> Markdown version of [/videos/100161-supercharge-your-jvm-performance-with-project-leyden?t=665](https://www.wearedevelopers.com/videos/100161-supercharge-your-jvm-performance-with-project-leyden?t=665). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Supercharge your JVM performance with Project Leyden

Are slow Java startup times penalizing your cloud-native apps? Project Leyden eliminates these bottlenecks by building an AOT cache, supercharging JVM performance without sacrificing peak throughput.

- **Speakers:** [Ana Maria Mihalceanu](https://www.wearedevelopers.com/@ana-maria-mihalceanu)
- **Event:** World Congress 2026 Europe
- **Published:** July 9, 2026
- **Duration:** 30:59
- **URL:** https://www.wearedevelopers.com/videos/100161-supercharge-your-jvm-performance-with-project-leyden

## Summary

Java's historic preference for peak performance via HotSpot profiling often resulted in slow startup and warm-up times, penalizing dynamic cloud-native applications. Project Leyden resolves these bottlenecks by intelligently shifting initialization work earlier—either at build time or during a pre-production training run. By dynamically observing class loading and hot methods to build an Ahead-of-Time (AOT) cache, Leyden dramatically reduces the time to the first useful HTTP response without sacrificing peak throughput. Implementing this optimization relies on executing targeted training runs, ideally utilizing smoke tests that represent typical startup paths rather than exhaustive regression suites that can bloat the cache size. Developers can adopt a discrete three-step workflow or a streamlined process combining training and assembly, while frameworks like Spring Boot amplify these gains through jar extraction tools, non-lazy singleton creation via context refresh, and dedicated Spring AOT integrations. Successfully operationalizing Project Leyden requires strict parity; the AOT binary demands identical JDK versions, hardware architectures, and ordered classpaths between CI/CD generation and production runtime. Security is paramount, requiring teams to treat this binary with the same stringent volume-mounting privileges as a primary JAR file. Furthermore, upgrading to JDK 26 introduces robust garbage collector compatibility—such as utilizing GC-agnostic streamable objects for ZGC users or massive heaps—eliminating previous memory pointer limitations and solidifying Leyden as a transformative leap in JVM scalability.

**Keywords:** project leyden, jvm performance optimization, java AOT caching, hotspot JVM warm-up, spring boot AOT integration, microservices startup latency, JDK 26 garbage collection, docker multi-stage java builds, ZGC streamable objects, java classpath configuration, kubernetes cache volumes, smoke testing cache generation, jar mode tools extraction, CI/CD cache generation

## Chapters

1. **Goals of Project Leyden for JVM performance** (00:02) — How Project Leyden addresses Java's historical trade-off between peak performance and initial application startup delays.
1. **Approaches to shifting JVM initialization work** (02:26) — Strategies for accelerating application readiness by shifting compilation work to earlier or later execution phases.
1. **Understanding the three-phase AOT caching workflow** (03:47) — Exploring the initial workflow involving training, assembly, and execution to improve runtime initialization metrics.
1. **Implementing the streamlined two-phase workflow in JDK 25** (06:13) — How JDK 25 condenses training and assembly into a single operational step for more efficient AOT profiling.
1. **Essential requirements for building an AOT cache** (08:27) — Strict environmental prerequisites for generating valid cache profiles ranging from hardware consistency to class path integrity.
1. **Optimizing Spring Boot applications with AOT caching** (11:05) — Leveraging specific framework tooling and expanded jar layers to streamline the AOT extraction process.
1. **Demonstrating the Spring Petclinic AOT cache generation** (12:40) — A practical walkthrough showing application compilation and accelerated startup commands in a local development environment.
1. **Achieving further performance gains by using Spring AOT** (13:58) — Combining the core JVM AOT cache with native compilation subsets to minimize deployment overhead constraints.
1. **Operational containerization approaches and CI/CD strategies** (15:28) — Deployment strategies for pre-generating application caches in CI pipelines and reusing them seamlessly across containerized environments.
1. **Understanding AOT cache limitations and deployment security vulnerabilities** (18:48) — Crucial warnings regarding file interoperability failures, security isolation needs, and architectural dependency mismatches.
1. **Resolving AOT garbage collection compatibility in JDK 26** (21:01) — How upcoming major releases resolve problematic heap layout disparities between different garbage collectors via streamable objects.
1. **Audience Q&A covering Project Leyden compatibility and pipelines** (24:55) — Addressing community inquiries regarding reflection API interactions, compilation pipelines, testing accuracy, and GraalVM alternative comparisons.

