Nvidia on Tuesday introduced the Jetson Orin Nano 2, a compact edge AI computer designed for entry-level robotics applications. The module delivers 78 trillion operations per second of AI processing power while maintaining a 15-watt power envelope, achieving double the inference performance of its predecessor with 40% lower energy consumption.
The device integrates an 8-core Arm CPU and 8 GB of memory, maintaining the form factor of earlier Jetson models. Nvidia positioned the Orin Nano 2 for use in autonomous drones, industrial inspection systems and consumer robotics, including delivery drones and home cleaning robots.
The company highlighted compatibility with large language models and visual language models optimized for edge inference, including Nvidia’s Cosmos and Nemotron suites alongside third-party models such as Gemma 4 and Qwen 3. Over 3 million developers have built solutions on Nvidia’s robotics platform, according to the company.
Early adopters include Matic Robotics, which plans to deploy the module in home cleaning robots to enable conversational AI, gesture detection and autonomous cleaning capabilities. Wing, Alphabet’s drone delivery unit, said it will evaluate the device for its fleet, currently running on the Jetson Orin Nano Super. Wing’s head of perception, Dinuka Abeywardena, noted the need for rapid, reliable real-world understanding in drone deliveries.
The Jetson Orin Nano 2 module and developer kit are scheduled for release in the first half of 2027.
