[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"article-meta-ai-hardware-wall-street-compute-demand-en":3,"article-related-meta-ai-hardware-wall-street-compute-demand-en":33,"series-industry-fa94d7c8-01ec-4cc1-9b33-a1c7ba78cda2":80},{"id":4,"slug":5,"title":6,"content":7,"summary":8,"source":9,"source_url":10,"author":11,"image_url":12,"cover_image":12,"category":13,"language":14,"translated_content":11,"related_article_id":15,"keywords":16,"key_takeaways":25,"views":29,"created_at":30,"published_at":31,"topic_cluster_id":32},"fa94d7c8-01ec-4cc1-9b33-a1c7ba78cda2","meta-ai-hardware-wall-street-compute-demand-en","Meta一句话引爆AI硬件，华尔街仍看多算力","\u003Cp data-speakable=\"summary\">华尔街把\u003Ca href=\"\u002Ftag\u002Fmeta\">Meta\u003C\u002Fa>的表态视作情绪冲击，但推理和\u003Ca href=\"\u002Ftag\u002Fagent\">agent\u003C\u002Fa>需求仍在扩张。\u003C\u002Fp>\u003Cp>Meta一句“卖算力”让AI硬件板块短线承压，但华尔街的拆解重点并不在“算力过剩”，而在需求结构变化。一个更直接的信号来自 \u003Ca href=\"https:\u002F\u002Fopenai.com\">OpenAI\u003C\u002Fa> 的 \u003Ca href=\"\u002Ftag\u002Fcodex\">Codex\u003C\u002Fa> 和 \u003Ca href=\"\u002Ftag\u002Fagentic-ai\">agentic AI\u003C\u002Fa> 相关数据：个人非开发者用户数增长137倍，组织用户数增长189倍，OpenAI内部用户数增长12倍。读完下面5项，你会知道为什么这更像是估值波动，而不是行业拐点。\u003C\u002Fp>\u003Ctable>\u003Cthead>\u003Ctr>\u003Cth>Item\u003C\u002Fth>\u003Cth>Signal\u003C\u002Fth>\u003Cth>Implication\u003C\u002Fth>\u003C\u002Ftr>\u003C\u002Fthead>\u003Ctbody>\u003Ctr>\u003Ctd>OpenAI 个人非开发者用户\u003C\u002Ftd>\u003Ctd>增长137倍\u003C\u002Ftd>\u003Ctd>推理需求扩张\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>OpenAI 组织用户\u003C\u002Ftd>\u003Ctd>增长189倍\u003C\u002Ftd>\u003Ctd>企业侧采用加速\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>OpenAI 内部用户\u003C\u002Ftd>\u003Ctd>增长12倍\u003C\u002Ftd>\u003Ctd>agent\u002F工具化使用升温\u003C\u002Ftd>\u003C\u002Ftr>\u003C\u002Ftbody>\u003C\u002Ftable>\u003Ch2>1. OpenAI 的增长数据说明了什么\u003C\u002Fh2>\u003Cp>这组数据最重要的地方，不是“涨了多少”，而是它指向了新的需求来源。个人非开发者、组织、内部用户都在快速增加，说明AI使用正在从开发者圈层外溢到更广的人群和更多工作流里。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1783081971766-41gl.png\" alt=\"Meta一句话引爆AI硬件，华尔街仍看多算力\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>这意味着算力需求不只来自训练大模型，还来自持续调用模型的推理场景。只要用户规模和调用频率继续上升，硬件需求就很难因为一两句市场表态而逆转。\u003C\u002Fp>\u003Cul>\u003Cli>个人非开发者用户：增长137倍\u003C\u002Fli>\u003Cli>组织用户：增长189倍\u003C\u002Fli>\u003Cli>OpenAI内部用户：增长12倍\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>2. 为什么“卖算力”不等于算力过剩\u003C\u002Fh2>\u003Cp>市场把Meta的说法解读成AI硬件需求降温，但这类表态更像公司层面的资本开支判断，不等于整个行业供需失衡。企业会调整采购节奏、预算结构和合作方式，但不会因此让推理需求自动消失。\u003C\u002Fp>\u003Cp>真正决定硬件周期的，是用户使用量、推理频次和新应用的渗透速度。现在被华尔街盯住的，是agent和推理的增长曲线，而不是单一公司的采购态度。\u003C\u002Fp>\u003Cul>\u003Cli>公司表态：影响短期情绪\u003C\u002Fli>\u003Cli>用户增长：影响长期需求\u003C\u002Fli>\u003Cli>推理调用：决定算力消耗\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>3. 推理和agent应用才是更大的需求侧\u003C\u002Fh2>\u003Cp>训练模型通常是一次性或阶段性的需求，推理则更像持续性消耗。随着更多人把AI用在写代码、检索、办公和自动化任务上，推理调用会变成更稳定的算力来源。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1783081970315-f959.png\" alt=\"Meta一句话引爆AI硬件，华尔街仍看多算力\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>agent应用尤其值得关注，因为它不是“问一次答一次”，而是可能在多个步骤里反复调用模型、工具和外部系统。换句话说，agent把一次任务拆成很多次计算，算力消耗自然更高。