US Restricts Nvidia Corp. From Selling AI Chips to China and Russia, Company Says

A sign is posted in front of the Nvidia headquarters in Santa Clara, Calif., on May 10, 2018. Justin Sullivan/Getty Images

Tech firm Nvidia Corp. has been told by Washington not to sell two of its artificial intelligence (AI) acceleration chips to Beijing and Moscow, potentially affecting millions of dollars’ worth of revenues.

“On August 26, 2022, the U.S. government, or USG, informed NVIDIA Corporation, or the Company, that the USG has imposed a new license requirement, effective immediately, for any future export to China (including Hong Kong) and Russia of the Company’s A100 and forthcoming H100 integrated circuits,” NVIDIA said in a regulatory filing with the Securities and Exchange Commission (SEC) dated Aug. 26.

“DGX or any other systems which incorporate A100 or H100 integrated circuits and the A100X are also covered by the new license requirement.”

Nvidia’s A100 and H100 processors enable AI developers to speed up research while being able to develop more advanced AI models. H100, expected to be launched this year, is six times faster than the A100 processor. DGX is a line of Nvidia servers and workstations used to accelerate deep learning applications.

Washington informed the company that the new licensing requirement will address the risk of these two chips being used for military purposes in Russia and China.

In its filing, Nvidia made it clear that it does not sell products in Russia. However, China is a big market for the company. The firm admitted that around $400 million in potential sales to China in the third fiscal quarter could be subject to the new license requirements.

California-headquartered AMD has also been provided with new requirements by the U.S. Department of Commerce that will affect shipments of MI250 integrated circuits to China, the company said to CNN.

Chip Sales to China

The order comes amid reports that China has been purchasing U.S.-designed semiconductors to boost its AI advancements.

In June, the Center for Security and Emerging Technology at Georgetown University (CSET) published a report stating that the Chinese People’s Liberation Army (PLA) is making “significant progress” in adapting AI for combat technologies.

“Of the 97 individual AI chips we could identify in public PLA purchase records, nearly all of them were designed by Nvidia, Xilinx (now AMD), Intel, or Microsemi,” the report said. Only one originated from China, coming from Fudan Electronics. However, this was a clone of an American design.

The report goes on to mention that it could not find “any records” of state-owned defense firms or PLA units awarding contracts to Chinese-designed AI chips. The analysis highlights the significance of the Chinese military’s dependence on American products to advance its interests.

Despite having the upper hand, Washington has continued allowing China to benefit from other U.S. technologies. According to a Wall Street Journal analysis, officials approved 94 percent of technology export applications to China in 2020. Though the approval rate declined last year, it remained elevated at 88 percent.