Vercel 推出 Fluid 運算架構：以「Scale to One」策略徹底解決 Serverless 冷啟動問題
Original: Scale to one: How Fluid solves cold starts
In Serverless architectures, "cold starts" have long been the biggest pain point affecting user experience and website performance metrics…
Vercel 發表全新的「Fluid」運算架構，旨在解決困擾 Serverless 開發者已久的「冷啟動（Cold Start）」問題。傳統 Serverless 採用「Scale to Zero（縮減至零）」以節省成本，卻會導致首個請求延遲；Fluid 則引入「Scale to One」機制，透過極低開銷維持至少一個微型運算實例處於熱機狀態，讓應用程式能瞬間回應，大幅提升網站的 TTFB 與使用者體驗。
In Serverless architectures, "cold starts" have long been the biggest pain point affecting user experience and website performance metrics such as TTFB and LCP. To help developers save costs, traditional Serverless platforms typically adopt a "Scale to Zero" strategy — when there is no traffic for a period of time, all computing resources are released. However, when the next request arrives, the system must re-provision containers, load code, and start the runtime environment, producing a cold start delay of several seconds.
Free shows the 3-line summary; Pro unlocks the full deep summary (~300 words) so you never have to click through.
See Pro plans →Want the original English / full article?
Read on Vercel Changelog →Summaries are AI-generated; the original article is authoritative.