China’s open source AI is ‘a catalyst for global progress’, Nvidia’s CEO Jensen Huang says

Nvidia’s co-founder and CEO Jensen Huang lauded China’s progress in open source artificial intelligence and pledged to work with Chinese companies, as the chip designer driving the growth of AI resumes shipping its sought-after chips to the country.

Large language models (LLM) developed by Chinese companies, including DeepSeek, Alibaba Group Holding, Tencent Holdings, MiniMax and Baidu, were “world class”, “developed here and shared openly”, and had spurred AI developments worldwide, Huang said.

Clad in a dark Tang-style suit instead of his signature black leather jacket, Huang was speaking as the guest of honour at the expo on his third trip to the Chinese capital city this year. His California-based chip-design firm, established in 1993, is poised to resume shipments of its made-for-China H20 chips to China.

The US government “assured Nvidia that licences will be granted” for exporting the H20 chip, a made-for-China product that was less powerful than Nvidia’s gold-standard acceleration chip, according to a Tuesday statement by the company.