US Officials Order Nvidia to Stop Selling Top AI Chips to China

Register now for free, unlimited access to

(Reuters) – Chip designer Nvidia (NVDA.O) said on Wednesday that U.S. authorities have instructed the company to stop exporting two of its major artificial intelligence-related computing chips to China. . It works like image recognition and hinders his Nvidia business in China.

Nvidia shares fell 6.6% after hours. The company said a ban affecting A100 and H100 chips designed to speed up machine learning tasks could prevent Nvidia from completing development of his H100, the flagship chip announced earlier this year. I was.

Shares of Nvidia rival Advanced Micro Devices (AMD.O) fell 3.7 percent after hours. An AMD spokesperson told Reuters the company had received new licensing requirements that would halt exports of its MI250 artificial intelligence chip to China, but it believes its MI100 chip would not be affected. He said he doesn’t believe the rules will have a material impact on the company’s business.

Register now for free, unlimited access to

According to Nvidia, U.S. officials said the new rule “addresses the risk that covered products may be used or diverted for ‘military end uses’ or ‘military end consumers’ in China. There is,” he said.

Asked for comment, the U.S. Department of Commerce did not specify what new standards it had set for AI chips that could no longer be shipped to China, but said it was reviewing China-related policies and practices. . hand.

“Although we are not in a position to outline specific policy changes at this time, we are aware of the additional technology, end-use, and end-user-related changes that are necessary to protect the national security and foreign policy interests of the United States. We are taking a comprehensive approach to implementing the measures,” a spokesperson told Reuters.

Technology company Nvidia’s logo is seen at its headquarters in Santa Clara, California, on February 11, 2015. REUTERS/Robert Galbraith

The announcement marks a significant escalation in the US crackdown on China’s technological prowess amid growing tensions over the fate of Taiwan, where chips for Nvidia and nearly every other major chip company are made. increase.

Without US-made chips from companies like Nvidia and AMD, Chinese organizations would be unable to cost-effectively perform the advanced computing used for image and speech recognition, among many other tasks. will not be possible.

Image recognition and natural language processing are common in consumer applications such as smartphones that can answer queries and tag photos. It also has military applications, such as scanning satellite imagery for weapons and bases, and filtering digital communications for intelligence gathering purposes.

Nvidia said it booked $400 million this quarter in sales of affected chips to China if the Chinese company decided not to buy Nvidia’s alternatives. It said it plans to apply for exemptions from the rules, but there is “no guarantee” that U.S. authorities will grant them.

Stacey Rasgon, a financial analyst at Bernstein, said the disclosure, which investors have been watching closely in recent years, comes from China, about 10% of Nvidia’s data center sales, and the hit to sales comes from Nvidia. This indicates that it is likely to be “manageable” for

“It’s not a change of[investment]thesis, but it doesn’t look good,” Rasgon said. “It’s a matter of what’s going on on both sides right now,” he said of possible escalation going forward.

The chip ban comes after Nvidia already predicted last week that its revenue would drop sharply this quarter on the back of a weakening gaming industry. Nvidia says he expects third-quarter revenue to fall to $5.9 billion, down 17% from the same period last year.

Register now for free, unlimited access to

Reporting by Eva Matthews and Nivedita Bal from Bangalore, Stephen Nellis and Jane Lee from San Francisco, Karen Freifeld from New York and Alexandra Alper from Washington.Edited by David Gregorio and Matthew Lewis

Our standards: Thomson Reuters Trust Principles.


Leave a Comment