Library for fast mapping of Java records to native memory

TypedMemory, a new Java library designed to map Java record types onto native memory, has been released as an experimental project targeting Java 25 and above. It aims to simplify high-performance off-heap memory management for systems, graphics, and data-oriented programming.

Developed by Mamba Studio, TypedMemory leverages the Java Foreign Function & Memory (FFM) API introduced in Java 25 to provide a strongly typed interface for off-heap memory. It allows developers to allocate, access, and manipulate contiguous native memory using Java records, reducing boilerplate and improving safety compared to manual memory management. The library supports features such as memory layout derivation, bulk operations, reinterpretation of memory segments, and nested structured data.

Currently, TypedMemory is in an early, experimental stage. Its core API is functional, but the project may undergo breaking changes as it matures. Future plans include support for pointer-typed fields beyond simple addresses and union types. The library is available via Maven Central, requiring Java 25 or higher and native access permissions.

Why It Matters

This development matters because it addresses longstanding challenges in Java’s low-level memory management. By providing a type-safe, expressive API for off-heap memory, TypedMemory enables more efficient native interop, data processing, and high-performance workloads such as graphics, simulation, and large dataset handling. Its design aims to reduce complexity and errors associated with manual memory layout handling, potentially influencing future Java native interop practices.

Background

Java has historically limited direct control over off-heap memory, often relying on JNI or unsafe APIs that are verbose and error-prone. The introduction of the FFM API in Java 25 offers a more structured approach, but high-level abstractions remain limited. TypedMemory builds on this foundation, providing a more developer-friendly interface for structured, high-performance off-heap data management. Its release follows ongoing industry interest in native interop and memory efficiency, especially for systems programming, graphics, and data science.

“TypedMemory aims to make off-heap programming in Java more natural by offering a strongly typed, low-level memory abstraction.”

— Mamba Studio

“TypedMemory leverages the Java Foreign Function & Memory API to provide a high-level, type-safe interface for off-heap data.”

— Java FFM API team

What Remains Unclear

It remains unclear how widely adopted TypedMemory will become, given its early-stage status and potential for breaking changes. The full set of planned features, such as support for pointer-typed fields and unions, is still under development. Additionally, performance benchmarks and real-world use cases are yet to be published, making the practical impact uncertain at this stage.

What’s Next

Next steps include the stabilization of the core API, the addition of planned features like pointer fields, and broader testing in real-world scenarios. Developers interested in native interop and high-performance data layouts are encouraged to experiment with the library and contribute feedback. Further releases are expected to refine the API and expand capabilities, with potential integration into larger Java projects.

Key Questions

What is TypedMemory?

TypedMemory is an experimental Java library that maps Java record types onto native memory, enabling efficient off-heap data management with a type-safe API.

What are the main features of TypedMemory?

It supports memory allocation, get/set operations, layout derivation, bulk copying, reinterpretation of memory segments, nested structured data, and wrapping existing memory segments.

Is TypedMemory stable and production-ready?

No, it is currently in an early, experimental stage. The API may change as the project develops, and it is primarily intended for testing and research purposes.

What are the requirements to use TypedMemory?

You need Java 25 or higher, with native access permissions enabled. The library is available via Maven Central for easy integration.