OpenAI and AWS among early adopters
Nvidia, currently the world’s most valuable corporation, has officially launched its next-generation AI computing architecture, Rubin, at the Consumer Electronics Show (CES). CEO Jensen Huang described Rubin as the new state of the art in AI hardware, according to TechCrunch.
Nvidia confirmed that the Rubin architecture has already entered full-scale production and will replace the current Blackwell generation, continuing the company’s rapid hardware development cycle following Hopper and Lovelace. Production volumes are expected to ramp up significantly in the second half of the year.
Read also: A critical moment for AI self-training is getting close according to Anthropic’s chief scientist
“Vera Rubin is designed to address the fundamental challenge we face: the exponential growth in AI compute demand. Today, I can tell you that Vera Rubin is in full production,” Huang said.
Major AI players have already committed to the new platform, including Amazon Web Services (AWS), OpenAI, and Anthropic. Rubin chips are expected to deliver 3.5x faster AI training speeds and up to 8x better energy efficiency compared to the previous generation.
Looking ahead, Huang estimates that global investments in AI infrastructure could reach $3–4 trillion over the next five years, underscoring the scale of demand for advanced AI hardware.
Photo: The Verge
