Jensen, CEO of NVIDIA Huang says the company’s next-generation AI superchip platform, Vera Rubin, should start arriving to customers later this year. “Today I can tell you that Vera Rubin is in full production,” Huang said during a press event Monday at the annual CES technology show in Las Vegas.
Rubin will reduce the cost of run AI models about a tenth of Nvidia’s current leading chip system, Black wellthe company told analysts and reporters in a call Sunday. Nvidia also said Rubin could drive some large models using about a quarter of the number of chips required by Blackwell. Taken together, these gains could make advanced AI systems much cheaper to operate and make it harder for Nvidia’s customers to justify abandoning its hardware.
Nvidia said on the call that two of its existing partners, Microsoft and CoreWeave, would be among the first companies to begin offering services powered by Rubin chips later this year. Two large AI data centers that Microsoft is currently building in Georgia and Wisconsin will eventually include thousands of Rubin chips, Nvidia added. Some Nvidia partners have begun running their next-generation AI models on early Rubin systems, the company said.
The semiconductor giant also said it was working with Red Hat, which makes open-source enterprise software for banks, automakers, airlines and government agencies, to offer more products running on the new Rubin chip system.
Nvidia’s latest chip platform is named Vera Rubinan American astronomer who reshaped the way scientists understand the properties of galaxies. The system features six different chips, including the Rubin GPU and a Vera CPU, both built using Taiwan Semiconductor Manufacturing Company’s 3-nanometer manufacturing process and the most advanced bandwidth memory technology available. Nvidia’s sixth-generation interconnect and switching technologies connect individual chips together.
Every element of this chip system is “completely revolutionary and the best of its kind,” Huang proclaimed during the company’s CES press conference.
Nvidia has been developing the Rubin system for years, and Huang first announced that the chips would arrive during a keynote speech in 2024. Last year, the company said that systems built on Rubin would begin arriving in the second half of 2026.
It’s unclear exactly what Nvidia means by saying Vera Rubin is in “full production.” Typically, production of such advanced chips, which Nvidia is building with longtime partner TSMC, starts at low volume while the chips go through testing and validation and ramp up at a later stage.
“This CES announcement about Rubin is meant to tell investors, ‘We’re on the right track,'” said Austin Lyons, an analyst at Creative Strategists and author of the semiconductor industry newsletter. Chipstrat. There were rumors on Wall Street that the Rubin GPU was behind schedule, Lyons says, so Nvidia is now fighting back by saying it has passed key development and testing milestones, and he is confident that Rubin is still on track to begin ramping up production in the second half of 2026.
In 2024, Nvidia had to delay delivery of its newest Blackwell chips due to a design flaw that caused them to overheat when connected together in server racks. Shipments for Blackwell were back on schedule by mid-2025.
As the AI industry rapidly grows, software companies and cloud service providers have faced fierce competition for access to Nvidia’s latest GPUs. Demand will likely be just as high for Rubin. But some companies are also hedging their bets by investing in their own custom chip designs. OpenAI, for example, said it is working with Broadcom to create custom silicon for its next generation of AI models. These partnerships highlight a longer-term risk for Nvidia: Customers who design their own chips can gain a level of control over their hardware that the company doesn’t offer.
But Lyons says today’s announcements demonstrate how Nvidia is evolving beyond just offering GPUs to become a “complete AI system architect, spanning compute, networking, memory hierarchy, storage and software orchestration.” Even as hyperscalers invest in custom silicon, he adds, Nvidia’s tightly integrated platform “is becoming increasingly difficult to replace.”