\u003C\u002Fp>\u003Cul>\u003Cli>写代码助手\u003C\u002Fli>\u003Cli>企业知识检索\u003C\u002Fli>\u003Cli>自动化工作流\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>4. 华尔街为什么仍然没有转向看空\u003C\u002Fh2>\u003Cp>机构的判断逻辑很简单：如果需求仍在扩张，硬件板块的下跌更可能是估值回撤，而不是基本面崩坏。尤其当新的AI使用场景不断出现时，市场很难用单一新闻定义整个周期。\u003C\u002Fp>\u003Cp>因此，当前更合理的解读是“需求结构在变”，而不是“需求结束了”。训练侧可能阶段性放缓，但推理侧和agent侧还在接力，这也是为什么分析师没有急着把这次波动定义为行业拐点。\u003C\u002Fp>\u003Ccode>需求 = 训练 + 推理 + agent调用\u003C\u002Fcode>\u003Ch2>5. 哪些信号比Meta表态更值得盯\u003C\u002Fh2>\u003Cp>如果你想判断AI硬件是不是真的见顶，重点不该是公司一句话，而是几个更硬的指标：用户数、调用频次、企业采用率和新产品的留存情况。它们比情绪新闻更能反映真实需求。\u003C\u002Fp>\u003Cp>当这些指标继续上行时，算力市场通常更像是在换挡，而不是熄火。也就是说，短期震荡可以有，长期供需反转还需要更强的数据证明。\u003C\u002Fp>\u003Cul>\u003Cli>用户增长速度\u003C\u002Fli>\u003Cli>企业部署数量\u003C\u002Fli>\u003Cli>agent任务完成率\u003C\u002Fli>\u003Cli>推理调用量\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>How to decide\u003C\u002Fh2>\u003Cp>如果你关注AI硬件股，短线交易者要盯情绪和估值，长线投资者要盯推理与agent的真实采用数据。前者决定波动，后者决定方向。\u003C\u002Fp>\u003Cp>如果你想判断这轮算力行情是否结束，优先看用户增长、组织采用和持续调用，而不是只看某家公司的采购口径。当前更像是市场在重新定价需求结构，而不是给算力周期下结论。\u003C\u002Fp>","1个信号解释华尔街为何不看空算力：推理和agent需求还在扩张。","zhuanlan.zhihu.com","https:\u002F\u002Fzhuanlan.zhihu.com\u002Fp\u002F2055972204626572740",null,"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1783081971766-41gl.png","industry","en","4217d41e-9694-4a49-9520-2f5d4b009496",[17,18,19,20,21,22,23,24],"Meta","AI硬件","算力","推理","agent","OpenAI","华尔街","需求增长",[26,27,28],"OpenAI数据显示，个人非开发者、组织和内部用户都在快速增长。","Meta的表态更像情绪冲击，不足以证明算力过剩。","推理和agent应用仍是更大的需求来源，决定长期算力消耗。",0,"2026-07-03T12:32:22.193465+00:00","2026-07-03T12:32:22.176+00:00","a1c158f8-b98b-4d99-aa84-35523d1f1876",{"tags":34,"relatedLang":39,"relatedPosts":43},[35,36,37],{"name":21,"slug":21},{"name":19,"slug":19},{"name":17,"slug":38},"meta",{"id":15,"slug":40,"title":41,"language":42},"meta-ai-hardware-wall-street-compute-demand-zh","Meta一句話震盪算力股，華爾街仍看多推理需求","zh",[44,50,56,62,68,74],{"id":45,"slug":46,"title":47,"cover_image":48,"image_url":48,"created_at":49,"category":13},"36fcb17a-f289-4dc5-94dc-ae2174b685ce","california-claude-deal-state-offices-en","California’s Claude deal puts AI inside state offices","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1783083770605-anqx.png","2026-07-03T13:02:20.665293+00:00",{"id":51,"slug":52,"title":53,"cover_image":54,"image_url":54,"created_at":55,"category":13},"7acd4647-667f-4218-a554-3734f6737efd","cloudflare-september-2026-crawl-defaults-en","Cloudflare sets September 15, 2026 crawl defaults","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1783071172244-ccw5.png","2026-07-03T09:32:29.80015+00:00",{"id":57,"slug":58,"title":59,"cover_image":60,"image_url":60,"created_at":61,"category":13},"423c0239-6cff-4206-9bbb-3db02da0c622","cloudflares-new-crawler-rules-shift-power-en","Cloudflare’s new crawler rules shift power to publishers","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1783069375501-xn87.png","2026-07-03T09:02:27.462999+00:00",{"id":63,"slug":64,"title":65,"cover_image":66,"image_url":66,"created_at":67,"category":13},"0362089c-2aa4-4fff-9b62-1a8b18c4d8ca","midjourney-body-scan-spa-ultrasonic-ct-en","Midjourney