Nvidia stops selling H200 chips in China, acting CEO Kress says

Nvidia will no longer sell its H200 chips in China, acting CEO Colette Kress said on Tuesday. That ends the company's participation in a controversial loophole in the U.S. ban on selling advanced chips to the country. China's access to leading AI accelerators was a major topic on Nvidia's first earnings call since CEO and co-founder Jensen Huang stepped down last week. The same day, Nvidia announced a board-level committee to oversee national security policies, export control compliance and global trade risks.
The H200 deal was brokered by Huang and President Donald Trump. It required the chips to be shipped through the U.S., which let the government collect a 25% tariff. China hawks, who wanted to deny Chinese labs computing power and protect the U.S. lead in AI, were angered by the pathway. Huang had long argued that it was more important for the U.S. to compete in China. In April he told podcaster Dwarkesh Patel that "to concede that market for the United States technology industry is a disservice to our country."
TechCrunch reports that Huang's ouster, in part over an affair with the CEO of a company under investigation for smuggling chips into China, has coincided with new rules issued Monday. They bar advanced U.S. chips from foreign data centers that serve Chinese customers remotely. Investment bank Jefferies forecast that the rules might trim between $44 billion and $62 billion from Nvidia's data center revenue next year. Kress, referring to the new export controls, said the backlog of demand for advanced chips like Vera Rubin means the rules would have no impact on the company's top-line revenue.
In the long term, TechCrunch writes, Nvidia may face the loss of potential customers in Singapore, Malaysia, Thailand and Indonesia. Critics of Nvidia's dependence on China claim that as much as 20% of its revenue comes from demand linked to the country. Annual revenue at its Singapore subsidiary hit $104 billion in 2025, before Nvidia stopped disclosing those figures. The article says that is clearly going to change.
The H200 pullback may matter less in practice than it sounds. Even before Nvidia decided to stop shipping, China had largely refused to accept delivery of the chips. According to Jacob Feldgoise, a researcher at Georgetown's Center for Security and Emerging Technology, the Chinese government would prefer to generate demand for its domestic chip industry, even if major AI compute users want the still-superior Nvidia equipment.
Meanwhile, Chinese frontier labs are finding ways to reach Nvidia chips elsewhere. All of China's leading AI models show evidence of having been trained on Nvidia Grace Blackwell chips, which are prohibited from export to China. That training may have taken place at data centers outside China, but sales to those facilities will be barred under the new rules. Export control attorney Dan Fisher-Owens told TechCrunch that large data centers suited to training AI models will have to certify they aren't serving any Chinese customers if they want to buy Nvidia chips. If they can't, the rules may push them toward other chip options and further eat into Nvidia's market share.
Key facts
- Acting CEO Colette Kress said Nvidia will no longer sell H200 chips in China, on the first earnings call since Jensen Huang stepped down last week.
- New U.S. rules issued Monday bar advanced U.S. chips from foreign data centers that serve Chinese customers remotely; Jefferies forecast a $44 billion to $62 billion hit to Nvidia's data center revenue next year.
- Kress said the backlog of demand for chips like Vera Rubin means the rules would have no impact on top-line revenue.
- The H200 arrangement, brokered by Huang and President Trump, shipped chips through the U.S. so the government could collect a 25% tariff; China had already largely refused delivery.
- Nvidia also announced a board-level committee on national security policies, export control compliance and global trade risks.
Why it matters
The H200 sales were the one sanctioned route for Nvidia's advanced chips into China, a deal brokered by Huang and Trump with a 25% tariff for the U.S. government. Ending it closes a pathway that China hawks had opposed, and it lands the same day as a new board-level committee on national security and export compliance and a week after Huang's departure. Together with the new rules on remote access from foreign data centers, it marks a clear shift in how Nvidia handles China.
Who it affects
Nvidia and its investors, first of all: Jefferies sees a possible $44 billion to $62 billion cut to next year's data center revenue, while Kress says there will be no top-line impact. Large foreign data centers suited to AI training will have to certify they aren't serving Chinese customers, and Nvidia may lose potential customers in Singapore, Malaysia, Thailand and Indonesia. Chinese AI labs are affected too, since their leading models show evidence of training on Grace Blackwell chips that may have been run in data centers outside China.
How to use it
There is nothing to adopt here, but there are things to watch. Operators of large AI training data centers outside China will have to certify that they aren't serving any Chinese customers, according to export control attorney Dan Fisher-Owens. Those who can't may be pushed toward other chip options. Investors can track whether Kress's no-impact claim holds against Jefferies' forecast.
How solid is it
The core fact, that Nvidia will stop selling H200s in China, comes from acting CEO Colette Kress on the earnings call, as reported by TechCrunch. The revenue figures are not equally firm. The $44 billion to $62 billion range is a Jefferies forecast, not a Nvidia number, and Kress disputes any top-line impact. The 20% China-linked revenue share and the $104 billion Singapore figure are claims by critics of Nvidia's China dependence, and Nvidia stopped disclosing the Singapore figures. The source does not state the reason Nvidia chose to stop H200 sales beyond the context given.
Risks and caveats
The source does not give the date from which the new export rules take effect. The account of Huang's ouster, in part over an affair with the CEO of a company under investigation for smuggling chips into China, is TechCrunch's own wording, and it names neither the person nor the company. The loss of customers in Southeast Asia is described only as something that may happen in the long term. China had already largely refused H200 deliveries, so the practical effect of the pullback may be limited. The rules may also push data centers that can't certify toward other chip options and further eat into Nvidia's market share.
“to concede that market for the United States technology industry is a disservice to our country.”
— Jensen Huang, in April, to podcaster Dwarkesh Patel