NVIDIA Unveils ‘Vera Rubin’ AI Chip, Promising 5x Reasoning boost
Table of Contents
NVIDIA CEO Jensen Huang revealed teh company’s next-generation artificial intelligence (AI) chip platform, ‘Vera Rubin,’ on Thursday, January 5th, signaling a significant leap forward in computing power and accelerating the race for AI dominance. The new chip, slated for full-scale production in the first half of 2026, boasts a 5x increase in reasoning capability and a 3.5x improvement in learning capacity compared to NVIDIA’s current flagship, the Blackwell Ultra.
A New Era of AI Computing
During a keynote address at the Fontainebleau hotel in Las Vegas, ahead of the CES 2026 technology exhibition, Huang showcased the palm-sized Vera Rubin chip.He stated the chip possesses the capacity to “remember all your conversations and research for the rest of your life,” highlighting its potential for personalized and advanced AI applications. The accelerated production timeline, moved from the latter half of 2026 to the first, underscores NVIDIA’s commitment to rapid innovation.
Huang, who previously championed the concept of “physical artificial intelligence” – AI that exists and interacts in the real world – predicted a continued bright future for the field. “Artificial intelligence is not just an request, it is your tool,” he explained, emphasizing the accessibility of AI development. “Everyone can develop artificial intelligence that suits them, and companies are also partners with us.”
From Robots to Autonomous Driving
The presentation differed from last year’s event, where Huang took the stage with 14 robots. This year, he appeared alone in a black leather jacket, focusing on the unveiling of NVIDIA’s latest AI concepts.A brief technical hiccup during the presentation, when the data screen temporarily went dark, prompted a lighthearted response from Huang, who quipped, “It looks like someone hit the jackpot (in Las Vegas).” The speech concluded with two small robots following him off stage. The event drew a large crowd, filling the 3,800-capacity Fontainebleau Theater.
Huang emphasized the transformative potential of Vera Rubin, stating it “can reinvent the entire structure of computing.” The chip is a combination of the central processing unit (CPU) ‘Vera,’ functioning as the brain, and the graphics processing unit (GPU) ‘Rubin.’
Intensified Competition in the AI Landscape
the official announcement of Vera Rubin’s mass production is expected to intensify competition among global technology companies vying for AI leadership. Rivals such as Qualcomm, developer of the tensor processing unit (TPU) used in Google’s ‘Gemini 3.0,’ and AMD, are now under pressure to accelerate their own AI chip development.
Huang also highlighted the importance of open source development in the AI market,citing China’s ‘Deepseek’ as an example. he noted that while open-source AI currently lags behind cutting-edge proprietary models by approximately six months in performance, its collaborative nature fosters rapid innovation. “As anyone can participate in development, new models are emerging and the speed at which thay are becoming smarter is accelerating,” he said. NVIDIA is furthering this approach by developing open models through its DGX supercomputer, aiming to facilitate collaboration and empower users to create “your own artificial intelligence.” A video exhibition showcased a small robot engaged in conversation,illustrating the potential of these advancements.
Alphamayo and the Future of Autonomous vehicles
NVIDIA also introduced ‘Alphamayo,’ its autonomous driving platform, representing a key example of “physical artificial intelligence.” Huang predicted that “One day, all vehicles in the world will be able to drive autonomously.” The platform will initially be integrated into vehicles manufactured by German automaker Mercedes-Benz, with plans for expansion into the United States market.
