Nvidia moves NemoClaw agents onto Jetson edge devices
Nvidia announced JetPack 7.2 and NemoClaw support for Jetson, pushing agentic AI from data centers into robotics, inspection, smart retail, traffic systems, and factory automation.
Read more
Nvidia's Jetson update is a practical physical-AI story that was easy to miss under the bigger RTX Spark headlines. At Computex, Nvidia announced JetPack 7.2 and NemoClaw support for Jetson, giving developers a production-grade path to run agentic workflows on edge systems rather than only servers or workstations. JetPack 7.2 adds Yocto project support, CUDA 13 on Jetson Orin, Multi-Instance GPU support on Jetson Thor, and a 20% performance bump for the Jetson AGX Orin 32GB module to 241 TOPS. Nvidia says NemoClaw can now deploy to Jetson with a single command, enabling agents for robotics, industrial inspection, smart retail, traffic management, humanoids, and factory operations. This matters because many useful AI agents need perception, latency, and local control at the edge, not just cloud reasoning.
Key details: Nvidia, June 1, 2026, JetPack 7.2, NemoClaw on Jetson, CUDA 13 on Jetson Orin, Yocto project support, MIG support on Jetson Thor, Jetson AGX Orin 32GB increased to 241 TOPS.
Continue swiping for more AI Brief stories.