Select a version
PROCESSOR-SDK-LINUX-RT-AM62A
RT-Linux SDK for Edge AI applications on AM62A
RT Linux Installer for sources, pre-built binaries and file system images with EdgeAI stack generated using Yocto build environment
RT Linux SD card image with EdgeAI Stack generated using Yocto build environment
RT Linux Target File System with EdgeAI Stack generated using Yocto build environment
Get access to latest bug fixes and feature enhancements with Yocto build environment
Products
Arm-based processors
Hardware development
Evaluation board
Documentation
Quick Start Guide for the AM62A Starter Kit (SK) board
Learn how to build embedded systems using AM62A processors
Edge AI SDK user manual
Yocto SDK user manual for developers
Instructions to build RT Linux EdgeAI image for AM62A Starter Kit (SK) board using Yocto build environment
Get the various RT Linux Performance Benchmarks for this release
Explore a selection of self-paced lessons for Edge AI development. Follow along with free cloud tools.
Build Sheet of supported features for AM62A
Software manifest file for AM62A
Release Information
Thank you for your interest in PROCESSOR SDK LINUX RT AM62A Software Development Kit (SDK) for Edge AI applications.
The Yocto build environment based RT-Linux SDK for Edge AI provides
- Upstream RT-Linux LTS Kernel with peripheral drivers, U-Boot bootloader.
- Pre-integrated firmware for hardware accelerated deep learning and camera ISP
- GStreamer plugins and Edge AI application stack
The SDK documentation provides simple steps to build, develop and run high performance deep learning applications using a live camera and display.
TI Linux CI/CD Snapshot
In order to improve access to the very latest bug fixes and feature enhancements, TI is enabling Continuous Integration and Continuous Deployment (CI/CD) for AM62A processors. The patches for these changes have always been available on public repositories hosted on git.ti.com, and these source repositories could be built using the Yocto build environment on a daily basis. The new TI Linux CI/CD Snapshot makes it even easier to consume these changes with pre-built binaries of Linux boot images to try out on TI Evaluation Modules (EVMs and SKs) before migrating or picking the patches. A detailed test report is also published to help users understand the status of the build and features being validated. The Linux boot images from a snapshot can be regenerated using sources by following the build instructions provided on the snapshot page.
The daily snapshot is not a replacement for the official Processors SDK releases which have full documentation, training and support. The snapshot is a new easy way to try bug fixes and feature enhancements planned for the next SDK while enabling much more tighter integration with customizations and application development. This improved integration should allow users to release their customized code built on the SDKs more quickly and confidently.
What's new
- First 2025 LTS Reference Release featuring the RT combined branch model, including RT Kernel with Real-Time Linux Interrupt Latency numbers
- Module safety for TIDL - TI Deep Learning Library
- Added support for 6 dimensions in Transpose and introduced several new operators: Unsqueeze, Acos, Atan, Cos, CosH, ELU, Neg, Tan, TanH, ScatterElements, and SiLU
- Out-of-the-Box EdgeAI Gallery App powered by QT 6.9 for seamless AI experiences
- Support for SELinux via meta-selinux with tisdk-edgeai-image
- Simplified the Yocto build process for tisdk-edgeai-image by eliminating the unnecessary edgeai branding step
- Support for suspend/resume of C7x in Edge AI Stack (currently context for VPAC and TIDL is not saved)
- EdgeAI DM R5 and C7x IPC firmwares are now delivered via ti-linux-firmware and default to remote endpoint 14, ensuring consistency across Sitara AM6x platforms with Linux RPMsg user-space applications.
- Falcon mode through R5 SPL