Nvidia Corp.’s suppliers have halted production of the company’s H200 artificial intelligence accelerator after China moved to block shipments of advanced chips, dealing another blow to the U.S. chipmaker’s access to one of its largest markets.
Recently, the suppliers had been working nonstop in anticipation of more than a million orders from China, aiming to meet March delivery targets. This week, however, Chinese customs officials informed agents that shipments of the H200 would not be allowed into the country.
Government authorities also warned local technology companies not to acquire the chips unless required, without specifying whether the ban is temporary or permanent.
Despite eager Chinese demand for Nvidia H200 chips, analysts suggest Beijing could be considering restrictions to advance local chip development or strengthen its negotiating position with the U.S. Some also claim the government wants to regulate which organizations can obtain top-tier computing resources and for what purposes, particularly in matters of national interest.
Earlier, Washington had restricted H200 shipments over fears they might give China an advantage in technology and defense. However, on Tuesday, officials from the Department of Commerce permitted Nvidia to sell its high-end AI chips in China, provided there’s enough supply in the US.
Before this, in December, President Trump had said he would only allow Nvidia to supply its H200 processors to designated customers in China and impose a 25% fee, stressing that he intended to protect national security and the nation’s AI dominance.
Over the course of 2025, Nvidia CEO Jensen Huang lobbied U.S. officials to allow sales of Nvidia’s powerful chips to China, arguing that global market participation is key to America’s competitive edge.
Thus, the firm on Tuesday welcomed the recent approval, saying it will benefit US manufacturing and jobs. Nonetheless, the green light came with conditions: chips must pass independent testing, sales to Chinese clients are capped relative to U.S. buyers, military use is restricted, and certain amounts must remain for domestic use. Plus, the recent Chinese block on shipments only complicates the matter further.
Current events also add another layer to a complex scenario in which the U.S.-designed, Taiwan-manufactured H200 chips can now be exported to China, with the U.S. government reportedly receiving part of the profits. Rather than being shipped straight from Taiwan to China, the U.S. government noted that the chips first go through a U.S. lab for testing, where a 25% tariff is imposed. This also affects AMD’s MI325X processor.
Analysts are split over whether selling the H200 to China makes strategic sense. Supporters argue it could slow China’s domestic chip development, while critics highlight its potential use in weapons systems.
According to Austin Lyons, a semiconductor analyst, China is concerned about its over-dependence on Nvidia, but local companies are expected to continue pursuing H200 chips until homegrown versions advance. Nvidia, he added, will be glad to gain revenue from China, even with lower margins caused by the U.S. government’s share.
Marc Einstein at Counterpoint Research also argued the proposal to take a cut of Nvidia’s sales might set a benchmark for Trump’s negotiations on other tariffs. He remarked, “It will be interesting to see if this tariff model expands to other sectors.”
Sharpen your strategy with mentorship + daily ideas - 30 days free access to our trading program