[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"article-cloud-infrastructure-spend-jumps-ai-demand-zh":3,"tags-cloud-infrastructure-spend-jumps-ai-demand-zh":33,"related-lang-cloud-infrastructure-spend-jumps-ai-demand-zh":47,"related-posts-cloud-infrastructure-spend-jumps-ai-demand-zh":51,"series-industry-53332d20-ac19-4066-bbfd-4164923130f7":88},{"id":4,"title":5,"content":6,"summary":7,"source":8,"source_url":9,"author":10,"image_url":11,"keywords":12,"language":21,"translated_content":10,"views":22,"is_premium":23,"created_at":24,"updated_at":24,"cover_image":11,"published_at":25,"rewrite_status":26,"rewrite_error":10,"rewritten_from_id":27,"slug":28,"category":29,"related_article_id":30,"status":31,"google_indexed_at":32,"x_posted_at":10,"tweet_text":10,"title_rewritten_at":10,"title_original":10,"key_takeaways":10,"topic_cluster_id":10,"embedding":10,"is_canonical_seed":23},"53332d20-ac19-4066-bbfd-4164923130f7","AI 需求把雲端支出推高 29%","\u003Cp>全球雲端基礎設施支出在 Q4 2025 衝到 \u003Cstrong>1109 億美元\u003C\u002Fstrong>。年增 \u003Cstrong>29%\u003C\u002Fstrong>。這不是小波動，是整個市場一起加碼。\u003C\u002Fp>\u003Cp>說白了，錢主要花在 AI。訓練、推論、資料處理，全部都在吃伺服器和 GPU。雲端大廠只能跟著擴容量，不然客戶一多就塞車。\u003C\u002Fp>\u003Cp>問題也很直接。企業一邊喊要上 AI，一邊又盯著成本。結果就是雲端支出往上跑，採購團隊卻還在算每個 Token 的成本。\u003C\u002Fp>\u003Ch2>為什麼帳單一直變大\u003C\u002Fh2>\u003Cp>這波支出成長，核心就是容量。\u003Ca href=\"https:\u002F\u002Faws.amazon.com\u002F\" target=\"_blank\" rel=\"noopener\">AWS\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fcloud.google.com\u002F\" target=\"_blank\" rel=\"noopener\">Google Cloud\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fazure.microsoft.com\u002F\" target=\"_blank\" rel=\"noopener\">Microsoft Azure\u003C\u002Fa> 這些 hy\u003Ca href=\"\u002Fnews\u002Fopenclaw-april-2026-update-xai-minimax-zh\">pe\u003C\u002Fa>rscaler，現在買的不只是機器，是整套 AI 基礎設施。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1775200180978-1ufe.png\" alt=\"AI 需求把雲端支出推高 29%\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>AI 訓練很吃 GPU。推論也不便宜，尤其是高頻請求、長上下文、RAG 這些場景。你如果把模型放進產品主流程，雲端帳單通常會先嚇你一跳。\u003C\u002Fp>\u003Cp>另一個原因是企業工作負載沒有停。傳統 SaaS、資料倉儲、串流分析、備份與災難復原，這些都還在。AI 只是把原本就不低的雲端支出，再往上疊一層。\u003C\u002Fp>\u003Cul>\u003Cli>GPU 採購和機房建置同步增加\u003C\u002Fli>\u003Cli>推論流量把即時算力需求拉高\u003C\u002Fli>\u003Cli>企業把更多資料搬上雲\u003C\u002Fli>\u003Cli>大型模型服務吃掉更多網路與儲存資源\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>誰在付錢，誰在賺錢\u003C\u002Fh2>\u003Cp>這筆錢不是平均分。