Edge Computing
Running code in data centers physically close to users, instead of one central origin.
Edge platforms like Cloudflare Workers, Vercel Edge Functions, Deno Deploy, and Fastly Compute@Edge run your code in dozens or hundreds of locations worldwide. A user in Tokyo gets served from Tokyo, not from your origin in Virginia.
The upside is dramatically lower latency for global users and natural protection against load spikes. The trade-offs are constraints: edge functions are usually short-lived, run on lightweight runtimes (V8 isolates, not full Node), and have limited filesystem and database options.
Edge is well-suited to API gateways, auth, A/B testing, personalization, and lightweight HTML rendering. Heavy compute (AI inference, video encoding) usually still runs centrally, with edge serving as the fast front door.