Nvidia on Sunday unveiled its next generation of artificial intelligence chips to succeed the previous model, which was announced just months earlier in March.
Nvidia CEO Jensen Huang announced the new AI chip architecture, dubbed “Rubin,” ahead of the COMPUTEX tech conference in Taipei.
Rubin comes months after the March announcement of the upcoming “Blackwell” model, which is still in production and expected to ship to customers later in 2024.
Huang’s announcement of Rubin appears to quicken the company’s already-accelerated pace of AI chip advancement.
Nvidia has pledged to release new AI chip models on a “one-year rhythm,” as Huang put it on Sunday. The company had previously been operating on a slower two-year update timeline for chips.
The turnaround from Blackwell to Rubin was a matter of less than three months, underscoring the competitive frenzy in the AI chip market and Nvidia’s sprint to preserve its dominant spot.
AMD and Intel are two major competitors working to catch up, though their gross margins trailed Nvidia’s in the most recent fiscal quarter. Companies like Microsoft, Google and Amazon are also vying for Nvidia’s top spot, even as they are simultaneously some of Nvidia’s biggest patrons. A flurry of startups are also working to enter the space.
“Today, we’re at the cusp of a major shift in computing,” Huang said Sunday. “With our innovations in AI and accelerated computing, we’re pushing the boundaries of what’s possible and driving the next wave of technological advancement.”
The Rubin chip platform will have new GPUs, the crucial graphic processing technology that helps train and launch AI systems. It will come with other new features like a central processor called “Vera,” though the Sunday announcement did not provide many details.
Shares of Nvidia were relatively flat at Friday’s market close with shares trading at $1,096.