What to make of the news from Nvidia at CES? Because everyone will make something of it, right – this being the world’s most valuable company, and its ghastly-jacketed owner-boss, Jenson Huang, being the big-tech oracle for the whole global economy. Well, given my (historical) beat at RCR, I have a particular view: that it is all about the difficult Industry 4.0 market (even if Nvidia avoids the label), and that Nvidia is late to the game; but also that the Santa Clara firm’s strategic tardiness and teflon magnetism will either drive new interest in Industry 4.0, or kill it for big tech once and for all. 

A quick recap, from yesterday in Vegas: Nvidia made it clear at CES where it sees the future – in enterprise AI, industrial robotics, and physical intelligence. Plus driverless cars, as a sexier articulation of the same. But this is all Industry 4.0 stuff, as much about IoT sensing as it is about AI sense-making. In Vegas, Huang showcased Rubin, a six‑chip rack-scale AI platform, alongside frameworks for autonomous systems, edge intelligence, and industrial deployments. These ‘things’, a way-away from the firm’s traditional heartlands in consumer GPUs and gaming (maybe why its stock barely budged after the keynote), have been in development for years. 

Which is not to say Nvidia has not been dropping this stuff into its roadmap already; nor that the whole mega-bucks bricks-and-mortar AI build-out, to house its AI-geared GPUs, has not (really) been in service of incoming enterprise workloads. But at CES, this was the story for Nvidia, even if it was translated in the popular press as AI for cars and taxis (foregrounding its chain-of-thought Alpamayo portfolio and work with Mercedes-Benz). But physical AI, as promoted by automation vendors for ages, and as presented by Huang on Monday, is about robotics to drive physical automation and industrial control.

And the ones in proper need of such pyrotechnics are industrial enterprises, busy rolling out their own private AI data centres and data networks. Which presents an interesting dynamic – the darling of the tech world at the sharp end of Industry 4.0, which big tech has been quitting the Industry 4.0 scene in droves over the last 12 months. Because it’s too much like hard work. Because it requires systems integration. Because it is a team sport and speculative pursuit, and not a big-box sale to big-ticket customers. They miss the point, and will miss the boat. Contrary to big-vendor balance sheets, industrial AI and automation are progressing steadily.

Predictive maintenance, robotics systems, edge applications are delivering measurable ROI in just about every hard-nosed industrial sector, and will eventually jump with higher-grade AI reasoning. It will take time, but it sounds like Nvidia also knows it’s where AI will pay.