{"id":109516,"date":"2025-05-21T14:52:30","date_gmt":"2025-05-21T06:52:30","guid":{"rendered":"https:\/\/www.ioiotimes.com\/?p=109516"},"modified":"2025-05-21T20:59:33","modified_gmt":"2025-05-21T12:59:33","slug":"red-hat%e6%8e%a8%e5%87%bared-hat-ai-inference-server%ef%bc%8c%e7%82%ba%e8%b7%a8%e6%b7%b7%e5%90%88%e9%9b%b2%e7%9a%84%e6%a8%a1%e5%9e%8b%e8%88%87%e5%8a%a0%e9%80%9f%e5%99%a8%e9%87%8b%e6%94%be","status":"publish","type":"post","link":"https:\/\/www.ioiotimes.com\/?p=109516","title":{"rendered":"Red Hat\u63a8\u51faRed Hat AI\u00a0Inference Server\uff0c\u70ba\u8de8\u6df7\u5408\u96f2\u7684\u6a21\u578b\u8207\u52a0\u901f\u5668\u91cb\u653e\u751f\u6210\u5f0fAI\u6f5b\u529b"},"content":{"rendered":"\n

Red Hat AI<\/strong>\u00a0Inference Server\u63a1\u7528\u6574\u5408Neural Magic\u6280\u8853\u7684vLLM\u4e26\u52a0\u4ee5\u5f37\u5316\uff0c\u70ba\u8de8\u6df7\u5408\u96f2\u74b0\u5883\u63d0\u4f9b\u66f4\u5feb\u3001\u66f4\u9ad8\u6548\u80fd\u4e14\u66f4\u5177\u6210\u672c\u6548\u76ca\u7684AI\u63a8\u8ad6<\/strong><\/h3>\n\n\n\n

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

\"\"<\/figure>\n\n\n\n

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

\u4e16\u754c\u9818\u5148\u958b\u653e\u539f\u59cb\u78bc\u8edf\u9ad4\u89e3\u6c7a\u65b9\u6848\u4f9b\u61c9\u5546 Red Hat \u4eca\u65e5\u5ba3\u5e03\u63a8\u51fa Red Hat AI Inference Server\uff0c\u9081\u51fa\u751f\u6210\u5f0f AI\uff08gen AI\uff09\u666e\u53ca\u81f3\u6df7\u5408\u96f2\u7684\u91cd\u8981\u4e00\u6b65\u3002\u4f5c\u70ba Red Hat AI \u7684\u5168\u65b0\u4f01\u696d\u7d1a\u63a8\u8ad6\u4f3a\u670d\u5668\uff0c\u6b64\u89e3\u6c7a\u65b9\u6848\u4e0d\u50c5\u6e90\u81ea\u65bc\u5f37\u5927\u7684 vLLM \u793e\u7fa4\u5c08\u6848\uff0c\u66f4\u900f\u904e Red Hat \u6574\u5408 Neural Magic \u6280\u8853\u52a0\u4ee5\u5f37\u5316\uff0c\u63d0\u4f9b\u66f4\u5feb\u7684\u901f\u5ea6\u3001\u66f4\u9ad8\u7684\u52a0\u901f\u5668\u6548\u7387\u8207\u66f4\u4f73\u7684\u6210\u672c\u6548\u76ca\uff0c\u4fc3\u9032\u5be6\u73fe Red Hat \u7684\u9858\u666f\uff0c\u4ea6\u5373\u80fd\u65bc\u4efb\u4f55\u96f2\u7aef\u74b0\u5883\u3001\u4efb\u4f55 AI \u52a0\u901f\u5668\u4e0a\u57f7\u884c\u5404\u7a2e\u751f\u6210\u5f0f AI \u6a21\u578b\u3002\u7121\u8ad6\u662f\u7368\u7acb\u90e8\u7f72\uff0c\u6216\u662f\u4f5c\u70ba Red Hat Enterprise Linux AI\uff08RHEL AI\uff09\u53ca Red Hat OpenShift AI \u7684\u6574\u5408\u5143\u4ef6\uff0c\u6b64\u7a81\u7834\u6027\u5e73\u53f0\u8ce6\u80fd\u4f01\u696d\u80fd\u66f4\u81ea\u4fe1\u5730\u5728\u751f\u7522\u74b0\u5883\u4e2d\u90e8\u7f72\u8207\u64f4\u5c55\u751f\u6210\u5f0f AI\u3002<\/strong><\/p>\n\n\n\n

