- President Donald Trump has imposed a 25 per cent tariff on certain advanced computing chips, including Nvidia’s H200 and AMD’s MI325X models, citing national security.
- The tariff excludes many chips used domestically for AI and tech infrastructure, raising questions about broader semiconductor policy and trade impact.
What happened: U.S. imposes targeted 25 per cent tariff on imported high‑end chips
On 14 January 2026, U.S. President Donald Trump signed a proclamation imposing a 25 per cent tariff on certain imported advanced computing chips, notably Nvidia’s H200 and AMD’s MI325X processors. The decision follows a nine‑month investigation under Section 232 of the Trade Expansion Act of 1962, which allows tariffs to be set on national security grounds.
The tariff is narrowly defined and does not apply to chips imported for U.S. data centres, startups, non‑data‑centre consumer or civil industrial applications, or public sector uses. This carve‑out aims to shield parts of the domestic tech ecosystem while still targeting the flow of high‑performance chips that could be re‑exported or sold abroad.
In addition, the proclamation requires that chips made in Taiwan — a key manufacturing hub for companies like Nvidia and AMD through partners such as Taiwan Semiconductor Manufacturing Co. (TSMC) — and destined for China pass through U.S. testing labs before export, subjecting them to the new tariff.
Also Read: US postpones China chip tariffs to 2027, seeking trade peace while raising stakes
Also Read: US delays additional tariffs on Chinese chips until June 2027
Why it’s important
The tariff move forms part of a broader effort by the Trump administration to reduce U.S. dependence on foreign semiconductor supply chains and encourage domestic production. The United States currently manufactures only around 10 per cent of the chips it consumes, leaving it reliant on overseas producers, particularly in Taiwan.
Proponents argue such tariffs can incentivise reshoring of semiconductor manufacturing and protect technology supply chains from geopolitical disruption. However, critics question the practical impact, noting that tariffs can raise costs for U.S. tech firms relying on advanced chips for AI research, cloud services and data‑centre operations. Exempting chips for domestic AI use suggests a tension between security objectives and industrial demand.
There is also uncertainty about how effective tariffs are in an industry where design and manufacturing are globally intertwined. For example, while Nvidia and AMD design many of their chips in the United States, actual fabrication often occurs abroad. Analysts worry that tariffs aimed at imports might not significantly change long‑term manufacturing patterns without complementary investments, such as those provided under the CHIPS and Science Act to incentivise U.S. production.
Additionally, the tariff’s narrow scope may leave significant portions of the semiconductor trade outside the levy, raising questions about broader trade and industrial strategy. Whether tariffs will expand to cover a wider range of chips and derivative products remains an open question, with potential ramifications for global AI and tech supply chains.
