{"id":393581,"date":"2026-08-19T09:32:32","date_gmt":"2026-08-19T08:32:32","guid":{"rendered":"https:\/\/droix.net\/blogs\/?p=365004"},"modified":"2026-09-27T06:50:10","modified_gmt":"2026-09-27T05:50:10","slug":"geekom-demonstre-cetru-mezglu-mini-pc-ai-klasteri-ar-vietejo-deepseek-modeli","status":"publish","type":"post","link":"https:\/\/droix.net\/blogs\/lv\/geekom-demonstre-cetru-mezglu-mini-pc-ai-klasteri-ar-vietejo-deepseek-modeli\/","title":{"rendered":"GEEKOM demonstr\u0113 \u010detru mezglu Mini PC AI klasteri ar viet\u0113jo DeepSeek modeli"},"content":{"rendered":"\n

GEEKOM<\/a> ir demonstr\u0113jis \u010detru mezglu AI klasteri, kas veidots no t\u0101 kompaktajiem A9 Mega Mini PC, par\u0101dot, k\u0101 maza izm\u0113ra sist\u0113mas var apvienot viet\u0113j\u0101m uz\u0146\u0113mumu AI lietojumprogramm\u0101m.<\/p>\n\n\n\n

Klasteris savieno \u010detras A9 Mega<\/a> vien\u012bbas caur USB4, \u013caujot t\u0101m darboties k\u0101 sadal\u012btai platformai bez tradicion\u0101la serveru skapja vai patent\u0113ta \u0101trgaitas t\u012bkla apr\u012bkojuma. Katru sist\u0113mu darbina AMD Ryzen AI Max+ 395 <\/a>procesors ar 16 Zen 5 CPU kodoliem, Radeon 8060S grafiku un vienoto atmi\u0146u.<\/p>\n\n\n\n

\n
\"\u010cetras<\/a>
\u010cetras A9 Mega vien\u012bbas<\/figcaption><\/figure>\n\n\n\n
\"\u010cetras<\/a>
\u010cetras A9 Mega vien\u012bbas<\/figcaption><\/figure>\n<\/figure>\n\n\n\n

Darbojoties ar Ubuntu, ROCm un DwarfStar programmat\u016bru, klasteris sadala DeepSeek V4 Flash modeli vis\u0101s \u010detr\u0101s ma\u0161\u012bn\u0101s. Tas nodro\u0161ina ar\u012b OpenAI sader\u012bgu API, \u013caujot integr\u0113t lietojumprogrammas, AI a\u0123entus un automatiz\u0101cijas r\u012bkus.<\/p>\n\n\n\n

Galven\u0101 priek\u0161roc\u012bba ir viet\u0113j\u0101 apstr\u0101de. Uz\u0146\u0113mumi var analiz\u0113t dokumentus, pirmkodu, l\u012bgumus un iek\u0161\u0113j\u0101s zin\u0101\u0161anas, nenos\u016btot sensit\u012bvu inform\u0101ciju publiskaj\u0101 m\u0101kon\u012b. GEEKOM nor\u0101da, ka sist\u0113ma ir test\u0113ta ar\u012b ar konteksta logiem l\u012bdz 250 000 mar\u0137ieriem, padarot to piem\u0113rotu gariem dokumentiem un sare\u017e\u0123\u012bt\u0101m darbpl\u016bsm\u0101m.<\/p>\n\n\n\n

\n
\"\u010cetras<\/a>
\u010cetras GEEKOM A9 Mega vien\u012bbas<\/figcaption><\/figure>\n\n\n\n
\"\u010cetras<\/a>
\u010cetras A9 Mega vien\u012bbas, skats no aizmugures<\/figcaption><\/figure>\n<\/figure>\n\n\n\n

