NVIDIA's Jetson update is the boring kind of important
JetPack 7.2 won't make headlines, but it's the kind of infrastructure work that actually moves industrial robotics forward.
画像クレジット: Image via source article. Used under fair use for news commentary. · source
NVIDIA just dropped JetPack 7.2 with agentic AI support for Jetson at COMPUTEX, and I'll be honest, my first reaction was "finally."
Look, here's the thing. When I was at Kuka, we spent an embarrassing amount of engineering hours wrestling with edge compute limitations. You'd have a perfectly good vision system, a solid motion controller, and then you'd hit a wall because the onboard compute couldn't handle the inference load without cooking itself or tanking cycle times. We used to joke that half our job was thermal management.
The actual news
The JetPack 7.2 release brings what NVIDIA's calling "agentic AI skills" to Jetson hardware, along with CUDA 13 support for the Orin lineup. There's also Multi-Instance GPU support on the Jetson AGX Orin, which is genuinely useful if you're running multiple inference tasks simultaneously (and in a real production environment, you always are).
NVIDIA's blog also mentions NemoClaw support, which appears to be their framework for getting language model reasoning into physical systems. The details on exactly how this works in practice are, well, sparse. It remains unclear how much of this is production-ready versus "works in the demo booth."
Separately, The Robot Report notes that NVIDIA also released open-source physical AI agent skills and an Isaac GR00T humanoid reference robot. The humanoid stuff is flashy, but I'm more interested in the tooling underneath.
関連記事
More in Industrial
New research tackles the speed problem that's kept diffusion planners in the lab. About time.
Robert "Bob" Macintosh · 2 hours ago · 3 min
A batch of new research papers show that vision-language-action models break down in predictable, clusterable ways. Anyone who's deployed industrial robots could've told you this.
Robert "Bob" Macintosh · 2 hours ago · 4 min
New research shows AI-powered robots can fail in ways we can't see coming, and the industry doesn't have a good answer yet.
Robert "Bob" Macintosh · 2 hours ago · 4 min
A batch of new research papers suggests we're finally cracking the code on getting robot policies out of simulation and onto real hardware without everything falling apart.