Details
- NVIDIA announced Jetson Orin Nano 2, an entry-level robotics computer delivering 2x the inference performance of Jetson Orin Nano Super in the same compact form factor.
- The module features 78 TOPS of AI compute, 8GB memory, and an 8-core Arm CPU, targeting robots, delivery and inspection drones, and vision AI systems.
- It improves energy efficiency, consuming 40% less power in 15W mode to deliver the same performance as its predecessor, and runs frontier-class LLMs and VLMs such as NVIDIA Cosmos, Nemotron, Gemma 4, and Qwen 3.
- More than 3 million developers use NVIDIA's robotics stack, with early adopters including Cognex, Doosan Bobcat, Matic, and Wing, alongside a broad hardware ecosystem building carrier boards and reference systems.
- The Jetson Orin Nano 2 module and developer kit are expected to be available in the first half of 2027, extending NVIDIA's push to bring generative and real-time reasoning AI to physical edge devices.
Impact
By doubling inference performance and improving power efficiency in an entry-level module, Jetson Orin Nano 2 strengthens NVIDIA’s position in edge robotics and physical AI platforms. Its ability to run modern LLMs and VLMs on compact devices is likely to accelerate adoption of autonomous drones and home robots, shaping R&D priorities toward edge-first generative AI and real-time perception over the next 12–24 months.
