Nvidia CES 2026: Rubin Platform Delivers 5x Blackwell, Mercedes Gets Autonomous AI
Jensen Huang unveils Rubin AI platform with 5x performance gains, announces Alpamayo autonomous driving partnership with Mercedes-Benz, and declares 'ChatGPT moment for physical AI' has arrived.
Jensen Huang delivered. The Nvidia CEO took the CES 2026 stage on Monday night and unveiled Rubin, the company's next-generation AI platform that promises 5x the performance of Blackwell while slashing token generation costs to one-tenth current levels.
That's the kind of performance leap that justifies Nvidia's $4.5 trillion valuation.
Rubin: Six Chips, One Platform
The Rubin platform represents Nvidia's most ambitious architecture yet—a six-chip system designed for extreme codesign between silicon and software. The centerpiece is the Vera Rubin GPU, which Huang positioned as the engine for next-generation AI data centers.
"Computing has been fundamentally reshaped as a result of accelerated computing, as a result of artificial intelligence," Huang told the packed Fontainebleau Las Vegas ballroom.
The performance claims are staggering: 5x more compute power than Blackwell, with dramatically lower inference costs. For AI companies already spending billions on training and deployment, that math changes everything.
Rubin enters production immediately, with availability to Nvidia partners in the second half of 2026. The timing is aggressive, but Nvidia has delivered on aggressive timelines before.
Alpamayo: The Autonomous Driving Bet
The bigger surprise was Alpamayo, an open portfolio of vision-language-action models, simulation blueprints, and datasets designed to enable Level 4 autonomous driving.
The first customer: Mercedes-Benz. The new Mercedes CLA will feature Alpamayo built on Nvidia's DRIVE full-stack autonomous vehicle platform, bringing AI-defined driving to US roads.
"There's no question in my mind now that this is going to be one of the largest robotics industries," Huang said. "Our vision is that someday every single car, every single truck will be autonomous."
The open model approach is strategically significant. Rather than keeping Alpamayo proprietary, Nvidia is making it available to the broader autonomous vehicle ecosystem. That's a platform play—own the infrastructure, let others build on top.
Physical AI Takes Center Stage
Huang declared the "ChatGPT moment for physical AI" is coming. It's a bold claim, but the CES presentation backed it up.
The keynote featured two BD-1 droids from Star Wars, Caterpillar's construction equipment, and Agibot's humanoid robots—all running on Nvidia technology. The company's open model portfolio now spans six domains: Clara for healthcare, Earth-2 for climate science, Nemotron for reasoning, Cosmos for robotics, GR00T for embodied intelligence, and Alpamayo for autonomous driving.
That's not a chip company talking. That's an AI infrastructure provider positioning itself across the full stack from silicon to simulation.
What Gamers Didn't Get
Notably absent from the nearly two-hour presentation: new GeForce RTX graphics cards. Nvidia used last year's CES to unveil the RTX 5000-series. This year, the company kept the focus entirely on enterprise AI and robotics.
For traders, that signals where Nvidia sees its growth. Consumer gaming remains a healthy business, but it's not the story Huang wanted to tell when he had the world's attention.
Stock Implications
Nvidia shares rose in premarket trading ahead of the keynote and should see continued momentum as investors digest the Rubin specs. The 5x performance claim alone, if validated, extends Nvidia's hardware lead over AMD and Intel.
The Alpamayo announcement opens a new addressable market. Morgan Stanley pegged the autonomous vehicle industry at $4 trillion by 2040. If Nvidia becomes the platform standard—as they did for AI training—the revenue opportunity is enormous.
Wedbush's Dan Ives called Nvidia, Microsoft, and Oracle "table-pounder" opportunities heading into the keynote. The Rubin reveal validates that thesis.
The Next Frontier
Huang framed 2026 as the year physical AI goes mainstream. That means robots that move through warehouses, vehicles that drive themselves, and industrial systems that optimize in real-time.
Nvidia is betting its next decade on that vision. Based on Monday night's presentation, they have the technology to back it up.
The preview expectations called for major announcements on physical AI. Huang delivered exactly that—and more.