{"id":460473,"date":"2025-05-22T09:00:06","date_gmt":"2025-05-22T09:00:06","guid":{"rendered":"http:\/\/savepearlharbor.com\/?p=460473"},"modified":"-0001-11-30T00:00:00","modified_gmt":"-0001-11-29T21:00:00","slug":"","status":"publish","type":"post","link":"https:\/\/savepearlharbor.com\/?p=460473","title":{"rendered":"<span>Red Hat \u043f\u0440\u0435\u0434\u0441\u0442\u0430\u0432\u0438\u043b\u0430 \u043e\u0442\u043a\u0440\u044b\u0442\u044b\u0439 \u043f\u0440\u043e\u0435\u043a\u0442 llm-d \u2014 \u044d\u0442\u043e \u0440\u0430\u0441\u043f\u0440\u0435\u0434\u0435\u043b\u0451\u043d\u043d\u0430\u044f \u0441\u0442\u0440\u0443\u043a\u0442\u0443\u0440\u0430 \u0432\u044b\u0432\u043e\u0434\u0430 LLM \u043d\u0430 \u043e\u0441\u043d\u043e\u0432\u0435 Kubernetes<\/span>"},"content":{"rendered":"<div><!--[--><!--]--><\/div>\n<div id=\"post-content-body\">\n<div>\n<div class=\"article-formatted-body article-formatted-body article-formatted-body_version-2\">\n<div xmlns=\"http:\/\/www.w3.org\/1999\/xhtml\">\n<figure class=\"full-width\"><img decoding=\"async\" src=\"https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/upload_files\/fc0\/431\/b42\/fc0431b42302472c66155b8543978a61.png\" width=\"900\" height=\"482\" sizes=\"auto, (max-width: 780px) 100vw, 50vw\" srcset=\"https:\/\/habrastorage.org\/r\/w780\/getpro\/habr\/upload_files\/fc0\/431\/b42\/fc0431b42302472c66155b8543978a61.png 780w,&#10;       https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/upload_files\/fc0\/431\/b42\/fc0431b42302472c66155b8543978a61.png 781w\" loading=\"lazy\" decode=\"async\"\/><\/figure>\n<p>Red Hat <a href=\"https:\/\/www.redhat.com\/en\/about\/press-releases\/red-hat-launches-llm-d-community-powering-distributed-gen-ai-inference-scale\" rel=\"noopener noreferrer nofollow\">\u043f\u0440\u0435\u0434\u0441\u0442\u0430\u0432\u0438\u043b\u0430<\/a> \u043e\u0442\u043a\u0440\u044b\u0442\u044b\u0439 <a href=\"https:\/\/llm-d.ai\/blog\/llm-d-announce\" rel=\"noopener noreferrer nofollow\">\u043f\u0440\u043e\u0435\u043a\u0442 llm\u2011d<\/a>. \u042d\u0442\u043e \u0432\u044b\u0441\u043e\u043a\u043e\u043f\u0440\u043e\u0438\u0437\u0432\u043e\u0434\u0438\u0442\u0435\u043b\u044c\u043d\u0430\u044f \u0440\u0430\u0441\u043f\u0440\u0435\u0434\u0435\u043b\u0451\u043d\u043d\u0430\u044f \u0441\u0442\u0440\u0443\u043a\u0442\u0443\u0440\u0430 \u0432\u044b\u0432\u043e\u0434\u0430 LLM \u043d\u0430\u00a0\u043e\u0441\u043d\u043e\u0432\u0435 Kubernetes. \u0418\u0441\u0445\u043e\u0434\u043d\u044b\u0439 \u043a\u043e\u0434 \u0440\u0435\u0448\u0435\u043d\u0438\u044f <a href=\"https:\/\/github.com\/llm-d\/llm-d\" rel=\"noopener noreferrer nofollow\">\u043e\u043f\u0443\u0431\u043b\u0438\u043a\u043e\u0432\u0430\u043d<\/a> \u043d\u0430\u00a0GitHub \u043f\u043e\u0434\u00a0\u043b\u0438\u0446\u0435\u043d\u0437\u0438\u0435\u0439 Apache License 2.0.