\u003Ca href=\"https:\u002F\u002Fwww.microsoft.com\u002Fen-us\u002F\" target=\"_blank\" rel=\"noopener\">Microsoft\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fwww.amazon.com\u002F\" target=\"_blank\" rel=\"noopener\">Amazon\u003C\u002Fa>、\u003Ca href=\"https:\u002F\u002Fwww.google.com\u002F\" target=\"_blank\" rel=\"noopener\">Google\u003C\u002Fa> 這類大廠，因為掌握雲端、AI 平台和銷售通路，最容易把需求轉成營收。\u003C\u002Fp>\u003Cp>但客戶端的感受通常沒那麼美好。很多團隊會先做 PoC，再發現正式上線後成本比預期高 2 倍，甚至 3 倍。這時候大家才開始認真看 cache、batching、量化和模型路由。\u003C\u002Fp>\u003Cp>\u003Ca href=\"https:\u002F\u002Fwww.nvidia.com\u002F\" target=\"_blank\" rel=\"noopener\">NVIDIA\u003C\u002Fa> 也在這波裡面吃到大量需求。因為雲端擴容的核心硬體，很多還是繞不開它的 GPU 生態。講白了，雲端大廠買卡，最後還是得看\u003Ca href=\"\u002Fnews\u002Ftrivy-docker-images-fresh-supply-chain-attack-zh\">供應鏈\u003C\u002Fa>。\u003C\u002Fp>\u003Cblockquote>“The AI boom is very real, and it’s very expensive.” — Satya Nadella\u003C\u002Fblockquote>\u003Ch2>跟其他雲端時期比，這次有什麼不同\u003C\u002Fh2>\u003Cp>以前雲端成長，主力是搬遷和數位轉型。現在不一樣，AI 是直接吃算力。差別在於，傳統雲端多半是穩定流量，AI 則是高峰明顯，而且單次請求成本高很多。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1775200172132-jn93.png\" alt=\"AI 需求把雲端支出推高 29%\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>另一個差別是投資節奏。過去企業常常先省錢，再慢慢上雲。現在很多公司是先買算力，再想怎麼把產品做出來。這種順序，會讓支出先衝很快。\u003C\u002Fp>\u003Cp>如果把幾家主要\u003Ca href=\"\u002Fnews\u002Fsora-shutdown-ai-vendor-risk-zh\">供應商\u003C\u002Fa>放一起看，差距也很明顯。\u003Ca href=\"https:\u002F\u002Fwww.oracle.com\u002Fcloud\u002F\" target=\"_blank\" rel=\"noopener\">Oracle Cloud\u003C\u002Fa> 主打資料庫與企業客戶。\u003Ca href=\"https:\u002F\u002Fwww.digitalocean.com\u002F\" target=\"_blank\" rel=\"noopener\">DigitalOcean\u003C\u002Fa> 則偏中小團隊。\u003Ca href=\"https:\u002F\u002Fwww.ibm.com\u002Fcloud\" target=\"_blank\" rel=\"noopener\">IBM Cloud\u003C\u002Fa> 走的是既有企業關係。AI 時代的錢，還是更偏向能拿到大規模 GPU 與網路資源的人。\u003C\u002Fp>\u003Cul>\u003Cli>傳統雲端：搬遷、儲存、SaaS 為主\u003C\u002Fli>\u003Cli>AI 雲端：GPU、推論、向量資料庫為主\u003C\u002Fli>\u003Cli>傳統工作負載：成本相對可預測\u003C\u002Fli>\u003Cli>AI 工作負載：成本波動更大\u003C\u002Fli>\u003C\u002Ful>\u003Ch2>台灣團隊該怎麼看\u003C\u002Fh2>\u003Cp>對台灣開發者來說，這代表兩件事。第一，AI 產品不只要看模型效果，也要看雲端成本。第二，架構能力會變重要，因為省 10% 成本，可能直接影響毛利。\u003C\u002Fp>\u003Cp>很多團隊現在都在改做法。有人把大模型請求拆成多段。有人先用小模型過濾，再丟大模型。也有人把熱門內容快取起來，避免每次都重算。這些都不是花招，是生存技能。\u003C\u002Fp>\u003Cp>如果你在做 SaaS、客服、搜尋或內容生成，我會建議你先算三個數字：每次請求成本、每月峰值流量、以及模型切換後的節省幅度。這三個數字，比 Demo 好不好看更重要。\u003C\u002Fp>\u003Ch2>接下來會怎麼走\u003C\u002Fh2>\u003Cp>我覺得接下來半年，雲端支出還會繼續偏高。原因很簡單，AI 需求沒有退燒，企業也還在試著把它塞進產品和內部流程。\u003C\u002Fp>\u003Cp>真正的變數是效率。如果模型更省 Token，或推論硬體更便宜，雲端帳單才可能慢下來。反過來說，只要大家還在拼命上 AI，雲端大廠的資本支出就很難停。\u003C\u002Fp>\u003Cp>對開發者來說，現在最實際的問題不是「要不要用 AI」，而是「怎麼用得不爆預算」。先把成本算清楚，再談規模，會比較像在做生意，不是在燒錢。