Jetson Orin Nano 2 consumes less power at the same performance level of its predecessor. Source: NVIDIA
As AI models become more efficient, more devices can become autonomous, but developers need compact, energy-efficient computers built for edge AI, asserted NVIDIA Corp. The company today introduced NVIDIA Jetson Orin Nano 2, a new entry-level computer for edge AI that it said enables millions of developers worldwide to build systems for physical AI applications.
“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,” stated 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 in smart drones, robots, and vision AI systems.”
“Over the past three months, we’ve seen the evolution of open models powering physical AI and how important they’ve become,” he added during a press briefing. “One year ago, frontier models which were 600 billion or 1 trillion parameter models. A year later, we’re able to bring that level of accuracy all the way down to intro-level edge AI products in Orin Nano 2. This is amazing.”
NVIDIA Jetson Orin Nano 2 is part of the company’s “three-computer, full-stack” approach to robotics, with Omniverse withCosmosprovidingsimulationfor testing, DJX supporting training, andJetsonas the “robot brain” providing runtime deployment.
“Jetson Orin Nano 2 delivers a significant leap inAIandvideoprocessing performance in a cost-effective, power-efficient system,” said NVIDIA. The system features 78 trillion operations per second (TOPS) of AI compute, 8GB of memory, and an eight-coreArmCPU.
Jetson Orin Nano 2 doubles the inference performance of Jetson Orin Nano Super through improved Tensor Cores and higher memory bandwidth, while maintaining the same compact form factor. In 15-watt mode, Jetson Orin Nano 2 consumes 40% less power to deliver the same peak-to-peak performance as its predecessor, Talla toldThe Robot Report.
NVIDIA noted that its existing open softwarestackcan run on Nano 2, which is built on the same GPU architecture as data centers. It is designed to be a drop-in for existing Orin customers.
In combination with Jetson agent skills and thecompany‘s technology ecosystem, developers can now run the latest large language models (LLMs) and vision language models (VLMs) optimized for memory-efficient edge inference. These include open models such as NVIDIA Cosmos and Nemotron, Gemma 4, and Qwen 3.
“This now suddenly unlocks a level of intelligence that was impossible — we’ve been dreaming about this for a decade in edge AI,” said Talla during a press briefing. “This is a significant moment in time, where now for all of these robotics and physical AI applications, we can actually put frontier AI models such as LLMs and VLMs on top of the other autonomous capabilities. We believe this will speed up the development and deployment of AI in the physical world.”
NVIDIA had already refreshed its higher-end Jetson offerings, including the Orin NX, Orin, and T3000/T2000 for mainstream applications and the Jetson T5000/T4000 for advanced uses.
NVIDIA said that more than 3 million developers arealready buildingon its robotics stack. Partners such asCognex,Doosan Bobcat, Matic, and Wing are evaluating or using Jetson Orin Nano 2 for real-world edge AI in compact devices.