China plans to allow top artificial intelligence companies to buy a limited amount of H200 chips from Nvidia, a sign the country is easing restrictions on the coveted US technology, according to the Information.
Chinese officials have informed companies such as Alibaba Group, ByteDance and DeepSeek that they will have permission to purchase some of the processors, which are used to develop AI models, the news site said July 8.
The companies need to say how many chips they need – and why – to get approval, according to the Information.
The chips have emerged as a geopolitical flashpoint for the world’s two largest economies.
Though US President Donald Trump granted Nvidia permission to sell the processors to China in December, Chinese authorities have been slow to allow the technology into the country.
Nvidia shares climbed 3.7 per cent to US$204.12 on July 8, part of a broader chip rally. It marked the stock’s biggest one-day increase in more than a month.
Though Nvidia is now up 9.4 per cent in 2026, that pales in comparison with the gains of most major semiconductor peers.
In a statement, Chinese embassy spokesperson Liu Chang said that the country’s position on US chip exports has been consistent.
“We advocate that China and the United States achieve mutual benefit and win-win outcomes through cooperation, and we oppose the politicisation, instrumentalisation, and weaponisation of technological and economic issues,” he said.
“We stand ready to work with all parties to jointly safeguard the stability of global industrial and supply chains.”
A representative for Santa Clara, California-based Nvidia did not respond to a request for comment.
The H200 import snags hinged in part on Chinese concerns that a flood of American-designed AI processors would hinder the government’s long-held goal of developing an indigenous chip industry.
Beijing has also cited fears that allowing American-made chips into the local market could pose a cybersecurity risk.
But soaring demand for AI processors helped spur Chinese authorities to allow some H200 purchases, according to the Information, which cited two unidentified people familiar with the matter.
Labs in China have been struggling, like their US counterparts, to find enough computing power to meet their growing needs to train and run new models.
Nvidia’s H200 hails from the company’s Hopper line of processors used for training and running AI models. It was the most powerful AI chip on the market until the introduction of the company’s next-generation Blackwell models in late 2024.
Nvidia is now working to release an even more powerful line-up called Rubin in the second half of 2026. Washington has continued to restrict sales of more capable AI chips to China on national security grounds.
The Chinese government, meanwhile, is still figuring out how many H200 chips to allow the companies to buy, but the total may end up being less than 200,000, the Information reported.
When it comes to AI infrastructure, that is not a large number. A single data centre can use more than 400,000 Blackwell processors.
In May, Nvidia Chief Financial Officer Colette Kress said the H200 chips had yet to generate any revenue for the company in China.
“We are uncertain whether any imports will be allowed into the country,” she said.
As a result, Nvidia did not include any China revenue from that market in its outlook, Kress said. BLOOMBERG