\u003C\u002Fp>","Q4 2025 全球雲端基礎設施支出達 1109 億美元，年增 29%。AI 訓練、推論與雲端工作負載一起拉高 hyperscaler 投資。","www.ciodive.com","https:\u002F\u002Fwww.ciodive.com\u002Fnews\u002Fcloud-infrastructure-spend-rises\u002F816003\u002F",null,"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1775200180978-1ufe.png",[13,14,15,16,17,18,19,20],"雲端基礎設施","AI需求","hyperscaler","GPU","推論成本","AWS","Azure","Google Cloud","zh",1,false,"2026-04-03T07:09:17.443915+00:00","2026-04-03T07:09:17.417+00:00","done","e9c6688d-241d-423a-bce3-8ff005ce7501","cloud-infrastructure-spend-jumps-ai-demand-zh","industry","642b6179-d79b-40f9-bba3-6cf0d0daffc8","published","2026-04-07T07:41:09.787+00:00",[34,36,38,39,40,42,44,46],{"name":16,"slug":35},"gpu",{"name":14,"slug":37},"ai需求",{"name":17,"slug":17},{"name":13,"slug":13},{"name":18,"slug":41},"aws",{"name":19,"slug":43},"azure",{"name":20,"slug":45},"google-cloud",{"name":15,"slug":15},{"id":30,"slug":48,"title":49,"language":50},"cloud-infrastructure-spend-jumps-ai-demand-en","Cloud infrastructure spend jumps 29% on AI demand","en",[52,58,64,70,76,82],{"id":53,"slug":54,"title":55,"cover_image":56,"image_url":56,"created_at":57,"category":29},"cd078ce9-0a92-485a-b428-2f5523250a19","circles-agent-stack-targets-machine-speed-payments-zh","Circle 推出 Agent Stack，瞄準機器速度支付","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1778871663628-uyk5.png","2026-05-15T19:00:44.16849+00:00",{"id":59,"slug":60,"title":61,"cover_image":62,"image_url":62,"created_at":63,"category":29},"96d96399-f674-4269-997a-cddfc34291a0","iren-signs-nvidia-ai-infrastructure-pact-zh","IREN 綁上 Nvidia AI 基建","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1778871057561-bukp.png","2026-05-15T18:50:37.57206+00:00",{"id":65,"slug":66,"title":67,"cover_image":68,"image_url":68,"created_at":69,"category":29},"de12a36e-52f9-4bca-8deb-a41cf974ffd9","circle-agent-stack-ai-payments-zh","Circle 推出 Agent Stack 做 AI 付款","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1778870462187-t9xv.png","2026-05-15T18:40:30.945394+00:00",{"id":71,"slug":72,"title":73,"cover_image":74,"image_url":74,"created_at":75,"category":29},"e6379f8a-3305-4862-bd15-1192d3247841","why-nebius-ai-pivot-is-more-real-than-hype-zh","為什麼 Nebius 的 AI 轉型比炒作更真實","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1778823044520-9mfz.png","2026-05-15T05:30:24.978992+00:00",{"id":77,"slug":78,"title":79,"cover_image":80,"image_url":80,"created_at":81,"category":29},"66c4e357-d84d-43ef-a2e7-120c4609e98e","nvidia-backs-corning-factories-with-billions-zh","Nvidia 