Deploy Agentic-Ready AI at the Edge with Memory Efficiency in NVIDIA JetPack 7.2
NVIDIA Technical, Monday, June 1st, 2026
NVIDIA JetPack 7.2 enables one-command deployment of agentic AI applications on Jetson devices with enhanced memory efficiency.
NVIDIA JetPack 7.2 introduces one-command deployment of NemoClaw for secure agentic AI applications on Jetson devices, supporting robotics, industrial automation, and edge AI systems. The release includes Jetson agent skills that automate development tasks such as Linux customization, memory optimization, and model benchmarking, reducing complexity and accelerating time to market.
JetPack 7.2 also brings Multi-Instance GPU (MIG) support to Jetson Thor for deterministic multiworkload execution, critical for real-time applications, plus official Yocto Project support for custom Linux distributions and Super Mode for Jetson AGX Orin 32 GB. These updates help developers maximize existing Jetson hardware value while reducing total cost of ownership.