u0432\u043e\u0434\u0438\u0442\u0435\u043b\u044c AI-\u043f\u0440\u043e\u0435\u043a\u0442\u043e\u0432 Google, \u0442\u0430\u043a\u0438\u0445 \u043a\u0430\u043a Tensorflow \u0438 JAX.","og:url":"https:\/\/prohoster.info\/sq\/blog\/news\/sozdatel-llvm-razvivaet-novyj-yazyk-programmirovaniya-mojo","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":"2023-05-08T10:48:43+00:00","article:modified_time":"2023-05-08T14:10:08+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\/sq\/wp-json\/wp\/v2\/posts\/108413","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/comments?post=108413"}],"version-history":[{"count":0,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/posts\/108413\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/media\/108414"}],"wp:attachment":[{"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/media?parent=108413"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/categories?post=108413"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/tags?post=108413"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}