Nvidia CEO Huang Announces Full Production of Next-Generation Chips, Boosting AI Capabilities

Nvidia’s CEO, Jensen Huang, revealed at the Consumer Electronics Show (CES) in Las Vegas that the company’s next generation of chips is now in full production. These new chips are expected to deliver five times the AI computing power of their predecessors, significantly enhancing chatbots and other AI applications.

Huang showcased the Vera Rubin platform, which consists of six Nvidia chips, including 72 graphics processing units and 36 central processors. This platform is set to debut later this year, and Nvidia executives confirmed that AI firms are already testing it in their labs. The chips can be grouped into "pods" containing over 1,000 Rubin chips, promising to improve the efficiency of generating AI "tokens" by a factor of 10.
 

Nvidia’s new chips also leverage proprietary data, contributing to their significant performance boost despite having only 1.6 times more transistors than previous models. While Nvidia currently dominates the AI model training market, it faces increased competition from rivals like Advanced Micro Devices (AMD) and customers such as Google, which are developing their own AI chips.

Huang emphasized that the new chips include “context memory storage” to help AI chatbots deliver quicker responses in lengthy conversations. Additionally, Nvidia introduced new networking switches featuring co-packaged optics, designed to improve machine-to-machine connections, competing with offerings from Broadcom and Cisco Systems.

Nvidia announced that its new Vera Rubin systems will be deployed by CoreWeave and expects major players like Microsoft, Amazon, Oracle, and Alphabet to adopt them. Huang also discussed software advancements for self-driving cars, such as Alpamayo, which will be open-sourced to foster transparency in the decision-making process of autonomous vehicles.

In another strategic move, Nvidia acquired talent and chip technology from the startup Groq, which has been helping Google design its own AI chips. This acquisition highlights the growing rivalry between Nvidia and Google, as Google continues to challenge Nvidia's AI dominance.

Huang further noted the strong demand for Nvidia’s H200 chips in China, despite the geopolitical concerns over the chip's distribution to the region. Nvidia has applied for licenses to ship the chips to China, awaiting government approvals.