How to use the AOT capabilities by Leyden and show test results which show that very significant progress has already been made.

Foojay.io's platform is a central hub for Java developers, offering insights into Java programming language, JVM ecosystem, and Java-related technologies. Through articles, tutorials, and community forums, Foojay.io offers insights into building scalable and maintainable Java applications. Developers can learn about Java language features, performance tuning tips, and best practices in Java development to write efficient and reliable software.

Foojay.io

Project Leyden brings AOT caching to OpenJDK (JDK 25+) to reduce Java startup time, warmup time, and memory footprint. The post explains how to set up an AOT cache using either a 2-step (joint training+assembly) or 3-step (separate training, assembly, production) workflow with specific JVM flags. Benchmark results on JDK 26 using a Quarkus-based math-heavy REST API and a simple Hibernate ORM REST API show that AOT caching cuts time-to-first-response roughly in half and significantly reduces JIT warmup jitter. Longer training runs (60,000 vs 1,000 requests) produce better-optimized caches with more compiled methods. Constraints include matching JVM version, CPU family, OS, and JVM arguments between training and production runs. The Leyden premain JVM additionally caches compiled code, promising even larger gains in future JDK releases.

How is Leyden improving Java Performance? Part 2 of

Should I start using AOT Cache in Java already?