Howdy everyone, its Your Friendly Neighborhood Navigator here with exciting news. AMD and Open Navigation have been collaborating over the last several months to create demonstrations of AMD’s powerful Ryzen AI, Embedded+, and Kria capabilities for the ROS 2 community!
We’ve started by showing how you can use Ryzen AI to power your autonomous robotics products in a variety of real-world and challenging situations. These devices contain the powerful, modern AMD x86 CPUs you know and love, married with a Radeon GPU, Neutral Processing Unit (NPU), and more to run hardware accelerated and AI applications on your robot products. We have benchmarked that all of autonomous navigation, 3D perception, localization, and autonomy applications can be performed with under 15% of total resource utilization, leaving plenty of room for many user applications, advanced AI, and more – all for an amazing price point & power budget. If you are working on robotics systems and need powerful x86 cores, we definitely recommend looking into Ryzen AI.
To this end, we’ve created 3 real-world demonstrations and released all of its source code and instructions to the community to get started from! We’ll be building on top of these in the future using the AI Acceleration and Hardware Acceleration capabilities of AMD’s technologies to make it easy to deploy AI and advanced robotics algorithms.
Check out GitHub - open-navigation/opennav_amd_demonstrations: Project containing demonstrations using AMD's Ryzen AI and other technologies with ROS 2 to see the videos, download datasets, learn how to setup and use ROS 2, Nav2 for Indoor, Urban, and 3D SLAM-based navigation!
Happy Ryzening!
Steve
P.S. as a bonus: We also include a series of guides for how to bootstrap a new robot: PC, software bringup, time synchronize multiple machines, and more!
