Nvidia Corporation has unveiled an updated version of its compact robotics computer — the Jetson Orin Nano 2. This device, aimed at edge AI computing, demonstrates a twofold increase in inference performance compared to its predecessor while maintaining the same form factor. The key feature is the ability to run modern language and multimodal models directly on board, without relying on cloud servers.
Technical Breakthrough in a Compact Form Factor
The new device is equipped with an 8-core processor based on the Arm architecture, 8 GB of RAM, and delivers performance of up to 78 TOPS. Thanks to improved tensor cores and increased memory bandwidth, inference efficiency has doubled relative to the Jetson Orin Nano Super model. Moreover, at a power consumption of 15 W, the device achieves the previous performance level while reducing energy costs by 40%.
This is critically important for autonomous systems — robots, drones, and computer vision systems that must process data in real time without a stable cloud connection. Reduced power consumption directly impacts the autonomy and operating time of such devices.
Physical AI: From Perception to Action
Nvidia positions the Jetson Orin Nano 2 as a platform for physical AI — systems capable not only of analyzing the environment but also of making decisions, interacting with people, and acting autonomously. The device supports models optimized for edge inference, including NVIDIA Cosmos, NVIDIA Nemotron, Gemma 4, and Qwen 3. This opens up opportunities for speech and image recognition, navigation, and complex decision-making directly on the device.
It is especially noteworthy that modern compact models already achieve accuracy comparable to large systems of the previous generation. This marks a paradigm shift: more and more generative AI functions are migrating from data centers to edge devices.
Ecosystem and Early Adoptions
By my estimates, Nvidia's robotics stack is already used by more than 3 million developers. Among the first companies testing the new device are machine vision systems manufacturer Cognex, industrial giant Doosan Bobcat, and robotics startup Matic, which uses the platform in home robots for simultaneous room mapping, object recognition, and autonomous task execution.
The autonomous drone scenario deserves special attention: Alphabet's Wing division, which uses the previous generation of Jetson in delivery systems, plans to evaluate the new version to accelerate data processing and improve energy efficiency.
Strategically, Nvidia aims to democratize access to generative and agentic AI, making it available not only to large corporations but also to developers of small autonomous devices. The Jetson Orin Nano 2 occupies the lower segment of the lineup but offers computing capabilities that previously required much more powerful hardware.
My analysis: This move by Nvidia is a clear signal to the market that the era of "intelligent" robots operating autonomously without cloud dependency is approaching faster than many expect. Doubling performance while reducing power consumption is not just an incremental improvement but a catalyst for the mass adoption of physical AI in consumer and industrial sectors. I expect that by the end of 2027, we will witness a "ChatGPT moment" for robotics, as some industry experts predict.