Key Points
- Nvidia now requires full upfront payment from Chinese buyers for its H200 AI chips to manage export risk.
- The payment rule applies to orders that cannot be canceled, changed, or refunded once placed.
- China has also asked some tech firms to pause H200 orders as regulators decide import terms.
Nvidia has introduced strict new terms for Chinese customers seeking its H200 artificial intelligence processors, sources say. The U.S. chipmaker now insists buyers pay the full cost of orders before shipment. This change reflects deep uncertainty over whether Beijing will formally approve the advanced chips’ import.
Under the updated policy, Chinese tech firms must commit capital upfront for H200 orders with no option to cancel, seek refunds, or adjust configurations after placement. Nvidia is negotiating these terms to hedge against regulatory or diplomatic shifts that could disrupt delivery.
In some limited cases, buyers may offer commercial insurance or provide asset collateral instead of cash to secure orders, according to people familiar with the private discussions. These alternatives remain rare and discretionary.
The H200 represents Nvidia’s high-performance AI chip tier, capable of leading large-scale model training at significantly faster speeds than older models. Chinese companies have preordered millions of units, but Nvidia’s current inventory is far smaller than total demand.
This enforcement shift goes beyond Nvidia’s usual deposit requirements. Previously, customers could place partial deposits to hold inventory. The full-payment demand marks a notable tightening tied to the unclear timeline for Chinese regulatory clearance.
Regulatory uncertainty is heightened because China’s industry ministry has not officially confirmed approval terms for H200 imports. Officials recently urged some domestic tech firms to pause their orders while they determine how many locally made chips must accompany imported ones.
The backdrop to this policy change lies in evolving U.S.–China technology trade rules. Last year, President Trump reversed a previous ban on advanced AI chip exports to China, allowing H200 shipments subject to a 25 percent revenue share paid to the U.S. government.
Even with Washington’s conditional export authorization, Nvidia faced setbacks in the past. A prior export ban on earlier AI chips led to a significant unsold inventory write-down. That experience likely influenced the company’s decision to require stronger financial commitments now.
Chinese regulators remain cautious. While some H200 shipments could be permitted this quarter for commercial use, Beijing is weighing domestic policy aims and technology self-reliance goals. Some plans reportedly exclude sales to sensitive sectors like the military and state infrastructure.
For Nvidia, this balancing act aims to capture robust Chinese demand without taking undue risk amid a shifting policy landscape. The company plans to fulfill early orders from existing stocks and has discussed expanding production through foundries like Taiwan Semiconductor Manufacturing Co. to meet future demand.
Chinese tech giants view the H200 as a key upgrade over older technology, even as domestic AI chips improve. Companies such as ByteDance and others see the H200’s advanced capabilities as crucial for competitive AI development.
The stricter payment terms shift financial risk from Nvidia to buyers. If approval delays or restrictions continue, customers may face sunk costs without delivery guarantees. This approach highlights how global tech trade tensions complicate the deployment of advanced computing hardware.
Analysts say the situation illustrates broader geopolitical tension around semiconductor supply chains. Nvidia must navigate U.S. export conditions, Chinese regulatory reviews, and demand from major clients eager to deploy cutting-edge AI infrastructure.
As talks with Chinese regulators progress, industry watchers will look for clearer signals about final import rules. The outcome could shape the flow of high-performance AI chips and influence strategies of global tech firms in one of the world’s largest markets.
