About TrendForce News

TrendForce News operates independently from our research team, curating key semiconductor and tech updates to support timely, informed decisions.

[News] TSMC, Samsung Reportedly Pause China Consumer Chip Shipments Under New U.S. Rules; 16nm Orders Affected


2026-11-20 Semiconductors editor

TSMC and Samsung have reportedly suspended shipments of certain chips to Chinese customers following tighter U.S. export controls, with the impact potentially extending beyond AI processors to consumer electronics. According to Financial Times, both chipmakers are said to have informed Chinese customers in recent days that some deliveries would be halted while they seek licenses from the U.S. Commerce Department.

As Financial Times notes, the new rules remove exemptions previously available for chips designed for non-AI applications, including smartphones. A lawyer briefed on the situation said TSMC and Samsung must now seek U.S. licenses for consumer chips manufactured using advanced processes. TSMC supplies smartphone processors to Chinese brands including Xiaomi, Oppo, and Vivo, the report adds.

Despite the expanded restrictions, consumer smartphone chips are still likely to receive U.S. export licenses, according to Financial Times, citing industry sources. However, lengthy approval procedures could disrupt production schedules.

TSMC Reportedly Halts Certain 16nm Orders, Tightens Chinese Customer Screening

Beyond the potential disruption to consumer chip shipments, the new restrictions are also prompting TSMC to tighten its manufacturing requirements for Chinese customers. Reuters, citing sources, reports that TSMC is tightening restrictions on Chinese chip designers, instructing affected customers to halt orders using 16nm and more advanced processes unless they obtain U.S. export licenses. Under the foundry’s reported guidance, affected Chinese designers will generally be limited to older 28nm and 40nm technologies, provided their chips do not use high-bandwidth memory (HBM) or require advanced packaging.

The changes also eliminate verification procedures previously available to certain Chinese customers. According to Reuters, TSMC could previously assess chip designs based on computing power and other technical specifications to determine whether they fell outside U.S. export controls. Under the new rules, affected advanced chips are instead presumed to require licenses unless an exception applies. The Financial Times similarly reports that Chinese chip designers can no longer rely on independent packaging and testing companies to verify that high-performance processors fall outside U.S. restrictions.

TSMC is also reportedly stepping up customer screening, examining operating locations, parent companies, major shareholders, packaging providers, and shipping routes. Reuters notes that the checks extend to orders placed through overseas subsidiaries or intermediaries in Taiwan, the U.S., Europe, Japan, and Southeast Asia.

U.S. Chip Curbs Threaten China’s AI Development and Supply Chains

The restrictions could particularly affect Chinese companies developing alternatives to NVIDIA processors, including independent chip designers and tech giants with in-house chips. Reuters notes that a U.S. government list of approved chip designers includes NVIDIA, AMD, Apple, MediaTek, and Realtek, but no Chinese companies.

The measures could add to existing supply challenges for Chinese AI chip developers. As Financial Times notes, Samsung’s Chinese customers include Baidu’s chip unit Kunlunxin and ByteDance, both of which have sought alternative chip supplies amid shortages of NVIDIA processors in China. Adding to the broader supply uncertainty, Commercial Times reports that NVIDIA has stopped shipping previously authorized Hopper-generation data center products to China.

Chinese technology companies are also assessing the broader impact on their AI operations. According to the AP, Alibaba has warned that the restrictions could affect its global data center expansion, while Tencent said reduced access to advanced computing capacity could delay training of its next-generation AI models.

Read more

(Photo credit: The White House)

Please note that this article cites information from Financial Times, Reuters, and AP.



Get in touch with us