Unleashing the Power of AI: NVIDIA Jetson Orin Nano 2 Takes Flight with Drones and Robots

NVIDIA has introduced the Jetson Orin Nano 2, a cutting-edge robotics computer designed to bring artificial intelligence to drones, robots, and vision systems at the edge. The company is targeting developers who seek to run generative AI models directly on a device rather than in a data center.

The rationale behind NVIDIA’s launch is based on the enhanced performance of small and medium AI models. These models now achieve accuracy levels previously only attained by larger frontier models a year ago.

Deepu Talla, NVIDIA’s VP of Robotics and Edge AI, stated, “The Jetson Orin Nano 2 computer makes this breakthrough accessible to millions of developers, providing the performance and efficiency required for real-time decision-making at the edge.”

The accuracy achieved by small and medium AI models enables compact edge hardware to interpret language and images and respond in real-time. Robots, delivery drones, inspection drones, and vision AI systems rely on hardware capable of handling these workloads without consuming excessive power.

Jetson Orin Nano 2 boasts 78 trillion operations per second of AI compute, 8GB of memory, and an eight-core Arm CPU. NVIDIA engineered the board to deliver increased AI and video-processing performance while maintaining low cost and power consumption.

The new board delivers twice the inference performance of the existing Jetson Orin Nano Super, thanks to improved Tensor Cores and higher memory bandwidth within the same compact form factor. Operating in 15-watt mode, Jetson Orin Nano 2 uses 40 percent less power than the Orin Nano Super while matching its performance.

Jetson Orin Nano 2 is designed to run on NVIDIA’s open software stack alongside Jetson agent skills and the broader Jetson AI ecosystem. The board is optimized to support large language models and vision language models for memory-efficient inference at the edge. NVIDIA cites its Cosmos, Nemotron, Gemma 4, and Qwen 3 models as examples that developers can deploy on the hardware.

Early partners such as Cognex, Doosan Bobcat, and Matic are among the first companies to adopt and explore Jetson Orin Nano 2. NVIDIA anticipates that these partners, along with others, will integrate edge AI into various applications, including home robots, vision AI systems, delivery drones, carrier boards, hardware systems, and reference designs.

Wing, Alphabet’s drone delivery subsidiary, currently utilizes Jetson Orin Nano Super and NVIDIA’s software stack in its delivery drone fleet. The company intends to assess Jetson Orin Nano 2 to enhance real-time AI perception and reasoning for quicker and safer deliveries from local businesses to residential yards.

Matic Robots, a consumer robotics company, is implementing Jetson Orin Nano 2 in its home cleaning robots. The board will enable Matic to incorporate conversational AI, gesture detection, precise mapping, semantic understanding of the home, and autonomous cleaning behavior.

NVIDIA has also enlisted a wide range of Jetson hardware partners working on carrier boards and reference designs for the new board. These partners are collaborating to help customers expedite their entry into the market.

In conclusion, NVIDIA’s Jetson Orin Nano 2 signifies a significant advancement in bringing physical AI to the edge, allowing for real-time decision-making in various applications. The board’s enhanced performance and energy efficiency make it a valuable tool for developers seeking to deploy AI models locally.