Introduction

NVIDIA Jetson platforms are widely used for embedded AI, multimedia, robotics, and accelerated edge computing applications. As Jetson use cases continue to expand, developers increasingly require deeper control over deployed systems and platform customization.

In many cases, this includes package integration, kernel customization, application integration, OTA mechanisms, system optimization, and BSP modifications. Because of this, Yocto has become an essential tool for building customized Jetson based systems.

Traditionally, Yocto support for Jetson platforms relied heavily on community maintained integrations such as meta-tegra. NVIDIA is now taking the next step by officially maintaining Yocto support directly aligned with JetPack releases and NVIDIA BSP components.

This wiki provides practical guidance for developers working with NVIDIA maintained Yocto integration, including platform setup, image generation, deployment workflows, multimedia validation, and BSP customization.


NVIDIA Yocto Layers

Official NVIDIA Yocto Layers

Below you can find the Yocto layers currently supported by NVIDIA and a brief description of each one. Click on any layer to open its repository.

Core Layer
meta-tegra
Main integration layer for Jetson platforms, including BSP support, flashing workflows, multimedia integration, and NVIDIA accelerated components.
Community
meta-tegra-community
Additional community-maintained integrations, external packages, board support, and extensions built around the Jetson Yocto ecosystem.
Reference
demo-distro
Reference distribution showcasing NVIDIA Yocto workflows, example configurations, multimedia support, and development examples.

What You Will Learn?

This guide walks through the process of getting started with the Yocto layers released by NVIDIA and building your first working Jetson image with them. By the end of this wiki, you will understand:

Build
Jetson Images
Create Linux images using NVIDIA maintained platform support and Yocto build infrastructure.
Deploy
Flash the Board
Generate deploy artifacts and program supported Jetson platforms.
Validate
Runtime Functionality
Confirm system access, boot stability, networking, services, and multimedia readiness.
Customize
Platform Integration
Integrate multimedia packages, APIs, device trees, overlays, BSP components, and custom device drivers.

Why Choose Yocto?

The Yocto Project is an open source framework used to create custom Linux distributions for embedded systems. Instead of relying on a fixed prebuilt operating system, Yocto allows developers to define the packages, services, libraries, kernel components, and platform configuration included in the final image.

This becomes especially important in embedded development, where systems often require reproducibility, maintainability, hardware optimization, and long-term platform control.

Benefits of using Yocto:

Reproducibility
Consistent Builds
Rebuild the same image configuration across developers, CI systems, validation environments, and production deployments with predictable results.
Optimization
Controlled Footprint
Include only the packages, services, and libraries required by the target product, helping reduce storage usage and unnecessary dependencies.
Integration
Hardware Enablement
Integrate kernel configuration, BSP components, GPU acceleration, multimedia support, and hardware-specific features directly into the image.
Maintenance
Long-Term Platform Control
Maintain a structured development environment for updates, validation, debugging, and long-term product support.

Why NVIDIA Jetson?

NVIDIA Jetson is a family of embedded computing platforms designed for AI, robotics, multimedia, industrial automation, and accelerated edge applications.

Jetson platforms combine GPU acceleration, multimedia hardware blocks, and low power ARM based processing into compact systems suitable for embedded and edge deployments.

AI
Edge Inference
Run AI and computer vision workloads directly on embedded edge devices with GPU acceleration.
Multimedia
Video Acceleration
Use dedicated multimedia hardware for video capture, encoding, decoding, streaming, and image processing.
Power
Low Power Deployment
Build compact systems optimized for robotics, industrial systems, smart cameras, and edge devices.
Platform
Embedded Integration
Combine Linux customization, hardware acceleration, multimedia integration, and platform-specific enablement within a unified embedded stack.