Nvidia Ties Southeast Asian Cloud Partner Status to Chip Location Tracking

The partner designation matters to cloud operators because it gives them priority access to Nvidia's most sought-after chips. By linking it to tracking, Nvidia is using its control over scarce supply as leverage in a region that US officials see as the main route for banned processors to reach Chinese buyers. The company has faced criticism that it has not done enough to stop such shipments.

Fleet Intelligence was launched nearly a year ago as a data centre management tool and a way to check where chips are installed. It is installed by customers, monitors hardware health and inventory, and records the physical location of accelerators inside a facility. Taking part has been voluntary, and the system does not automatically share location data with Nvidia or the US government. Nvidia declined to say whether the new requirement would extend beyond Southeast Asia or apply to companies that are already partners, according to Bloomberg.

The tracking rule follows other steps to tighten control over who gets Nvidia hardware in Asia. Earlier this year the company removed half of the Asian companies from its list of pre-approved buyers, and it has recently told cloud customers not to rent more than half of their Nvidia computing capacity to Chinese clients. Asked about that cap, an Nvidia spokesperson told the Financial Times that "Longstanding US policy encourages every cloud to build on American infrastructure and to win business everywhere they can." Nvidia says it complies with export laws and expects the government to identify bad actors.

US officials believe Chinese companies have obtained billions of dollars of restricted Nvidia processors through smuggling networks operating in Southeast Asia. In March, Super Micro co-founder Wally Liaw was arrested on charges of plotting to smuggle $2.5 billion of Nvidia GPU servers into China, and federal prosecutors have unsealed a $300 million smuggling indictment against Earthmade Computer Inc.

The case drawing the most attention involves Megaspeed, a Singapore-based company that won Nvidia Cloud Partner status within a week of being founded in 2023 and has been under investigation by US and Singaporean authorities for more than a year. The New York Times reported last week that US officials suspect a senior Megaspeed executive of spying for Beijing and of using her ties to Nvidia chief executive Jensen Huang. Huang has called his meetings with Megaspeed "completely normal" and said Nvidia has "counted every one" of the GPUs deployed there. Megaspeed's main landlord has dropped it as a tenant, and two regional data centre operators have disputed Nvidia's account of where its chips were located.

The pressure is also coming from Congress. Senior House Republicans this week asked Huang for a closed-door briefing on espionage concerns. Nvidia has not said whether it carried out promised follow-up inspections of Megaspeed's data centres.

The tracking rule does not cover chips already sold to companies outside the partner programme, and Nvidia has not said whether existing partners in Southeast Asia will have to sign up.