出資 Corning 工廠擴產","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1778822450270-trdb.png","2026-05-15T05:20:27.701475+00:00",{"id":83,"slug":84,"title":85,"cover_image":86,"image_url":86,"created_at":87,"category":29},"31d8109c-8b0b-46e2-86bc-d274a03269d1","why-anthropic-gates-foundation-ai-public-goods-zh","為什麼 Anthropic 和 Gates Foundation 應該投資 A…","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1778796636474-u508.png","2026-05-14T22:10:21.138177+00:00",[89,94,99,104,109,114,119,124,129,134],{"id":90,"slug":91,"title":92,"created_at":93},"ee073da7-28b3-4752-a319-5a501459fb87","ai-in-2026-what-actually-matters-now-zh","2026 AI 真正重要的事","2026-03-26T07:09:12.008134+00:00",{"id":95,"slug":96,"title":97,"created_at":98},"83bd1795-8548-44c9-9a7e-de50a0923f71","trump-ai-framework-power-speech-state-preemption-zh","川普 AI 框架瞄準電力、言論與州權","2026-03-26T07:12:18.695466+00:00",{"id":100,"slug":101,"title":102,"created_at":103},"ea6be18b-c903-4e54-97b7-5f7447a612e0","nvidia-gtc-2026-big-ai-announcements-zh","NVIDIA GTC 2026 重點拆解","2026-03-26T07:14:26.62638+00:00",{"id":105,"slug":106,"title":107,"created_at":108},"4bcec76f-4c36-4daa-909f-54cd702f7c93","claude-users-spreading-out-and-getting-better-zh","Claude 用戶更分散，也更會用","2026-03-26T07:22:52.325888+00:00",{"id":110,"slug":111,"title":112,"created_at":113},"bd903b15-2473-4178-9789-b7557816e535","openclaw-raises-hard-question-for-ai-models-zh","OpenClaw 逼問 AI 模型價值","2026-03-26T07:24:54.707486+00:00",{"id":115,"slug":116,"title":117,"created_at":118},"eeac6b9e-ad9d-4831-8eec-8bba3f9bca6a","gap-google-gemini-checkout-fashion-search-zh","Gap 把結帳搬進 Gemini","2026-03-26T07:28:23.937768+00:00",{"id":120,"slug":121,"title":122,"created_at":123},"0740e53f-605d-4d57-8601-c10beb126f3c","google-pushes-gemini-transition-to-march-2026-zh","Google 把 Gemini 轉換延到 2026 年 3…","2026-03-26T07:30:12.825269+00:00",{"id":125,"slug":126,"title":127,"created_at":128},"e660d801-2421-4529-8fa9-86b82b066990","metas-llama-4-benchmark-scandal-gets-worse-zh","Meta Llama 4 分數風波又擴大","2026-03-26T07:34:21.156421+00:00",{"id":130,"slug":131,"title":132,"created_at":133},"183f9e7c-e143-40bb-a6d5-67ba84a3a8bc","accenture-mistral-ai-sovereign-enterprise-deal-zh","Accenture 攜手 Mistral AI 賣主權 AI","2026-03-26T07:38:14.818906+00:00",{"id":135,"slug":136,"title":137,"created_at":138},"191d9b1b-768a-478c-978c-dd7431a38149","mistral-ai-faces-its-hardest-year-yet-zh","Mistral AI 迎來最硬的一年","2026-03-26T07:40:23.716374+00:00"]