Saska\u0146\u0101 ar uz\u0146\u0113muma teikto klasteris sasniedza aptuveni 14,61 mar\u0137ieri sekund\u0113 pie vienas vienlaic\u012bbas, ar 95. procentiles laiku l\u012bdz pirmajam mar\u0137ierim aptuveni 0,42 sekundes t\u0101 32 un 128 mar\u0137ieru testos. Organiz\u0101cijas var s\u0101kt ar vienu vai div\u0101m A9 Mega<\/a> sist\u0113m\u0101m un papla\u0161in\u0101t l\u012bdz \u010detr\u0101m, pieaugot pras\u012bb\u0101m.<\/p>\n\n\n\n

Demonstr\u0101cija izce\u013c kompakto datoru potenci\u0101lu k\u0101 m\u0113rogojam\u0101m viet\u0113j\u0101m AI platform\u0101m uz\u0146\u0113mumiem, laboratorij\u0101m, klas\u0113m un malas skait\u013co\u0161anas vid\u0113m.<\/p>\n\n\n\n

Avots: GEEKOM, izmantojot PR Newswire<\/a>. Veiktsp\u0113jas r\u0101d\u012bt\u0101ji un iesp\u0113jas ir balst\u012bti uz uz\u0146\u0113muma zi\u0146otajiem testiem.<\/em><\/p>\n","protected":false},"excerpt":{"rendered":"

GEEKOM ir demonstr\u0113jis \u010detru mezglu AI klasteri, kas veidots no t\u0101 kompaktajiem A9 Mega Mini PC. Savienots caur USB4 un darbin\u0101ts ar AMD Ryzen AI Max+ 395 procesoriem, sist\u0113ma lok\u0101li darbojas ar DeepSeek V4 Flash, pied\u0101v\u0101jot uz\u0146\u0113mumiem m\u0113rogojamu veidu, k\u0101 apstr\u0101d\u0101t sensit\u012bvus dokumentus, kodu un AI darba slodzes, piln\u012bb\u0101 nepa\u013caujoties uz m\u0101ko\u0146a infrastrukt\u016bru.<\/p>\n","protected":false},"author":859,"featured_media":393582,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2508,2983],"tags":[32639,3283,2617],"class_list":["post-393581","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-zinas","category-mini-dators-lv","tag-a9-mega-16","tag-geekom-lv","tag-mini-dators"],"lang":"lv","translations":{"lv":393581,"en":365004,"de":370301,"en-gb":370672,"en-ie":370679,"en-au":370686,"fr":371387,"ar":382327,"bg":384272,"cs":385446,"es":387053,"da":387374,"pt-pt":388137,"hr":389905,"hu":390629,"el":403888,"et":403954,"fi":404020,"it":404085,"lt":404150,"nl":404259,"no":404325,"pl":404390,"ro":404456,"sk":404522,"sl":404588,"sv":404653},"pll_sync_post":{},"_links":{"self":[{"href":"https:\/\/droix.net\/blogs\/wp-json\/wp\/v2\/posts\/393581","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/droix.net\/blogs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/droix.net\/blogs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/droix.net\/blogs\/wp-json\/wp\/v2\/users\/859"}],"replies":[{"embeddable":true,"href":"https:\/\/droix.net\/blogs\/wp-json\/wp\/v2\/comments?post=393581"}],"version-history":[{"count":1,"href":"https:\/\/droix.net\/blogs\/wp-json\/wp\/v2\/posts\/393581\/revisions"}],"predecessor-version":[{"id":393587,"href":"https:\/\/droix.net\/blogs\/wp-json\/wp\/v2\/posts\/393581\/revisions\/393587"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/droix.net\/blogs\/wp-json\/wp\/v2\/media\/393582"}],"wp:attachment":[{"href":"https:\/\/droix.net\/blogs\/wp-json\/wp\/v2\/media?parent=393581"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/droix.net\/blogs\/wp-json\/wp\/v2\/categories?post=393581"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/droix.net\/blogs\/wp-json\/wp\/v2\/tags?post=393581"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}