\u63a8\u8ad6\uff08Inference\uff09\u662f AI \u7684\u95dc\u9375\u57f7\u884c\u5f15\u64ce\uff0c\u9810\u5148\u8a13\u7df4\u6a21\u578b\u5f97\u4ee5\u501f\u52a9\u6b64\u6b65\u9a5f\u5c07\u8cc7\u6599\u8f49\u5316\u70ba\u5be6\u969b\u5f71\u97ff\u7684\u7d50\u679c\u3002\u63a8\u8ad6\u4f5c\u70ba\u4f7f\u7528\u8005\u4e92\u52d5\u7684\u6a1e\u7d10\uff0c\u9700\u8981\u8fc5\u901f\u4e14\u6e96\u78ba\u7684\u56de\u61c9\u3002\u96a8\u8457\u751f\u6210\u5f0f AI \u6a21\u578b\u65e5\u76ca\u8907\u96dc\uff0c\u52a0\u4e0a\u751f\u7522\u74b0\u5883\u90e8\u7f72\u898f\u6a21\u64f4\u589e\uff0c\u63a8\u8ad6\u53ef\u80fd\u6210\u70ba\u4e00\u5927\u74f6\u9838\uff0c\u4e0d\u50c5\u6703\u6d88\u8017\u5927\u91cf\u786c\u9ad4\u8cc7\u6e90\uff0c\u66f4\u53ef\u80fd\u5c0e\u81f4\u56de\u61c9\u901f\u5ea6\u9072\u7de9\u4e26\u63d0\u5347\u71df\u904b\u6210\u672c\u3002\u70ba\u4e86\u5927\u898f\u6a21\u91cb\u653e AI \u771f\u6b63\u7684\u6f5b\u529b\uff0c\u4e26\u4e14\u66f4\u5f9e\u5bb9\u5730\u61c9\u5c0d\u5176\u6f5b\u5728\u7684\u8907\u96dc\u6027\uff0c\u5f37\u5927\u7684\u63a8\u8ad6\u4f3a\u670d\u5668\u5df2\u4e0d\u518d\u662f\u5962\u4f88\u54c1\uff0c\u800c\u662f\u5fc5\u8981\u689d\u4ef6\u3002<\/p>\n\n\n\n

\u70ba\u61c9\u5c0d\u4e0a\u8ff0\u6311\u6230\uff0cRed Hat \u5168\u65b0\u63a8\u51fa Red Hat AI Inference Server\uff0c\u8a72\u958b\u653e\u5f0f\u63a8\u8ad6\u89e3\u6c7a\u65b9\u6848\u662f\u5c08\u70ba\u9ad8\u6548\u80fd\u8a2d\u8a08\uff0c\u4e26\u642d\u914d\u9802\u5c16\u7684\u6a21\u578b\u58d3\u7e2e\uff08model compression\uff09\u8207\u6700\u4f73\u5316\u5de5\u5177\u3002\u6b64\u5275\u65b0\u80fd\u63d0\u4f9b\u53cd\u61c9\u66f4\u52a0\u9748\u654f\u7684\u4f7f\u7528\u8005\u9ad4\u9a57\uff0c\u540c\u6642\u4f01\u696d\u5728\u9078\u64c7 AI \u52a0\u901f\u5668\u3001\u6a21\u578b\u53ca IT \u74b0\u5883\u6642\u5f97\u4ee5\u4eab\u6709\u524d\u6240\u672a\u6709\u7684\u81ea\u7531\u5ea6\uff0c\u9032\u800c\u5145\u5206\u91cb\u653e\u751f\u6210\u5f0f AI \u7684\u8f49\u578b\u52d5\u80fd\u3002<\/p>\n\n\n\n

Red Hat \u526f\u7e3d\u88c1\u66a8AI \u4e8b\u696d\u90e8\u7e3d\u7d93\u7406 Joe Fernandes \u8868\u793a\uff1a\u300c\u63a8\u8ad6\u662f\u751f\u6210\u5f0f AI \u771f\u6b63\u5c55\u73fe\u50f9\u503c\u7684\u5730\u65b9\uff0c\u5728\u9019\u500b\u968e\u6bb5\uff0c\u7279\u5b9a\u7684\u6a21\u578b\u80fd\u70ba\u4f7f\u7528\u8005\u4e92\u52d5\u63d0\u4f9b\u5feb\u901f\u3001\u6e96\u78ba\u7684\u56de\u61c9\uff0c\u4f46\u9019\u500b\u904e\u7a0b\u5fc5\u9808\u4ee5\u6709\u6548\u4e14\u5177\u6210\u672c\u6548\u76ca\u7684\u65b9\u5f0f\u5be6\u73fe\u3002Red Hat AI Inference Server \u65e8\u5728\u6eff\u8db3\u5927\u898f\u6a21\u3001\u9ad8\u6548\u80fd\u3001\u9ad8\u56de\u61c9\u6027\u63a8\u8ad6\u7684\u9700\u6c42\uff0c\u540c\u6642\u7dad\u6301\u4f4e\u8cc7\u6e90\u8017\u7528\uff0c\u9032\u800c\u63d0\u4f9b\u901a\u7528\u63a8\u8ad6\u5c64\uff0c\u652f\u63f4\u5728\u4efb\u4f55\u74b0\u5883\u3001\u4efb\u4f55\u52a0\u901f\u5668\u4e0a\u57f7\u884c\u7684\u4efb\u4f55\u6a21\u578b\u3002\u300d<\/p>\n\n\n\n