## Related Moments

- [Optimizing Java startup processes for fast serverless workloads](https://www.wearedevelopers.com/videos/1008-serverless-java-in-action-cloud-agnostic-design-patterns-and-tips) (from "Serverless Java in Action: Cloud Agnostic Design Patterns and Tips")
- [Optimizing Java performance with tiered compilation concepts](https://www.wearedevelopers.com/videos/240-just-in-time-compilation-in-jvm) (from "Just-in-time Compilation in JVM")
- [Optimizing container workloads with Quarkus and lightweight JVMs](https://www.wearedevelopers.com/videos/89-development-of-reactive-applications-with-quarkus) (from "Development of reactive applications with Quarkus")
- [Prioritizing application startup efficiency over systemic throughput metrics](https://www.wearedevelopers.com/videos/104-cloud-chaos-and-microservices-mayhem) (from "Cloud Chaos and Microservices Mayhem")
- [Load testing and native AOT pre-compilation performance](https://www.wearedevelopers.com/videos/100207-in-depth-net-azure-functions-isolated-mode-performance-and-durable-ai-agents) (from "In-depth .NET Azure Functions: Isolated mode, performance and durable AI agents")
- [Evaluating performance tradeoffs between JVM and native images](https://www.wearedevelopers.com/videos/443-going-serverless-using-the-spring-framework-ecosystem) (from "Going serverless using the Spring Framework ecosystem")

## Related Articles

- [Top 10 Java Libraries](https://www.wearedevelopers.com/magazine/364-top-10-java-libraries)
- [Using Java 17 latest features in real world projects](https://www.wearedevelopers.com/magazine/121-using-java-17-latest-features-in-real-world-projects)
- [MLops – Deploying, Maintaining And Evolving Machine Learning Models in Production](https://www.wearedevelopers.com/magazine/115-mlops-deploying-maintaining-and-evolving-machine-learning-models-in-production)
- [Stephan Gillich - Bringing AI Everywhere](https://www.wearedevelopers.com/magazine/489-stephan-gillich-bringing-ai-everywhere)

## Related Jobs

- [Softwareentwickler Java](https://www.wearedevelopers.com/jobs/ext/1245896-softwareentwickler-java) at **Aok Systems Gmbh**
- [Java Softwareentwickler - Spring Boot & Microservices](https://www.wearedevelopers.com/jobs/ext/1306817-java-softwareentwickler-spring-boot-microservices) at **Aok Systems Gmbh**
- [Senior Backend Engineer (Java)](https://www.wearedevelopers.com/jobs/ext/19369-senior-backend-engineer-java) at **Bonial International GmbH**
- [Softwareentwickler Künstliche Intelligenz mit Schwerpunkt Projektkoordination](https://www.wearedevelopers.com/jobs/ext/1247541-softwareentwickler-kunstliche-intelligenz-mit-schwerpunkt-projektkoordination) at **Aok Systems Gmbh**
- [Lead Software Engineer - Data Engineering](https://www.wearedevelopers.com/jobs/ext/2000968-lead-software-engineer-data-engineering) at **Dynatrace**
- [Principal Engineer - AI Search & Vector Infrastructure](https://www.wearedevelopers.com/jobs/ext/353953-principal-engineer-ai-search-vector-infrastructure) at **Redis**