The Biden administration on Tuesday announced additional limits on the types of advanced semiconductors that US companies can sell to China, lifting restrictions issued last October to limit China’s progress on supercomputing and artificial intelligence.
It appears that these rules could block most shipments of advanced semiconductors from the United States to Chinese data centers, which use them to build models capable of artificial intelligence. More U.S. companies wanting to sell advanced chips or the machinery used to make them in China will have to notify the government of their plans, or obtain a special license.
To prevent the risk of advanced American chips reaching China via third countries, the United States will also require chip makers to obtain licenses to ship to dozens of other countries that are subject to a US arms embargo.
The Biden administration argues that China’s access to such advanced technology is dangerous because it could aid the country’s military in tasks such as guiding hypersonic missiles, setting up advanced surveillance systems or cracking top-secret US codes.
Leading AI experts have warned that the technology could pose an existential threat to humanity if it is not properly managed.
But artificial intelligence also has valuable business applications, and tougher restrictions could hit Chinese companies trying to develop AI chatbots such as ByteDance, TikTok’s parent company, or Internet giant Baidu, industry analysts said. In the long run, the limitations could also weaken China’s economy, given that AI is transforming industries ranging from retail to health care.
The limit is also likely to cut into the money that US chipmakers like Nvidia, AMD and Intel make from selling advanced chips to China. Some chipmakers make as much as a third of their revenue from Chinese buyers and have lobbied against tougher restrictions in recent months.
US officials said the rules would exempt chips that were solely for use in commercial applications such as smartphones, laptops, electric vehicles and gaming systems. Most of the rules will go into effect in 30 days, although some will go into effect sooner than that.
In a statement, the Semiconductor Industry Association, which represents major chip makers, said it was evaluating the impact of the updated rules.
“We recognize the need to protect national security and believe that maintaining a healthy U.S. semiconductor industry is an essential component to achieving that goal,” the group said. “Overly broad, unilateral controls risk harming the U.S. semiconductor ecosystem without advancing national security because they encourage foreign customers to look elsewhere.”
Speaking with reporters on Monday, a senior administration official said the United States has seen people try to work around rules before, and recent breakthroughs in generative AI have made regulators more aware of this. That’s how the so-called big language models behind it were being developed and used.
Commerce Secretary Gina M. Raimondo said the changes were made “to ensure that these rules are as effective as possible,” and that she expected the rules to be updated at least annually as technology advanced.
Referring to the People’s Republic of China, he said, “The goal is the same goal it has always been, which is to limit the PRC’s access to advanced semiconductors that can fuel breakthroughs in artificial intelligence and sophisticated computers that power the PRC military. are important for applications. ,
“It is more important than ever to control technology as it relates to national security,” he said.
The tougher rules could anger Chinese officials as the Biden administration tries to improve ties and prepare for a possible meeting between President Biden and China’s top leader Xi Jinping in California next month.
The Biden administration is trying to counter China’s growing mastery of many cutting-edge technologies by pumping money into new chip factories in the United States. It is also trying to impose strict but narrow restrictions on exports of technology to China that could have military uses, while allowing other trade to flow freely. US officials describe this strategy as “protecting American technology”.a small yard and a high fence,
But determining which technologies truly pose a threat to national security has been a controversial task. Major semiconductor companies such as Intel, Qualcomm and Nvidia have argued that overly restrictive trade restrictions could deprive them of the revenue they need to invest in new plants and research facilities in the United States.
Some critics say the limits could also embolden China’s efforts to develop alternative technologies, ultimately weakening US influence globally. Chinese researchers have made significant progress in developing domestic versions of advanced chips, but experts say the country remains years behind Western capabilities for semiconductor manufacturing.
The changes announced Tuesday have a particularly significant impact on Nvidia, one of the biggest beneficiaries of the artificial intelligence boom.
In response to the Biden administration’s first major restrictions on artificial intelligence chips a year ago, Nvidia designed new chips, the A800 and H800, for the Chinese market, which were slower but still used by Chinese companies to train AI models. Could be used for. A senior administration official said the new rules would restrict those sales.
Nvidia has said China typically generates 20 percent to 25 percent of the company’s data center revenue, which includes products other than AI-enabling chips.
In addition to those expanded sanctions, the United States will create a “gray list” that will require manufacturers of certain less advanced chips to notify the government if they are selling them to China, Iran or other countries subject to the US arms embargo.
Administration also kept Thirteen Chinese companies and subsidiaries involved in the development of integrated circuits for advanced computing have been included in an “entity list” that requires US companies to obtain special permission before sending material to them.
In a note to clients last week, Julian Evans-Pritchard, head of China economics at research firm Capital Economics, said the effects of the controls will become more pronounced as non-Chinese companies introduce more advanced versions of their existing products and as As their data sets grew larger, the amount of computing power needed to train AI models increased.
“The result is that China’s ability to reach the technological frontier in large-scale AI model development will be hindered by US export controls,” Mr Evans-Pritchard wrote. This could have a huge impact on the Chinese economy, he said, because “we think AI has the potential to be a game changer for productivity growth over the next few decades.”