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

<\/a>vLLM<\/strong>\uff1a\u64f4\u5145\u63a8\u8ad6\u5275\u65b0<\/strong><\/h3>\n\n\n\n

Red Hat AI Inference Server \u662f\u5efa\u7f6e\u65bc\u5f15\u9818\u696d\u754c\u7684 vLLM \u5c08\u6848\u4e4b\u4e0a\u3002\u6b64\u793e\u7fa4\u5c08\u6848\u662f\u7531\u52a0\u5dde\u5927\u5b78\u67cf\u514b\u840a\u5206\u6821\u65bc 2023 \u5e74\u4e2d\u555f\u52d5\uff0c\u53ef\u63d0\u4f9b\u9ad8\u50b3\u8f38\u91cf\u7684\u751f\u6210\u5f0f AI \u63a8\u8ad6\u3001\u652f\u63f4\u5927\u578b\u5167\u5bb9\u8f38\u5165\u3001\u591a GPU \u6a21\u578b\u52a0\u901f\u4e26\u652f\u63f4\u9023\u7e8c\u6279\u6b21\u8655\u7406\u7b49\u773e\u591a\u529f\u80fd\u3002<\/p>\n\n\n\n

vLLM \u4e0d\u50c5\u5ee3\u6cdb\u652f\u63f4\u516c\u958b\u53ef\u7528\u7684\u6a21\u578b\uff0c\u66f4\u80fd\u5f9e Day 0 \u5373\u6574\u5408 DeepSeek\u3001Gemma\u3001Llama\u3001Mistral\u3001Phi \u7b49\u6a21\u578b\uff0c\u4ee5\u53ca\u958b\u6e90\u4f01\u696d\u7d1a\u63a8\u7406\u6a21\u578b\uff08reasoning models\uff09\u5982 Llama Nemotron<\/a>\uff0c\u63a8\u52d5\u5176\u6210\u70ba\u672a\u4f86 AI \u63a8\u8ad6\u5275\u65b0\u7684\u5be6\u8cea\u6a19\u6e96\u3002\u9802\u5c16\u6a21\u578b\u7684\u4f9b\u61c9\u5546\u6b63\u7a4d\u6975\u64c1\u62b1 vLLM\uff0c\u9032\u4e00\u6b65\u978f\u56fa vLLM \u5728\u5851\u9020\u751f\u6210\u5f0f AI \u672a\u4f86\u6642\u626e\u6f14\u7684\u95dc\u9375\u89d2\u8272\u3002<\/p>\n\n\n\n

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

\u3000\u3000\u3000<\/p>\n\n\n\n

\ud83d\udfe6COMPUTEX \u6d3b\u52d5– \u770b MSI \u5c55\u524d\u6703\u62ff\u8d85\u840c\u597d\u79ae<\/a>
\ud83d\udfe6\u73fe\u5728\u5c31\u52a0\u5165 ioioTIMES \u81c9\u66f8\u7c89\u7d72\u5718 \u66f4\u591a\u4e92\u52d5\u3001\u66f4\u591a\u597d\u5eb7\u650f\u62b5\u52a0!!
\ud83d\udfe6\u6211\u5011\u6709LINE TODAY<\/a>\u983b\u9053\u4e86\uff0c\u5feb\u4f86\u8ffd\u8e2a\u6211\u5011\u5427!!–\u6700\u65b0\u79d1\u6280\u65b0\u805e \u76e1\u5728\u4f60<\/strong>\u624b<\/h4>\n","protected":false},"excerpt":{"rendered":"

Red Hat AI\u00a0Inferen<\/p>\n","protected":false},"author":3,"featured_media":109517,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"footnotes":""},"categories":[13],"tags":[13670,11808,1586,13671],"class_list":["post-109516","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-ai-inference-server","tag-neural-magic","tag-red-hat","tag-vllm"],"_links":{"self":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/109516"}],"collection":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=109516"}],"version-history":[{"count":2,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/109516\/revisions"}],"predecessor-version":[{"id":109542,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/109516\/revisions\/109542"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/media\/109517"}],"wp:attachment":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=109516"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=109516"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=109516"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}