Azul says Azul Prime now delivers 2x to 5x faster application warm-up than standard OpenJDK, a change aimed at cutting the cloud capacity enterprises keep idle while Java services start up.
The gain comes from Cloud Native Compiler, part of Azul Optimizer Hub. Instead of each Java Virtual Machine (JVM) optimising code on its own, the compiler caches just-in-time compilations across a fleet and streams the optimised code to each new instance at start-up, before it takes traffic.
Why Java warm-up costs money
A new JVM instance starts on slower, less optimised code and learns which paths matter as it runs. Teams usually cope by over-provisioning compute or accepting slower performance after each deployment. Azul points to Cast AI‘s 2026 Kubernetes report, which put CPU over-provisioning at 69%, up from 40% a year earlier. It also cites Datadog‘s 2025 containers report, which found nearly two-thirds of Kubernetes organisations now scale automatically.
Fleet-wide learning without code changes
Azul says new instances receive the code the fleet has already optimised, and that switching the feature on takes one configuration change, with no rewrite, recompile or re-architecture. The company says operators can cut standby capacity and autoscale more dynamically as a result. It names fraud detection, real-time ad bidding, digital payments, multiplayer gaming and e-commerce as workloads where first-request performance matters most.
Scott Sellers, Azul’s co-founder and chief executive, said teams have treated slow warm-up as a given. “You stop paying the warm-up tax on every single instance,” he said.
Availability
Cloud Native Compiler comes at no additional charge with Azul Prime. Azul says its channel partners can use it to offer cost savings through professional services and managed deployments. The company has worked on warm-up since 2014, when it introduced ReadyNow, and extended it in 2023 with ReadyNow Orchestrator. The release does not detail the benchmark conditions behind the 2x to 5x range.
