The US Considers Restricting Chip Shipments to China

According to the report, with knowledge of the situation, the US authorities are considering strengthening an export control regulation aimed at reducing China’s access to artificial intelligence processors by blocking the amount of computational power those chips may have. This is an effort to preserve leadership in the area of artificial intelligence (AI) and potential military uses. This change comes as a cause of the Biden administration’s release of comprehensive guidelines, which was released in October 2022. 

Restricting China’s semiconductor industry and supporting its own US chip sector through significant subsidies was the main intention. The Wall Street Journal states that the Semiconductor companies, such as Nvidia, and Advanced Micro Devices (AMD), may have their shipments halted to Chinese customers by the US Commerce Department in early July. As a result of these limitations, significant semiconductor makers may face a severe downfall in share prices. 

Samsung Mexico Appoints Karen Goldberg Carrillo as New Marketing Chief of MX division

Continuing with the report

, in after-hours trading, shares of both Nvidia (NVDA.O) and Advanced Micro Devices (AMD.O) have experienced declines of over 2% and approximately 1.5%, respectively. The semiconductor sector companies are encountering significant challenges due to the ongoing trade tensions between China and the Biden Administration. If new restrictions come into effect, they will have a substantial impact on Nvidia’s A800 and H800 graphic unit processing chips. Without a specific US export license, the new limitations may even prohibit the export of Nvidia’s A800.

This limitation will bring two main restrictions on AI processors.

1. Chip intercommunication speed – an essential feature that will call for the chaining together of thousands of chips in AI systems like ChatGPT.
2. Processors’ computational capacity – Nvidia’s H800 chip which was created for the Chinese market, has the equal computational capacity to their worldwide chip at certain settings utilized in AI work, but its chip-to-chip rates are constrained.