Nvidia is no longer shipping the AI chips that Washington had licensed it to sell to China. Interim chief executive Colette Kress confirmed the halt on the earnings call on Tuesday, November 17, in one plain sentence, and then offered no reason for it at all.
“Our guidance does not have any licensed Hopper data center products included, and we are no longer shipping licensed Hopper data center products to China,” Kress said, as quoted in the call transcript and in Reuters reporting.
The timing makes the sentence land harder. In the quarter that ended on October 25, Nvidia sold 2.4 billion dollars worth of those same licensed Hopper products to China. Now the next quarter forecast assumes no data center compute revenue from China at all.
The announcement
The products at issue belong to the Hopper family. That family includes the H200, a powerful processor used to train and run advanced AI models. Hopper products are data center products. Neither the earnings call nor the earnings release mentioned gaming graphics cards, and readers should not read this halt as a halt on every Nvidia product sold in China.
Kress spoke on the first earnings call since Jensen Huang resigned. Huang stepped down as president, chief executive and board member on November 12, and the board named Kress, the longtime finance chief, as interim chief executive. Our earlier feature on that change is at Huang Exit Asks Who Leads AI Now.
The wider numbers show why a 2.4 billion dollar channel can close without denting the headline story. Nvidia reported total revenue of 115.8 billion dollars for the third quarter of fiscal 2027, up 103 percent from a year earlier, with data center revenue of 107.6 billion dollars. The licensed Hopper sales to China were about 2.1 percent of total revenue. For the fourth quarter, Nvidia guided revenue to 130 billion dollars, plus or minus 3 percent, for a quarter that includes an extra week, and said outright that it is not assuming any data center compute revenue from China.
Important details
The licensed route was never a simple sale. Washington first restricted advanced AI chip exports to China in 2022 and tightened those controls over time. In January 2026 the United States eased its review for the H200, allowing export applications to be looked at one by one under security conditions. Washington later approved H200 sales to ten Chinese companies, including Alibaba, Tencent and ByteDance, the owner of TikTok, according to earlier Reuters reporting.
Approval in Washington did not mean approval in Beijing. Nvidia had already told investors that turning licences into sales was hard. Its second quarter filing said licences granted from February 2026 allowed small shipments of H200 products to specific customers in China, and that Chinese government restrictions then stopped Nvidia from selling all of the products those approvals covered. The company also recorded a 400 million dollar charge linked to excess H200 inventory and purchase obligations in the first half of fiscal 2027.
The route itself carried a cost. Eligible chips had to pass inspection in the United States before delivery, and that step made shipments subject to a 25 percent tariff on import. Reuters also reported earlier in November that about 600,000 H200 chips were sitting in Hong Kong while Beijing held back approval for large imports to the mainland. Beijing has restricted H200 imports as it pushes to cut reliance on United States technology, even as Chinese firms face strong demand for AI compute.
| Detail | What was reported |
|---|---|
| Products halted | Licensed Hopper data center products, including the H200 |
| Sales in the quarter ended October 25 | 2.4 billion dollars to China |
| Total Nvidia revenue that quarter | 115.8 billion dollars |
| Share of total revenue | About 2.1 percent |
| Fourth quarter outlook | 130 billion dollars, plus or minus 3 percent |
| China in that outlook | No data center compute revenue assumed |
| Reason given for the halt | None |
| Restart date given | None |

- 2022The United States introduces export controls on advanced AI chips for China
- January 2026Washington eases H200 licensing review to a case by case basis under security conditions
- February 2026Licences allow small H200 shipments to specific customers in China, per Nvidia filings
- November 5, 2026Reuters reports about 600,000 H200 chips held in Hong Kong awaiting import approval
- November 12, 2026Jensen Huang resigns and Colette Kress becomes interim chief executive
- November 16, 2026The Commerce Department issues new rules on advanced chips in overseas data centers that serve Chinese customers remotely
- November 17, 2026Kress confirms licensed Hopper shipments to China have stopped
A squeeze from both sides, and a new rule beside it
It would be tidy to pin the halt on one cause. Nvidia has not done that, and this report will not do it for them. Kress gave no explanation, and no filing cited in the reporting below ties the decision to a single order, a single customer, or a single rule.
What is documented is pressure from several directions at once. Chinese import limits left approved chips waiting in Hong Kong. The tariff and inspection route raised costs that Nvidia said it had been unable to pass on in full. Then, on November 16, the day before the earnings call, the Commerce Department Bureau of Industry and Security issued new rules aimed at advanced United States chips in foreign data centers that provide remote computing capacity to Chinese customers. Those rules also require export licences for chip shipments to several other countries named in the coverage. Nvidia has not said that this rule caused the halt, and the timing alone does not prove that it did.
On the same day as the call, Nvidia also created a board committee to oversee national security policy, export compliance and global trade risks. That move does not explain the halt either. It does show a company putting export decisions under closer board level watch at the exact moment its most sensitive sales route closes.
Chinese buyers, meanwhile, have been building paths that do not run through Nvidia at all. Our report on Meituan LongCat trained without Nvidia chips covered one large model effort built on domestic hardware, and Google TPU chips show how hard rivals outside China are pushing their own accelerators. A halt that lasts will feed both trends, even if Nvidia never intended it to.

What is next
The confirmed steps are short. Shipments of licensed Hopper data center products to China have stopped. The fourth quarter outlook includes none of those sales and no data center compute revenue from China. No restart date has been given.
The open questions are the ones buyers and labs now have to plan around. Is the halt permanent, or is it a pause while licences, import approvals and the new remote computing rules settle? Did customers cancel, did Beijing block, or did Nvidia choose to walk away from a small and costly channel while demand elsewhere is far larger than supply? Kress said the backlog for newer products is deep, a point other coverage of the call also noted, but she did not use that point to explain the China decision.
For Chinese labs, the practical effect arrives first, whatever the cause turns out to be. Approved H200 supply was already slow, costly and uncertain. It is now, by the words of the person running Nvidia, not shipping at all. That leaves stock on hand, domestic chips, and remote compute routes that regulators moved to close the day before she spoke.
Facts checked November 19, 2026, against Nvidia earnings reporting and the sources below. Figures for revenue, guidance and licensed Hopper sales are as reported by Nvidia and cited in that reporting. Claims about import limits, approved buyers and chips held in Hong Kong are attributed to Reuters reporting where they appear, and Nvidia has given no reason for the halt.
Sources
- Reuters, Nvidia stops shipping US approved AI chips to China, November 18, 2026: https://935thelloyd.com/2026/11/18/nvidia-stops-shipping-us-approved-ai-chips-to-china/
- Gulf Business, Nvidia stops shipping US approved AI chips to China, November 19, 2026: https://gulfbusiness.com/en/2026/news/nvidia-stops-shipping-us-approved-ai-chips-to-china/
- Nvidia, financial results for third quarter fiscal 2027, November 17, 2026, as reported in earnings coverage: https://nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-third-quarter-fiscal-2027
- MarketBeat, Nvidia Q3 2027 earnings results and call transcript, November 17, 2026: https://www.marketbeat.com/earnings/reports/2026-11-17-nvidia-co-stock/
- Hero and body images: Dirty Static illustrations
END OF TRANSMISSION
Stay curious. Question everything.


