https://vercel.com/docs/fluid-compute#how-to-enable-fluid-compute
The doc says it has better cold start, but I'm not sure what to expect. Let's say I have an Express + Node server that connects to a db on startup. On traditional scale-to-zero platforms, the cold starts take thousands of milliseconds. Does Fluid Compute stop the instance at all? If it does, would the cold start be much faster than something like AWS lambda?
If you're having trouble deploying an Express app, this guide can help.
https://vercel.com/guides/using-express-with-vercel
You can also ask v0 for suggestions tailored to your own project setup.
Hi , welcome to the Vercel Community!
Fluid compute bytecode caching that improves the cold start on subsequent requests for your application. In addition to this, Fluid Compute allows multiple invocations to share a single function instance, which further reduces the cold start problem.
There's also a short explainer video for Fluid Compute on our YouTube channel that I'll highly recommend watching.
