US Intensifies Efforts to Combat Nvidia Chip Smuggling to China Through Malaysia

The U.S. authorities are intensifying their efforts to combat the illegal export of Nvidia chips to China via Malaysia.

Officials suspect that advanced semiconductors are still making their way to China despite export controls. Malaysia’s Minister of Trade, Zafrul Aziz, informed the Financial Times that the U.S. government is demanding strict monitoring of all shipments of high-performance Nvidia chips to Malaysia.

Currently, a task force is being established in the country to enhance regulations in the rapidly growing data center sector. Malaysia has emerged as one of the fastest-growing markets for data centers, attracting over $25 billion in investments over the past year and a half.

This rapid expansion has seemingly caught the attention of U.S. regulators, who are concerned about potential evasion of export restrictions. These concerns appear to be valid, as authorities in Singapore recently arrested nine individuals suspected of involvement in fraudulent sales of servers equipped with Nvidia chips.

The U.S. government is increasingly tightening restrictions on chip exports to China through a new three-tiered licensing system for AI chips. This system grants preferences to 18 close allies, including Germany and Japan, and imposes stringent quantitative limits on approximately 120 countries. China, along with Russia, Iran, and North Korea, will face a complete export ban.

The Trump administration is planning to impose further restrictions, including limitations on maintenance activities conducted by companies like Tokyo Electron and ASML at Chinese semiconductor manufacturing facilities. New measures will also strengthen sanctions against companies such as ChangXin Memory Technologies and SMIC.

For cloud service providers like Microsoft, Google, and Amazon, new regulations now limit AI computing capacity outside the U.S. to 50% of the total capacity. In nations that are not within the group of closest allies, only 7% of computing capacity is allowed per country, although companies can apply for special permits.

Nvidia has strongly opposed these measures, labeling them as «far-reaching interference» and warning that they could jeopardize the U.S.’s leadership in AI development.

[Source]