Nvidia ships a small batch of H200 AI chips to China after US export approval, report says

Nvidia has shipped a small number of its H200 artificial intelligence chips to customers in China after winning approval to export the product, according to a report from Bloomberg published by Yahoo Finance. The report attributes the development to a Trump administration official who oversees export controls, describing the move as limited in scale.

The chips in question are Nvidia’s H200, a data-center processor designed to accelerate machine-learning and generative AI workloads. For Nvidia and its customers, H200 is part of the company’s broader effort to supply compute used in training and running AI models, with demand tied to cloud operators and large enterprises building AI systems.

The report characterizes the shipment as “small” and frames it as the result of US approval rather than a broad relaxation of restrictions. That distinction matters because Nvidia’s ability to supply advanced AI accelerators to China has been constrained by US export rules that govern both which products can be shipped and under what licensing conditions.

Under the reported scenario, the US approval indicates that at least some licenses can be granted for certain shipments, even as the overall framework remains restrictive. However, the article does not describe how many chips were shipped, which specific customers received them, or how the approval was reached within the licensing process.

Nvidia has not, in the information provided for this report, issued a specific public statement detailing the volume of any China-bound H200 shipments. Likewise, the report does not provide documentary details such as license terms, ship dates, or end-use conditions that typically accompany export authorizations.

For the broader AI semiconductor market, any incremental easing of export limitations can affect near-term supply expectations and customer planning. Still, the reported scale described as “small” suggests the impact is likely to be incremental rather than a reversal of the restrictions that have shaped Nvidia’s China revenue outlook.

What remains unclear is whether the approvals will extend beyond a limited batch, whether additional licenses are pending, and whether the approvals apply to H200 units only or other Nvidia AI systems. The report also does not say how US officials are balancing national-security concerns against commercial demand as they evaluate future requests.

Going forward, investors and customers will likely watch for further disclosures that could come from Nvidia, US export-control authorities, or regulators, including any signs of expanded licensing for additional shipments or product variants. The pace and breadth of approvals will be key to determining whether this is a one-off delivery or the start of a longer-term change in access.