NVIDIA Jetson Orin Nano 2 Brings Frontier AI to Entry-Level Robotics
The new edge AI module doubles inference performance and slashes power consumption to enable real-time LLM and VLM deployment on compact devices.
NVIDIA has introduced the Jetson Orin Nano 2, an entry-level edge AI computer designed to bring advanced reasoning to robotics and physical AI applications. The module enables the deployment of frontier AI models, including Large Language Models (LLMs) and Vision Language Models (VLMs), directly on edge hardware for real-time interaction.
The hardware delivers 78 trillion operations per second (TOPS) of AI compute, paired with 8GB of memory and an eight-core Arm CPU. NVIDIA reports that the Orin Nano 2 doubles the inference performance of the Jetson Orin Nano Super, a leap achieved through higher memory bandwidth and improved Tensor Cores. Efficiency is a primary focus; in 15-watt mode, the module consumes 40% less power than its predecessor while delivering the same peak-to-peak performance.
A Full-Stack Robotics Strategy
The release is a key component of NVIDIA's "three-computer, full-stack" robotics strategy. This ecosystem integrates Omniverse with Cosmos for simulation, DJX for training, and the Jetson line for runtime deployment. By refreshing its entry-level hardware following updates to the Orin NX and T-series modules, NVIDIA is migrating data-center-grade GPU architecture into compact, energy-efficient form factors.
This push comes as the company's ecosystem scales. NVIDIA reports that more than 3 million developers are building on its robotics stack, and over 10,000 companies are either developing or shipping products powered by Jetson hardware. The Orin Nano 2 is built to support a wide array of open models, including NVIDIA Cosmos, Nemotron, Gemma 4, and Qwen 3.
The Shift to Physical AI
Bringing frontier-level intelligence to entry-level hardware lowers the barrier for autonomous reasoning in drones and home robotics. By moving beyond simple programmed tasks toward complex semantic understanding, devices can operate without relying on cloud connectivity. Deepu Talla, VP of Robotics and Edge AI at NVIDIA, noted that today’s small and medium frontier models have reached the accuracy of last year’s largest models, unlocking real-time intelligence for edge devices. Talla added that this capability enables a "new class of applications" for companion robots, which have historically been limited by poor intelligence.
Industry partners view this as a critical step for specialized autonomy. Dinuka Abeywardena, Head of Perception at Wing, stated that drone delivery specifically depends on AI that enables a fast and reliable understanding of the physical world.
Availability and Outlook
The Jetson Orin Nano 2 module and its accompanying developer kit are expected to be available in the first half of 2027. As NVIDIA continues to shrink the gap between cloud-based AI and edge execution, the industry will be watching to see how quickly these efficiency gains translate into commercially available consumer robots and industrial autonomous systems.