Jensen Huang says Chinese AI models perform excellently, open source will drive efficiency gains

·by Henderson·AI
Jensen Huang says Chinese AI models perform excellently, open source will drive efficiency gains

NVIDIA CEO Jensen Huang said at a forum hosted by DBS Bank that Chinese AI models including DeepSeek, Kimi and GLM have performed excellently. He emphasized he is a staunch supporter of open source, believing it is the best way to drive efficiency gains. He added that the market will decide which models are best suited for different application scenarios.

Jensen Huang predicts quantum computing will go mainstream within the next three to five years

Huang also mentioned that quantum computing could enter mainstream applications within the next three to five years, and that software will play an important role in driving efficient use of related infrastructure. Notably, this is not the first time Huang has made such remarks. In July this year, in an interview with Axios, he discussed China's AI development, stating that China's open-source AI models are very excellent, and the U.S. should not fear China's open-source AI models but should instead worry about voices calling for banning these models.

The report notes that Huang dismissed the notion that AI competition is a race with a finish line. He said we will use AI forever, and in this field the U.S. will be here for the long term, and China will be here for the long term. Huang's remarks came as Chinese companies released the next-generation model Kimi K3, which shook global markets again following DeepSeek.

In Huang's view, the market first misunderstood DeepSeek's impact, and Wall Street this time misunderstood Kimi's impact. He believes cheaper, more open models will enable more people and enterprises to benefit from AI, thereby increasing demand for NVIDIA's chips, data centers and computing power. He noted that free AI should bring significant benefits to hardware, chips and data centers alike.

H
About the author
Henderson