On Wednesday, the US administration implemented a 25% tariff on a category of advanced computing chips, an action detailed in a White House fact sheet. The tariff specifically encompasses high-performance microchips such as Nvidia's H200 and AMD's MI325X. This step is part of a wider initiative aimed at consolidating the country's position in the technology sector, particularly in the field of artificial intelligence (AI).
Notably, the tariff includes a conditional clause granting exemptions for chips imported with the purpose of supporting and expanding the US's own technological supply chain. Despite this provision, there is currently no public clarity regarding the criteria that importers must satisfy to secure such exemptions. Attempts by media outlets to obtain further details from the White House have yet to receive responses.
The White House document also indicated the potential for imposing more extensive tariffs on a broader range of semiconductors and related derivative products in the near future. This aligns with the administration's strategic focus on fostering domestic manufacturing capabilities within the tech sector.
Nvidia, whose graphics processing units are pivotal to powering AI-driven data centers, has been central to the ongoing AI surge. The company's CEO, Jensen Huang, has emerged as a notable figure within the context of US government AI initiatives. This prominence places Nvidia at the intersection of the US's trade and technology policy.
The justification for these tariffs is grounded in national security concerns, specifically invoked through Section 232 of the Trade Expansion Act of 1962. This section empowers the president to apply import restrictions if certain products threaten national security interests.
Earlier statements by the administration indicated that Nvidia's H200 chip would still be allowed entry into the Chinese market; however, the US government would impose a 25% duty on these sales. Nvidia has not publicly commented on the tariff implementation following inquiries.
AMD, another major chip manufacturer affected by the tariff, issued a statement affirming their compliance with all relevant US export control laws and policies. No further details were provided.
This tariff strategy is not without precedent under the current administration. In August, President Trump threatened to apply 100% tariffs on chips and semiconductor imports but highlighted that companies with tangible commitments to domestic manufacturing would be exempt from these steep levies.
Throughout his tenure, the president has emphasized AI as a critical component of his agenda, signing multiple executive orders and unveiling plans designed to reduce regulatory hurdles and accelerate AI adoption and innovation within the United States.
Nevertheless, some industry leaders, including Nvidia's CEO Jensen Huang, have expressed concerns over stringent export controls targeting China. They argue that such restrictions may unintentionally encourage technological advancements within China’s domestic chip industry, complicating US objectives.
This dynamic has positioned key semiconductor firms, especially Nvidia, amid complex trade tensions between the US and China. In August, both Nvidia and AMD disclosed an arrangement under which 15% of sales revenue from chips sold in China would contribute to the US government. This agreement appeared to pertain primarily to earlier generations of chips rather than the latest models.
Initial public indications of imposing a 25% tariff on chip imports emerged from President Trump in February, with a target implementation around April. However, the formal investigative process under Section 232, necessary for finalizing such tariffs, was only initiated several months later.
It is important to recognize that these tariffs operate independently of another set of levies currently under legal examination before the Supreme Court.
Given the ongoing developments in this area, further updates are anticipated as more information becomes available.