New US Ban To Restrict Access of China’s AI Firms to Chips

The United States is currently exploring the prospects of introducing additional restrictive measures regarding the export of chips necessary for developments in the field of artificial intelligence to China.

New US Ban To Restrict Access of China’s AI Firms to Chips

These measures are being considered in connection with the concerns of the American leadership about the potential use of advanced semiconductors as part of the modernization process of the Chinese military industry.

On Tuesday, June 27, the media reported that the US Department of Commerce may decide to impose a ban on the supply of chips from a number of manufacturers, including Nvidia, as early as July.

The restrictive measures are part of a large-scale US strategy that aims to limit the development of the Chinese technology sector in the field of artificial intelligence. In this case, priority attention is paid to the military industry, which, through the use of advanced semiconductors, can enter a new stage of existence from a technical point of view. But a potentially possible ban will also be a factor of negative impact on the commercial sector of artificial intelligence. In this field, many companies have teams of specialists that cover both China and the United States.

In September last year, export controls were introduced, within the framework of which a measure was implemented to restrict sales to the Chinese side of cutting-edge Nvidia A100 and H100 chips, which are necessary for high-performance computing.

As a kind of response to the fall ban, Nvidia has released a less powerful A800 artificial intelligence chip. This decision created a way to circumvent export restrictions. If the measures currently being discussed come into force, the A800 chip will also be in the zone of strict control. In this case, an export license may be required to send units of the mentioned products to China. This means that the supply of A800 chips will be possible, but subject to certain conditions that could potentially reduce the volume of exports.

Nvidia has not yet commented on the intentions of the American authorities regarding the restriction of semiconductor supplies to China.

Large language models, the most popular among which is currently GPT-4, are developing and forming the need for increasing computing power. Chinese technology companies, taking into account the not minimal level of probability of tightening export restrictions, are stocking up on Nvidia artificial intelligence chips in advance. For example, ByteDance has placed orders for graphics processors from an American manufacturer worth more than $1 billion. This was reported by the Chinese media.

Journalists also claim that the A100 chip, which is prohibited for export, can be purchased on the black market in China. The cost of this microelectronic device in an illegal trading space is 20 thousand dollars. At the same time, at the official level, the chip costs half as much.

The media reports that the US government is currently also considering the possibility of establishing a ban on the rental of cloud services by Chinese companies that operate in the field of artificial intelligence. If this decision is made, firms from China will find themselves in an extremely difficult situation. Cloud services are a way for Chinese companies to circumvent the ban on the export of some chips.

The definition of an AI firm is very broad and belongs to the category of terms that provide a vast space in terms of the possibility of interpretation. In the context of the discussed ban on the use of cloud services, Chinese companies that are not directly related to the sphere that is the target of restrictions may be under attack.

Firms from China, when operating at the global level, often use the services of American cloud providers, rejecting similar offers from Alibaba or Tencent, as Western regulators are showing increasing caution about Chinese servers.

As we have reported earlier, AWS and NVIDIA Collaborate on Next Generation AI Infrastructure.

Serhii Mikhailov

2345 Posts 0 Comments

Serhii’s track record of study and work spans six years at the Faculty of Philology and eight years in the media, during which he has developed a deep understanding of various aspects of the industry and honed his writing skills; his areas of expertise include fintech, payments, cryptocurrency, and financial services, and he is constantly keeping a close eye on the latest developments and innovations in these fields, as he believes that they will have a significant impact on the future direction of the economy as a whole.