Nvidia, the leading manufacturer of AI chips, has stated that it expects a $5.5 billion loss after the U.S. government imposed stricter export controls targeting its H20 processor, designed specifically for China.
The company claimed it was informed last week that the chip, which was previously consistent with existing export restrictions, requires a special license to be sold in China, including Hong Kong, for the "indefinite future."
Nvidia's H20 chip is one of its most popular AI products in China. These new restrictions, part of rising US-China trade tensions, aim to prevent the chips from being used in Chinese supercomputers. As a result, Nvidia stated that the expected expenses are dependent on inventory, purchase commitments, and related reserves for the affected chips.
The impact was immediate, and Nvidia shares fell around 6% in after-hours trading. The news also caused larger market ripples, with chip makers including AMD falling 7%, South Korea's Samsung Electronics and SK Hynix down up to 3%, and Dutch chip equipment maker ASML losing 5% in early European trading.
Despite the setback, analysts like Marc Einstein of Counterpoint Research believe Nvidia is financially secure enough to absorb it. Still, the developments highlight the importance of AI processors in the geopolitical tug-of-war for technological domination.
Since 2022, the U.S. has steadily tightened controls over semiconductor exports to China, claiming national security risks. At the same time, former President Donald Trump recently announced increased tariffs on imported semiconductors, which could further pressure the global chip supply chain. Although the semiconductor sector has been mostly spared from Trump's tariffs, additional taxes could affect it directly.
To mitigate long-term risks, Nvidia has announced plans to invest up to $500 billion in U.S.-based AI infrastructure over the next four years.
The company, which designs its chips in the US but relies on Taiwanese manufacturers for production, is moving its focus to domestic capabilities as the global tech scene grows more fragmented.