NVIDIA Debuts Jetson Orin Nano 2 to Power Real-Time Generative AI on Edge Robotics

NVIDIA Corp. announced the Jetson Orin Nano 2, a compact robotics computer delivering 78 TOPS of AI performance for edge devices and autonomous systems, with twice the inference performance of the previous Jetson Orin Nano Super while using 40% less power in 15W mode. The platform supports generative AI models locally, reducing cloud dependence, and has been adopted by partners including Cognex, Doosan Bobcat, Matic Robots, and Alphabet's Wing.

TL;DR — Key Takeaways - NVIDIA’s Jetson Orin Nano 2 delivers 78 TOPS of AI performance in a compact, low-power platform aimed at robotics and edge AI. - The system offers roughly twice the inference performance of the Jetson Orin Nano Super while using 40% less power at matched performance in 15W mode. - The platform is designed to run modern generative AI models locally, reducing dependence on cloud infrastructure for real-time workloads. NVIDIA Corp. announced on Tuesday the Jetson Orin Nano 2, a compact robotics computer engineered to bring frontier-class generative artificial intelligence AI performance to entry-level edge devices and autonomous systems. Designed for deployment of physical machines such as delivery drones, vision systems, and consumer robotics, the new processor provides developers with the low-power computational capabilities required to run modern AI models locally without relying on cloud infrastructure, according to NVIDIA https://www.hpcwire.com/off-the-wire/nvidia-unveils-jetson-orin-nano-2-for-robotics-and-edge-ai/ . As AI models become smaller and more efficient, edge devices can execute complex real-time tasks, including contextual reasoning, speech interpretation, and image processing. The Jetson Orin Nano 2 delivers 78 trillion operations per second TOPS of AI compute, supported by an 8-core Arm CPU and 8GB of memory. According to NVIDIA, the device achieves twice the inference performance of the previous Jetson Orin Nano Super model through upgraded Tensor Cores and expanded memory bandwidth, while maintaining the same physical form factor. When operating in its 15-watt mode, the processor consumes 40 percent less power to match the performance output of its predecessor. “Today’s small and medium frontier models have reached the accuracy of last year’s largest frontier models, unlocking real-time intelligence for edge devices,” said Deepu Talla, vice president of robotics and edge AI at NVIDIA. “The Jetson Orin Nano 2 computer puts that breakthrough within reach of millions of developers, delivering the performance and energy efficiency needed for real-time reasoning at the edge.” The module operates on NVIDIA’s open software stack and supports popular open-source models optimized for edge inference, including NVIDIA Cosmos, NVIDIA Nemotron, Gemma 4, and Qwen 3. The ecosystem currently supports more than 3 million developers working across robotics and edge AI applications. Several technology and industrial partners have already begun integrating the new platform, including Cognex, Doosan Bobcat, Matic Robots, and Alphabet drone delivery subsidiary Wing. Wing plans to evaluate the Jetson Orin Nano 2 for its delivery drone fleet to improve onboard perception and enable safer, faster autonomous flights. “Drone delivery depends on AI that can enable fast, reliable understanding of the real world,” said Dinuka Abeywardena, head of perception at Wing. “Wing is exploring Jetson Orin Nano 2 to give us a path to more responsive, energy-efficient drones.” Similarly, consumer robotics startup Matic Robots is adopting the processor to enhance its home cleaning units with conversational AI, gesture recognition, precise semantic mapping, and real-time navigation. Navneet Dalal, co-founder and CEO of Matic Robots, noted that the platform enables the execution of state-of-the-art AI models directly within a compact hardware footprint required for dynamic indoor environments.