NVIDIA’s CEO, Jensen Huang, has stirred excitement ahead of the upcoming GTC 2026 conference by announcing that a groundbreaking new chip—described as “never before seen in the world”—will make its debut during his keynote address in San Jose, California, on March 15. This strategic move is anticipated to further cement NVIDIA’s dominance in the AI infrastructure market. Industry watchers speculate that the new offering could be an advanced variant of the Rubin series or the next iteration in the Feynman series, potentially overcoming critical bottlenecks in latency and memory bandwidth.

On January 5 (local time), Huang showcased the next-generation Rubin platform at CES 2026, highlighting a fivefold performance improvement, full production readiness, and scheduled shipments for the latter half of the year. He also expressed optimism about the adoption of the H200 chip by Chinese buyers, stressing the importance of timely product launches to stay competitive in the Chinese market. Earlier, the Trump administration had allowed NVIDIA to export H200 chips to China, albeit with a condition of a 25% profit deduction. Meanwhile, Chinese companies are vigorously pursuing the development of indigenous AI chips to gain a foothold in the market.

In response to the White House AI chief’s assertion that China had rejected U.S. H200 chips, a spokesperson for the Chinese Foreign Ministry advised seeking clarification from Chinese authorities.