Netlify Edge Functions go 5x faster with Firecracker MicroVMs
Original: 5x faster Edge Functions: V8 isolates to Firecracker MicroVMs
Why This Matters
Moving to MicroVMs signals a broader industry shift away from shared isolates toward stronger, faster edge compute primitives.
Netlify rebuilt its Edge Functions infrastructure, replacing V8 isolates with Firecracker MicroVMs running inside its own edge network. The result: median warm invocation latency dropped from 25–40ms to 5–6ms, p99 latency improved 47.4%, and log delivery is 5x faster. About one billion Edge Functions run on Netlify daily.
Netlify has overhauled the infrastructure powering Edge Functions, switching from a hosted V8 isolate execution service to Firecracker MicroVMs running directly inside its own edge network. The project was built in collaboration with Unikraft, which has published its own account of the work.
The numbers are stark. Warm invocations — covering routing to a compute node, entering a MicroVM, running the function, and producing response headers — now land at roughly 5–6ms at the median (p50), down from 25–40ms on the previous setup. P99 latency improved 47.4%, availability sits at 99.998%, and edge function log delivery is 5x faster.
Cold starts, which occur when a region's compute node hasn't seen a given request before, average about 9ms and affect only 1.2% of invocations.
At scale, the platform handles tens to hundreds of thousands of edge function invocations per second — serving workloads like Sunweb's page personalization and Loto-Québec's cookie-based traffic routing, among hundreds of thousands of other sites.
Netlify says the migration is invisible to developers: URL imports, npm packages, Node built-ins, netlify.toml declarations, and local development all work as before. The switch also improves security isolation and opens the door to heavier compute workloads at the edge.