<\/p>\n<p>\u041f\u0440\u043e\u0435\u043a\u0442 llm\u2011d \u043d\u0430\u043f\u0440\u0430\u0432\u043b\u0435\u043d \u043d\u0430\u00a0\u0440\u0430\u0441\u0448\u0438\u0440\u0435\u043d\u0438\u0435 \u0432\u043e\u0437\u043c\u043e\u0436\u043d\u043e\u0441\u0442\u0435\u0439 \u0440\u0430\u0441\u043f\u0440\u0435\u0434\u0435\u043b\u0451\u043d\u043d\u043e\u0433\u043e \u0432\u044b\u0432\u043e\u0434\u0430 \u0438 \u043c\u0430\u0441\u0448\u0442\u0430\u0431\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f \u0441\u0438\u0441\u0442\u0435\u043c \u0433\u0435\u043d\u0435\u0440\u0430\u0442\u0438\u0432\u043d\u043e\u0433\u043e \u0418\u0418. <\/p>\n<p>\u0420\u0435\u0448\u0435\u043d\u0438\u0435 llm\u2011d \u0440\u0430\u0437\u0440\u0430\u0431\u0430\u0442\u044b\u0432\u0430\u0435\u0442\u0441\u044f Red Hat \u0441\u043e\u0432\u043c\u0435\u0441\u0442\u043d\u043e \u0441\u00a0Nvidia, AMD, Intel, IBM Research, Google Cloud, CoreWeave, Hugging Face \u0438 \u0434\u0440\u0443\u0433\u0438\u043c\u0438 \u043f\u043e\u0441\u0442\u0430\u0432\u0449\u0438\u043a\u0430\u043c\u0438 \u0438 \u043e\u0440\u0433\u0430\u043d\u0438\u0437\u0430\u0446\u0438\u044f\u043c\u0438 \u0432\u00a0\u0441\u0444\u0435\u0440\u0435 \u0418\u0418. <\/p>\n<p>\u041f\u0440\u043e\u0433\u0440\u0430\u043c\u043c\u043d\u043e\u0435 \u043e\u0431\u0435\u0441\u043f\u0435\u0447\u0435\u043d\u0438\u0435 llm\u2011d \u0441\u043e\u0437\u0434\u0430\u043d\u043e \u043d\u0430\u00a0\u043e\u0441\u043d\u043e\u0432\u0435 Kubernetes \u0438 \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u0443\u0435\u0442 vLLM \u0434\u043b\u044f\u00a0\u0440\u0430\u0441\u043f\u0440\u0435\u0434\u0435\u043b\u0451\u043d\u043d\u043e\u0433\u043e \u0432\u044b\u0432\u043e\u0434\u0430. Llm\u2011d \u0442\u0430\u043a\u0436\u0435 \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u0443\u0435\u0442 LMCache \u0434\u043b\u044f\u00a0\u0432\u044b\u0433\u0440\u0443\u0437\u043a\u0438 KV\u2011cache (key\u2011value cache offloading), \u0441\u0435\u0442\u0435\u0432\u043e\u0439 \u043c\u0430\u0440\u0448\u0440\u0443\u0442\u0438\u0437\u0430\u0446\u0438\u0438 \u0441\u00a0\u043f\u043e\u0434\u0434\u0435\u0440\u0436\u043a\u043e\u0439 \u0418\u0418, \u0432\u044b\u0441\u043e\u043a\u043e\u043f\u0440\u043e\u0438\u0437\u0432\u043e\u0434\u0438\u0442\u0435\u043b\u044c\u043d\u044b\u0445 \u043a\u043e\u043c\u043c\u0443\u043d\u0438\u043a\u0430\u0446\u0438\u043e\u043d\u043d\u044b\u0445 API \u0438 \u0434\u0440\u0443\u0433\u0438\u0445 \u0444\u0443\u043d\u043a\u0446\u0438\u0439, \u043f\u043e\u043c\u043e\u0433\u0430\u044e\u0449\u0438\u0445 \u043d\u0430\u0439\u0442\u0438 \u0443\u0431\u0435\u0434\u0438\u0442\u0435\u043b\u044c\u043d\u043e\u0435 \u0440\u0435\u0448\u0435\u043d\u0438\u0435 \u0434\u043b\u044f\u00a0\u0440\u0430\u0441\u043f\u0440\u0435\u0434\u0435\u043b\u0451\u043d\u043d\u043e\u0433\u043e \u0432\u044b\u0432\u043e\u0434\u0430 \u0438 \u043c\u0430\u0441\u0448\u0442\u0430\u0431\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f \u0418\u0418\u2011\u0441\u0438\u0441\u0442\u0435\u043c. <\/p>\n<p>\u041e\u0441\u043d\u043e\u0432\u043d\u044b\u0435 \u043e\u0441\u043e\u0431\u0435\u043d\u043d\u043e\u0441\u0442\u0438 llm\u2011d:<\/p>\n<ul>\n<li>\n<p>\u043f\u043b\u0430\u043d\u0438\u0440\u043e\u0432\u0449\u0438\u043a \u0432\u044b\u0432\u043e\u0434\u0430\\\u0438\u043d\u0442\u0435\u0440\u0444\u0435\u0440\u0435\u043d\u0441\u0430 (Inference