is building a body-scan spa","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1783067585797-r9j9.png","2026-07-03T08:32:42.100537+00:00",{"id":69,"slug":70,"title":71,"cover_image":72,"image_url":72,"created_at":73,"category":13},"b43e5f46-a4d7-4441-9aee-890cc83b8141","midjourney-force-hollywood-ai-cards-table-en","Midjourney Is Right to Force Hollywood’s AI Cards Onto the Table","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1783065773451-i5ie.png","2026-07-03T08:02:25.173434+00:00",{"id":75,"slug":76,"title":77,"cover_image":78,"image_url":78,"created_at":79,"category":13},"6b2d04b8-9053-44d8-8ec1-1e614b55ab8b","midjourney-body-scanner-diagnosis-wrong-bet-en","Midjourney should stay out of diagnosis and focus on body-composition…","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1783063975300-wt2i.png","2026-07-03T07:32:27.132957+00:00",[81,86,91,96,101,106,111,116,121,126],{"id":82,"slug":83,"title":84,"created_at":85},"d35a1bd9-e709-412e-a2df-392df1dc572a","ai-impact-2026-developments-market-en","AI's Impact in 2026: Key Developments and Market Shifts","2026-03-25T16:20:33.205823+00:00",{"id":87,"slug":88,"title":89,"created_at":90},"5ed27921-5fd6-492e-8c59-78393bf37710","trumps-ai-legislative-framework-en","Trump's AI Legislative Framework: What's Inside?","2026-03-25T16:22:20.005325+00:00",{"id":92,"slug":93,"title":94,"created_at":95},"e454a642-f03c-4794-b185-5f651aebbaca","nvidia-gtc-2026-key-highlights-innovations-en","NVIDIA GTC 2026: Key Highlights and Innovations","2026-03-25T16:22:47.882615+00:00",{"id":97,"slug":98,"title":99,"created_at":100},"0ebb5b16-774a-4922-945d-5f2ce1df5a6d","claude-usage-diversifies-learning-curves-en","Claude Usage Diversifies, Learning Curves Emerge","2026-03-25T16:25:50.770376+00:00",{"id":102,"slug":103,"title":104,"created_at":105},"69934e86-2fc5-4280-8223-7b917a48ace8","openclaw-ai-commoditization-concerns-en","OpenClaw's Rise Raises Concerns of AI Model Commoditization","2026-03-25T16:26:30.582047+00:00",{"id":107,"slug":108,"title":109,"created_at":110},"b4b2575b-2ac8-46b2-b90e-ab1d7c060797","google-gemini-ai-rollout-2026-en","Google's Gemini AI Rollout Extended to 2026","2026-03-25T16:28:14.808842+00:00",{"id":112,"slug":113,"title":114,"created_at":115},"6e18bc65-42ae-4ad0-b564-67d7f66b979e","meta-llama4-fabricated-results-scandal-en","Meta's Llama 4 Scandal: Fabricated AI Test Results Unveiled","2026-03-25T16:29:15.482836+00:00",{"id":117,"slug":118,"title":119,"created_at":120},"bf888e9d-08be-4f47-996c-7b24b5ab3500","accenture-mistral-ai-deployment-en","Accenture and Mistral AI Team Up for AI Deployment","2026-03-25T16:31:01.894655+00:00",{"id":122,"slug":123,"title":124,"created_at":125},"5382b536-fad2-49c6-ac85-9eb2bae49f35","mistral-ai-high-stakes-2026-en","Mistral AI: Facing High Stakes in 2026","2026-03-25T16:31:39.941974+00:00",{"id":127,"slug":128,"title":129,"created_at":130},"9da3d2d6-b669-4971-ba1d-17fdb3548ed5","cursors-meteoric-rise-pressures-en","Cursor's Meteoric Rise Faces Industry Pressures","2026-03-25T16:32:21.899217+00:00"]