Washington has imposed unprecedented restrictions on US citizens and residents who work for Chinese semiconductor companies.

In its ongoing tech war with China, the United States has taken multiple steps in the past decade to hamper Beijing’s ascendance in the technology sector. 

Washington has already stonewalled Huawei from the 5G market, imposed sanctions on another Chinese telecom firm ZTE and bullied Dutch company ASML into holding back delivery of chip-making equipment to Chinese firms. 

But the policy announced by President Joe Biden on October 7 is by far the toughest as it goes beyond restricting technology transfer to China’s military, the primary cause of concern for Washington. 

“The latest US export controls mainly target US persons who work for semiconductor firms in China. This measure in my view is a huge blow,” says Dr Monique Chu, a lecturer of Chinese politics at the University of Southampton. 

Since the early-2000s, China’s semiconductor firms have relied heavily on foreign talent, including expats educated in the West and engineers from neighbouring Taiwan. 

“And many of these Chinese returnees and Taiwanese engineers working in China's industry either have US green cards or dual nationalities,” Dr Chu tells TRT World

Under latest export controls, US citizens, green card holders and residents would require a government licence if they want to work with China’s semiconductor companies in any capacity. 

This includes even support service, a routine work that companies offer when they sell any product to a client. 

US lawmakers have made it clear that such licences or permits won’t be given in the first place, effectively stopping American citizens and residents from working in China. 

US President Joe Biden delivers remarks during a signing event for the CHIPS and Science Act of 2022, on the South Lawn of the White House in Washington, United States, August 9, 2022.
US President Joe Biden delivers remarks during a signing event for the CHIPS and Science Act of 2022, on the South Lawn of the White House in Washington, United States, August 9, 2022. (Reuters)

Hitting the heart

Chinese semiconductor firms have long relied on foreign talent to catch up with their Western peers such as Intel and Qualcomm. 

The Shanghai-based Semiconductor Manufacturing International Corporation (SMIC), China’s largest contract chip maker, was established by Richard Chang, a Taiwanese-American. 

In the early years, around one-third of SMIC’s engineers were foreigners, including many Americans, Dr Chu wrote in her book The East Asian Computer Chip War

Before the latest curbs, US officials mainly focused on blocking chip designers and manufacturers from selling their products to the Chinese military. 

Now a blanket ban has been imposed on the sale of equipment and tools without which it is nearly impossible to make advanced chips, especially those needed for artificial intelligence (AI) applications. 

Washington is banking on its leverage over the global semiconductor supply chain to stop China from acquiring the capability of making chips at nodes of 14 nanometers (nm) or beyond. 

A nanometer is the size of a transistor, the building blocks of chips -- the smaller the size, the more transistors can be piled on to a chip, making them more efficient and cost-effective. 

Leaders in the integrated circuit market such as Taiwan’s TSMC have already achieved the capability to produce 5nm chips on a commercial scale. Whereas, under the new rules, China’s SMIC will face difficulty producing even 14nm chips. 

“The latest curbs are much broader and prevent American capital equipment companies, even those offshore, from selling or servicing equipment in Chinese chip factories or fabs,” says Douglas Fuller, a professor at Copenhagen Business School and author of Paper Tigers, Hidden Dragons: Firms and the Political Economy of China's Technological Development

“The US is preventing EDA tool vendors from selling to chip design firms in China that are designing advanced chips, such as AI chips. Without the EDA software, it will be very difficult to design these chips.” 

EDA or Electronic Design Automation refers to the sophisticated blueprint of the chip-making process, including the complex layout of the integrated circuit. No chip can be manufactured without an EDA and only a handful of companies, such as Synopsys, specialise in this field. Almost all of them are based in the US. 

Similarly, the advanced photolithography machines essential for manufacturing modern chips use components and technology developed in the US. That’s one reason Washington has stopped the Netherlands-based ASML from selling its extreme ultraviolet lithography (EUVL) equipment to Chinese firms. 

The urgency to act stems from the fear that America and its allies could fall behind in the race for advanced chips, which will power everything from self-driven cars, supercomputers, data centres to space rockets.

In August, Biden signed the CHIPS and Science Act, promising billions of dollars in subsidies to convince the likes of Samsung and TSMC to open up production facilities in the US. 

While the US has been in control of back-end processes -- the EDA tools, machines and chemicals -- required at the various stages of semiconductor manufacturing, the actual production has moved to Asia. 

For instance NVIDIA, a giant in semiconductors, relies on TSMC’s factories for the production of its powerful GPUs (graphic processing units). 

“In the next five years at least, it will be impossible (for China) to sidestep these chip manufacturing sanctions unless American allies, Japan and the Netherlands, defy the US,” says Douglas. 

But both Tokyo and Amsterdam have their own interests to protect and will likely abide by the US restrictions, he says.

Source: TRT World