Scheduler), \u043e\u043f\u0442\u0438\u043c\u0438\u0437\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u044b\u0439 \u0434\u043b\u044f\u00a0vLLM;<\/p>\n<\/li>\n<li>\n<p>\u0434\u0435\u0437\u0430\u0433\u0440\u0435\u0433\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u043e\u0435 \u043e\u0431\u0441\u043b\u0443\u0436\u0438\u0432\u0430\u043d\u0438\u0435 \u0441\u00a0vLLM;<\/p>\n<\/li>\n<li>\n<p>\u043a\u044d\u0448\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u0435 \u0434\u0435\u0437\u0430\u0433\u0440\u0435\u0433\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u044b\u0445 \u043f\u0440\u0435\u0444\u0438\u043a\u0441\u043e\u0432 \u0441\u00a0vLLM;<\/p>\n<\/li>\n<li>\n<p>\u0432\u043e\u0437\u043c\u043e\u0436\u043d\u043e\u0441\u0442\u044c \u0430\u0432\u0442\u043e\u043c\u0430\u0442\u0438\u0447\u0435\u0441\u043a\u043e\u0433\u043e \u043c\u0430\u0441\u0448\u0442\u0430\u0431\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f \u043f\u043e\u00a0\u043e\u0431\u043e\u0440\u0443\u0434\u043e\u0432\u0430\u043d\u0438\u044e, \u0440\u0430\u0431\u043e\u0447\u0435\u0439 \u043d\u0430\u0433\u0440\u0443\u0437\u043a\u0435 \u0438 \u0442\u0440\u0430\u0444\u0438\u043a\u0443.<\/p>\n<\/li>\n<\/ul>\n<figure class=\"full-width\"><img decoding=\"async\" src=\"https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/upload_files\/5e2\/86f\/80a\/5e286f80aaf7e7709634fade54e8c0e0.png\" width=\"772\" height=\"577\" sizes=\"auto, (max-width: 780px) 100vw, 50vw\" srcset=\"https:\/\/habrastorage.org\/r\/w780\/getpro\/habr\/upload_files\/5e2\/86f\/80a\/5e286f80aaf7e7709634fade54e8c0e0.png 780w,&#10;       https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/upload_files\/5e2\/86f\/80a\/5e286f80aaf7e7709634fade54e8c0e0.png 781w\" loading=\"lazy\" decode=\"async\"\/><\/figure>\n<\/div>\n<\/div>\n<\/div>\n<p><!----><!----><\/div>\n<p><!----><!----><br \/> \u0441\u0441\u044b\u043b\u043a\u0430 \u043d\u0430 \u043e\u0440\u0438\u0433\u0438\u043d\u0430\u043b \u0441\u0442\u0430\u0442\u044c\u0438 <a href=\"https:\/\/habr.com\/ru\/articles\/911690\/\"> https:\/\/habr.com\/ru\/articles\/911690\/<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<div><!--[--><!--]--><\/div>\n<div id=\"post-content-body\">\n<div>\n<div class=\"article-formatted-body article-formatted-body article-formatted-body_version-2\">\n<div xmlns=\"http:\/\/www.w3.org\/1999\/xhtml\">\n<figure class=\"full-width\"><\/figure>\n<p>Red Hat <a href=\"https:\/\/www.redhat.com\/en\/about\/press-releases\/red-hat-launches-llm-d-community-powering-distributed-gen-ai-inference-scale\" rel=\"noopener noreferrer nofollow\">\u043f\u0440\u0435\u0434\u0441\u0442\u0430\u0432\u0438\u043b\u0430<\/a> \u043e\u0442\u043a\u0440\u044b\u0442\u044b\u0439 <a href=\"https:\/\/llm-d.ai\/blog\/llm-d-announce\" rel=\"noopener noreferrer nofollow\">\u043f\u0440\u043e\u0435\u043a\u0442 llm\u2011d<\/a>. \u042d\u0442\u043e \u0432\u044b\u0441\u043e\u043a\u043e\u043f\u0440\u043e\u0438\u0437\u0432\u043e\u0434\u0438\u0442\u0435\u043b\u044c\u043d\u0430\u044f \u0440\u0430\u0441\u043f\u0440\u0435\u0434\u0435\u043b\u0451\u043d\u043d\u0430\u044f \u0441\u0442\u0440\u0443\u043a\u0442\u0443\u0440\u0430 \u0432\u044b\u0432\u043e\u0434\u0430 LLM \u043d\u0430\u00a0\u043e\u0441\u043d\u043e\u0432\u0435 Kubernetes. \u0418\u0441\u0445\u043e\u0434\u043d\u044b\u0439 \u043a\u043e\u0434 \u0440\u0435\u0448\u0435\u043d\u0438\u044f <a href=\"https:\/\/github.com\/llm-d\/llm-d\" rel=\"noopener noreferrer nofollow\">\u043e\u043f\u0443\u0431\u043b\u0438\u043a\u043e\u0432\u0430\u043d<\/a> \u043d\u0430\u00a0GitHub \u043f\u043e\u0434\u00a0\u043b\u0438\u0446\u0435\u043d\u0437\u0438\u0435\u0439 Apache License 2.0.<\/p>\n<p>\u041f\u0440\u043e\u0435\u043a\u0442 llm\u2011d \u043d\u0430\u043f\u0440\u0430\u0432\u043b\u0435\u043d \u043d\u0430\u00a0\u0440\u0430\u0441\u0448\u0438\u0440\u0435\u043d\u0438\u0435 \u0432\u043e\u0437\u043c\u043e\u0436\u043d\u043e\u0441\u0442\u0435\u0439 \u0440\u0430\u0441\u043f\u0440\u0435\u0434\u0435\u043b\u0451\u043d\u043d\u043e\u0433\u043e \u0432\u044b\u0432\u043e\u0434\u0430 \u0438 \u043c\u0430\u0441\u0448\u0442\u0430\u0431\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f \u0441\u0438\u0441\u0442\u0435\u043c \u0433\u0435\u043d\u0435\u0440\u0430\u0442\u0438\u0432\u043d\u043e\u0433\u043e \u0418\u0418. <\/p>\n<p>\u0420\u0435\u0448\u0435\u043d\u0438\u0435 llm\u2011d \u0440\u0430\u0437\u0440\u0430\u0431\u0430\u0442\u044b\u0432\u0430\u0435\u0442\u0441\u044f Red Hat \u0441\u043e\u0432\u043c\u0435\u0441\u0442\u043d\u043e \u0441\u00a0Nvidia, AMD, Intel, IBM Research, Google Cloud, CoreWeave, Hugging Face \u0438 \u0434\u0440\u0443\u0433\u0438\u043c\u0438 \u043f\u043e\u0441\u0442\u0430\u0432\u0449\u0438\u043a\u0430\u043c\u0438 \u0438 \u043e\u0440\u0433\u0430\u043d\u0438\u0437\u0430\u0446\u0438\u044f\u043c\u0438 \u0432\u00a0\u0441\u0444\u0435\u0440\u0435 \u0418\u0418. <\/p>\n<p>\u041f\u0440\u043e\u0433\u0440\u0430\u043c\u043c\u043d\u043e\u0435 \u043e\u0431\u0435\u0441\u043f\u0435\u0447\u0435\u043d\u0438\u0435 llm\u2011d \u0441\u043e\u0437\u0434\u0430\u043d\u043e \u043d\u0430\u00a0\u043e\u0441\u043d\u043e\u0432\u0435 Kubernetes \u0438 \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u0443\u0435\u0442 vLLM \u0434\u043b\u044f\u00a0\u0440\u0430\u0441\u043f\u0440\u0435\u0434\u0435\u043b\u0451\u043d\u043d\u043e\u0433\u043e \u0432\u044b\u0432\u043e\u0434\u0430. Llm\u2011d \u0442\u0430\u043a\u0436\u0435 \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u0443\u0435\u0442 LMCache \u0434\u043b\u044f\u00a0\u0432\u044b\u0433\u0440\u0443\u0437\u043a\u0438 KV\u2011cache (key\u2011value cache offloading), \u0441\u0435\u0442\u0435\u0432\u043e\u0439 \u043c\u0430\u0440\u0448\u0440\u0443\u0442\u0438\u0437\u0430\u0446\u0438\u0438 \u0441\u00a0\u043f\u043e\u0434\u0434\u0435\u0440\u0436\u043a\u043e\u0439 \u0418\u0418, \u0432\u044b\u0441\u043e\u043a\u043e\u043f\u0440\u043e\u0438\u0437\u0432\u043e\u0434\u0438\u0442\u0435\u043b\u044c\u043d\u044b\u0445 \u043a\u043e\u043c\u043c\u0443\u043d\u0438\u043a\u0430\u0446\u0438\u043e\u043d\u043d\u044b\u0445 API \u0438 \u0434\u0440\u0443\u0433\u0438\u0445 \u0444\u0443\u043d\u043a\u0446\u0438\u0439, \u043f\u043e\u043c\u043e\u0433\u0430\u044e\u0449\u0438\u0445 \u043d\u0430\u0439\u0442\u0438 \u0443\u0431\u0435\u0434\u0438\u0442\u0435\u043b\u044c\u043d\u043e\u0435 \u0440\u0435\u0448\u0435\u043d\u0438\u0435 \u0434\u043b\u044f\u00a0\u0440\u0430\u0441\u043f\u0440\u0435\u0434\u0435\u043b\u0451\u043d\u043d\u043e\u0433\u043e \u0432\u044b\u0432\u043e\u0434\u0430 \u0438 \u043c\u0430\u0441\u0448\u0442\u0430\u0431\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f \u0418\u0418\u2011\u0441\u0438\u0441\u0442\u0435\u043c. <\/p>\n<p>\u041e\u0441\u043d\u043e\u0432\u043d\u044b\u0435 \u043e\u0441\u043e\u0431\u0435\u043d\u043d\u043e\u0441\u0442\u0438 llm\u2011d:<\/p>\n<ul>\n<li>\n<p>\u043f\u043b\u0430\u043d\u0438\u0440\u043e\u0432\u0449\u0438\u043a \u0432\u044b\u0432\u043e\u0434\u0430\\\u0438\u043d\u0442\u0435\u0440\u0444\u0435\u0440\u0435\u043d\u0441\u0430 (Inference Scheduler), \u043e\u043f\u0442\u0438\u043c\u0438\u0437\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u044b\u0439 \u0434\u043b\u044f\u00a0vLLM;<\/p>\n<\/li>\n<li>\n<p>\u0434\u0435\u0437\u0430\u0433\u0440\u0435\u0433\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u043e\u0435 \u043e\u0431\u0441\u043b\u0443\u0436\u0438\u0432\u0430\u043d\u0438\u0435 \u0441\u00a0vLLM;<\/p>\n<\/li>\n<li>\n<p>\u043a\u044d\u0448\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u0435 \u0434\u0435\u0437\u0430\u0433\u0440\u0435\u0433\u0438\u0440\u043e\u0432\u0430\u043d\u043d\u044b\u0445 \u043f\u0440\u0435\u0444\u0438\u043a\u0441\u043e\u0432 \u0441\u00a0vLLM;<\/p>\n<\/li>\n<li>\n<p>\u0432\u043e\u0437\u043c\u043e\u0436\u043d\u043e\u0441\u0442\u044c \u0430\u0432\u0442\u043e\u043c\u0430\u0442\u0438\u0447\u0435\u0441\u043a\u043e\u0433\u043e \u043c\u0430\u0441\u0448\u0442\u0430\u0431\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f \u043f\u043e\u00a0\u043e\u0431\u043e\u0440\u0443\u0434\u043e\u0432\u0430\u043d\u0438\u044e, \u0440\u0430\u0431\u043e\u0447\u0435\u0439 \u043d\u0430\u0433\u0440\u0443\u0437\u043a\u0435 \u0438 \u0442\u0440\u0430\u0444\u0438\u043a\u0443.<\/p>\n<\/li>\n<\/ul>\n<figure class=\"full-width\"><\/figure>\n<\/div>\n<\/div>\n<\/div>\n<p><!----><!----><\/div>\n<p><!----><!----><br \/> \u0441\u0441\u044b\u043b\u043a\u0430 \u043d\u0430 \u043e\u0440\u0438\u0433\u0438\u043d\u0430\u043b \u0441\u0442\u0430\u0442\u044c\u0438 <a href=\"https:\/\/habr.com\/ru\/articles\/911690\/\"> https:\/\/habr.com\/ru\/articles\/911690\/<\/a><br \/><\/br><\/br><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[],"tags":[],"class_list":["post-460473","post","type-post","status-publish","format-standard","hentry"],"_links":{"self":[{"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=\/wp\/v2\/posts\/460473","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=460473"}],"version-history":[{"count":0,"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=\/wp\/v2\/posts\/460473\/revisions"}],"wp:attachment":[{"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=460473"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=460473"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=460473"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}