At COMPUTEX NVIDIA recently introduced JetPack 7.2 and NVIDIA NemoClaw support on NVIDIA Jetson. NemoClaw is NVIDIA’s agentic AI framework. The release will bring NemoClaw on the production-grade Jetson stack and transfer agentic AI from servers and workstations into the physical world of robotics, and other industrial applications. Alternatively, JetPack 7.2 adds agentic AI skills, Yocto project support, NVIDIA CUDA 13 on NVIDIA Jetson Orin, and Multi-Instance GPU (MIG) support on NVIDIA Jetson Thor.
Agentic AI is here, and Jetson’s programmability and high performance enable developers to instantly deploy physical AI agents in production at the edge,” said Deepu Talla, vice president of robotics and edge computing at NVIDIA. “With purpose-built skills for agentic development and workflows, developers can accelerate time to market, cut total cost of ownership and deploy at scale — all on a memory-optimized platform.
Three Layers Run Agentic AI
The release consists of three layers. The base, JetPack 7.2 consisting of the operating system (OS), compute, deterministic performance. A middle, new layer of agent skills, automating developer tasks. Additionally, a third layer with NemoClaw, is at the top.
The first or base layer of the release combines updates for Jetson and other applications:
JetPack 7.2 brings major upgrades to the Jetson software foundation. Yocto-based OS support gives industrial customers a leaner, more customizable Linux foundation — important for memory-bound deployments. CUDA 13 on Jetson Orin brings the latest compute stack to existing devices. MIG plus real-time kernel on Jetson Thor lets developers reserve dedicated GPU resources for deterministic workloads, like robot perception systems that can’t pause for unrelated AI inference. Jetson AGX Orin 32GB also gets a performance boost to 241 TOPS of AI compute, up 20% above its original spec.
Agent skills, the middle layer, accelerates the building of Jetson based-systems. These agent skills cover Linux customization, memory optimization, model benchmarking and similar developer tasks. Therefore the agent-deployable skills developed from NVIDIA documentation and design guides overall take less time to complete.
The top layer deploys NemoClaw to Jetson with a single command. The combination places agentic AI on a production-grade robotics and vision AI stack, reducing task automation for industrial systems to a minimum. NVIDIA Metropolis VSS blueprint skills introduces visual reasoning agents for developers to utilize that further enhance visual reasoning understanding.
Practical Applications
The Jetson platform is already in use in areas such as robotics, drones, industrial applications, humanoid systems, healthcare devices, and agriculture applications. For example, the humanoid robot, Solomon, uses NVIDIA NemoClaw to coordinate AI agents. Advantech is creating an agentic factory brain for its manufacturing facilities. The factory operations will be driven by AI-native operations incorporating NVIDIA NemoClaw, NVIDIA Nemotron 3 and NVIDIA Jetson Thor technologies.
Another Nemoclaw application is SandStar, which uses “NVIDIA Jetson Orin NX and NemoClaw to power AI vending machines and smart retail operations with AI vision, LLM-driven interaction, standard operating procedure monitoring and store optimization across 30+ countries.”