NVIDIA’s Jetson Orin Nano 2 brings edge AI to drones and robots
NVIDIA has launched the Jetson Orin Nano 2, a compact edge AI computer built to run generative AI models directly on drones, robots, and vision systems instead of relying on distant data centers. The board targets developers who need real-time reasoning without the bulk or power draw of larger systems, marking a shift toward smaller, more efficient AI hardware.
Smaller models, bigger impact
Until recently, only large AI models could deliver the accuracy required for tasks like interpreting language or processing images in real time. NVIDIA argues that smaller and medium-sized models have now caught up, making it feasible to deploy them on low-power edge devices. The Jetson Orin Nano 2 packs 78 trillion operations per second of AI compute, 8GB of memory, and an eight-core Arm CPU into the same compact form factor as its predecessor. Despite its size, it delivers twice the inference performance of the earlier Jetson Orin Nano Super while using 40% less power—15 watts compared to the Super’s 25 watts. The gains come from improved Tensor Cores and higher memory bandwidth, enabling faster processing without sacrificing efficiency.
Software and partners pave the way
The board runs on NVIDIA’s open software stack, which includes Jetson Agent Skills and the broader Jetson AI ecosystem. It supports large language models and vision-language models optimized for memory-efficient edge inference, such as NVIDIA’s Cosmos and Nemotron, as well as Gemma 4 and Qwen 3. Early adopters include Cognex, Doosan Bobcat, and Matic, with more than three million developers already building on NVIDIA’s robotics stack. Wing, Alphabet’s drone delivery subsidiary, currently uses the Jetson Orin Nano Super in its fleet and is evaluating the new board to enhance real-time AI perception and reasoning for faster, safer deliveries.
Why it matters
The Jetson Orin Nano 2 signals a turning point for edge AI, making it possible to run sophisticated AI models on compact, low-power hardware. For industries like robotics and drone delivery, this means smarter, more responsive systems that don’t depend on cloud connectivity—opening doors to new applications and business models. The hardware’s efficiency and performance could accelerate adoption across sectors, from home robots to industrial inspection, while keeping costs and energy use in check.
Source: AI News. AI-assisted editorial synthesis — TechnoExpress.