TSMC has reportedly stopped shipping chips to Chinese designers after new U.S. export rules flagged decade-old 16nm and 14nm nodes, according to Tom's Hardware. The change affects chip designers using those process nodes, including makers of smartphone processors, autonomous driving chips and other consumer electronics, not just data center hardware. Washington's earlier export controls had targeted only high-performance AI chips such as Nvidia's Blackwell and the upcoming Vera Rubin chips.

Under the new rules, front-end foundries and outsourced semiconductor assembly and test companies, also known as packaging and testing companies, must treat any semiconductor built on a 16nm or 14nm node as an export-controlled advanced computing chip by default. The ruling states that even a designed chip with fewer than 30 billion transistors and no HBM cannot be assumed not to be built for high-performance computing. Designers using 16nm or newer nodes, even when their chips are not meant for data centers, must obtain documentation from an approved designer certifying the chip's Total Processing Performance or performance density to avoid added scrutiny from the fab.

The 16nm and 14nm nodes have existed for roughly 10 years and appear in consumer devices such as smartphones and TVs as well as in automotive applications. A source told the publication that the new rule is very strict and that its scope reaches far beyond data center chips to smartphone processors, autonomous driving and other consumer electronics types if they use 14nm or 16nm or more advanced technology. According to the source, the move would delay shipments to many chip developers worldwide until they clear compliance and would hit many emerging startups in many Asian nations.

Separately, the U.S. now requires export licenses to ship advanced AI chips to Indonesia, Malaysia, Singapore, Brazil and Thailand. The IFR says those countries are often used as transshipment points for chips diverted to China, as shown by the Supermicro and Megaspeed investigations, which used one of these countries as the destination for Nvidia servers before they were forwarded to their final destination in China. The new rules mark an expansion from Washington's previous controls, which focused only on high-performance AI chips.

Chip smuggling has been a long-simmering issue in the United States and appears to have boiled over after the revelation of Jensen Huang's affair with Megaspeed executive Alice Huang, who is also suspected of being instrumental in diverting $7.5 billion worth of Nvidia chips and passing information to Chinese intelligence. Sources say that while the alleged relationship with a direct subordinate was the apparent primary cause of his departure from Nvidia, the real reason behind his ousting was the mass smuggling of 1.5 million Nvidia AI GPUs that happened right under his nose.

More hardware news from TechManNews.