The US-China chip ban, explained

computer chip with chinese flag

It’s been more than a month since the US Commerce Department issued new rules that halted exports of some advanced chips—those with military or AI applications—to Chinese customers.

China has yet to respond – but Beijing has multiple options in its arsenal. It’s unlikely, experts say, that U.S. action will be the last fighting word in an industry that’s becoming more geopolitically sensitive by the day.

This is not the first time that the US government has restricted the flow of chips to its perceived adversaries. Earlier, the US stopped selling chips to individual Chinese customers. In response to Russia’s aggression in Ukraine earlier this year, the United States (along with several other countries, including South Korea and Taiwan) placed Russia under a chip embargo.

But none of these previous U.S. chip bans were as sweeping as the new rules issued on Oct. 7. “This announcement is probably the most extensive export control in decades,” said Sujay Sivakumar, an analyst at the Center for International and Strategic Studies in Washington.

The rules, to Chinese consumers, prohibit the sale of advanced chips with both high performance (at least 300 trillion operations per second, or 300 teraoops) and fast interconnect speeds (typically, at least 600 gigabits per second). Nvidia’s A100, by comparison, is capable of 600 terabytes and matches interconnect speeds of 600 Gb/s. Nvidia’s more impressive H100 can reach around 4,000 trillion operations and 900 Gb/s. Both chips, intended for data centers and AI trainers, cannot be sold to Chinese customers under the new rules.

Additionally, the rules restrict the sale of fabrication equipment if it is knowingly used to manufacture certain classes of advanced logic or memory chips. These include logic chips produced at nodes of 16 nanometers or less (which Intel, Samsung, and TSMC have done since the early 2010s); NAND long-term memory integrated circuits with at least 128 layers (state of the art today); or DRAM short-term memory integrated circuits manufactured at 18 nanometers or less (which Samsung began manufacturing in 2016).

Chinese chipmakers have barely scratched the surface of that number. SMIC launched 14-nm mass production this year despite facing existing US sanctions. YMTC started shipping 128-layer NAND chips last year.

The rules restrict not only US companies, but also citizens and permanent residents. US employees of Chinese semiconductor firms have had to pack up. ASML, a Dutch manufacturer of fabrication equipment, has asked US employees to stop servicing Chinese customers.

Speaking of Chinese consumers, most – including office workers, gamers, small chip designers – probably won’t feel the controls. “Most of the chip trade and chip manufacturing in China is overwhelming,” said Christopher Miller, a historian who studies the semiconductor trade at Tufts University.

Controlled types of chips instead go into supercomputers and large data centers, and are desirable for training and running large machine-learning models. Above all, the U.S. hopes to prevent Beijing from using the chips to enhance its military power—and potentially preempt an attack on Taiwan, where most of the world’s semiconductors and microprocessors are manufactured.

To close a possible bypass, the controls also apply to non-US companies that rely on US-made equipment or software. For example, Taiwanese or South Korean chipmakers cannot sell advanced chips to Chinese customers that are built on US-made technology.

It is possible to apply to the US government for an exemption from at least some of the restrictions. For example, Taiwanese fab juggernaut TSMC and South Korean chipmaker SK Hynix have already received temporary exemptions for one year. “It’s hard to say what happens after that,” said Patrick Schroeder, a researcher at Chatham House in London And the Commerce Department has already said such licenses will be the exception, not the rule (although Commerce Department Undersecretary Alan Estevez has suggested that about two-thirds of licenses are approved).

More export controls could be the route. Estevez indicated that the government is considering restrictions on technology in other sensitive areas — specifically citing quantum information science and biotechnology, both of which have seen China-based researchers make major advances in the past decade.

The Chinese government has so far responded with harsh words and little action. “We don’t know if their response will be an immediate response or if they have a long-term approach to deal with it,” Sivakumar said. “It’s speculation at this point.”

Beijing may work with foreign companies whose revenues in the lucrative Chinese market are now threatened. “I’m not really aware of a specific company that thinks it’s going to be a winner,” says Sivakumar. This week, in the eastern city of Hefei, the Chinese government hosted a chipmakers conference whose participants included US firms AMD, Intel and Qualcomm.

Nvidia has already responded by introducing a China-specific chip, the A800, which appears to be a modified A100 cut to meet the requirements. Analysts say Nvidia’s approach could be a model for other companies to maintain Chinese sales.

There may be other tools that the Chinese government can use. While China may be dependent on foreign semiconductors, foreign electronics manufacturers depend on China for rare-earth metals — and China supplies the supermajority of the world’s rare earths.

China has a history of reducing its rare-earth supply for geopolitical leverage. In 2010, a Chinese fishing boat collided with two Japanese coast guard vessels, sparking an international incident when Japanese authorities arrested the boat’s captain. In response, the Chinese government halted rare-earth exports to Japan for several months.

Of course, much of the conversation centered on US actions and Chinese responses. But for third parties, the entire dispute provides a constant reminder of how volatile and volatile the chip supply can be. In the European Union, which accounts for less than 10 percent of the world’s microchip market, the debate has fueled interest in the potential European Chips Act, a plan to invest heavily in manufacturing in Europe. “For Europe in particular, it’s important not to get caught up in this US-China trade issue,” Schröder said.

“The way the semiconductor industry has evolved over the past few decades has predicated a relatively stable geopolitical order,” says Sivakumar. “Of course, the ground reality has changed.”