On Friday, American officials told Nvidia that the technology company needs a U.S. government license in order to export products whose performance equals or exceeds the capabilities of its A100 graphics processing chips to China, Russia, or Hong Kong.
When the company revealed this on Wednesday, its shares fell by 4% and fell again Thursday morning. The company had allegedly booked $400 million in sales to China of products that fall under the new requirement, which may now have to be scrapped.
This move by the U.S. government constitutes a significant crackdown on Chinese technological capabilities. It comes amid increased uncertainty regarding the fate of the Republic of Taiwan, which has faced recent escalations and encroachments by Chinese communist forces.
The majority of Nvidia's chips and those of virtually every other major chip manufacturer are produced in Taiwan, upon which both the U.S. and China are heavily reliant. The U.S. regards the potential of Chinese technological supremacy as both a strategic threat and a challenge to American interests.
Nvidia alleged that part of the American motive behind the new rule was addressing "the risk that the covered products may be used in, or diverted to, a 'military end use' or 'military end user' in China."
The A100 chips are used to power high-performing data centers for AI, data analytics, and supercomputing. Such technology would enable the communist Chinese regime to continue to execute advanced work like image recognition or to further power the kind of systems it has employed in its Xinjiang concentration camps.
As for military uses, such chips can be utilized to analyze satellite imagery for weapons and military installations, as well as to sift through digital communications for intelligence-gathering purposes.
A group of experts chaired by former Google CEO Eric Schmidt stated last year that the U.S. was "not prepared to defend or compete in the AI era."
The group's 756-page National Security Commission on Artificial Intelligence report indicated that the possibility of Chinese technological supremacy poses a threat to the homeland: "We will not be able to defend against AI-enabled threats without ubiquitous AI capabilities and new warfighting paradigms."
The report also indicated that the CCP "has embarked on a multi-pronged campaign of licit and illicit technology transfer to become a 'science and technology superpower' by 2050." The communist regime deliberately seeks to target "U.S. critical sectors, companies, and research institutions," stealing their technology and costing America $300 billion to $600 billion per year.
The new rule intended as a check on Chinese ambitions does not only target Nvidia. Rival tech company Advanced Micro Devices, whose shares similarly dropped after the announcement, told Reuters that it too received similar license requirements. It is no longer permitted to export its MI250 artificial intelligence chips to China.
The communist Chinese regime responded on September 1, condemning the curb on Nvidia's exports and demanding its reversal. "China firmly opposes this," said Shu Jueting, a CCP spokesperson. "The U.S. side should immediately stop its erroneous practices ... and do more to contribute to world economic stability."
China is heavily reliant upon the Republic of Taiwan's high-end chips, as it is presently unable to produce them despite spending heavily in an effort to develop the capability.
The U.S. is similarly reliant. Recognizing its dependency on foreign supply, in 2021 Congress included incentives in its national defense bill to establish leading-edge chip manufactures stateside.
Last month, President Joe Biden signed the CHIPS Act into law, sending $53 billion to U.S. chip manufacturers. In August, Intel announced it would spend over $20 billion building chip factories in Ohio. The trouble is that, as with the Intel factories, American chip manufacturers generally will not be able to ramp up production for years.