Nvidia Ties Cloud Partner Status to GPU Tracking in Southeast Asia Amid Smuggling Scrutiny

Nvidia is requiring cloud providers in Southeast Asia to enroll in its Fleet Intelligence monitoring system as a condition for official Cloud Partner status, Bloomberg reported, citing people familiar with the requirements. The move links the chipmaker's public endorsement and preferred access to its most sought-after AI accelerators to participation in a hardware-tracking program that can reveal where individual processors are operating. Nvidia was not immediately available for comment.

Partner benefits conditioned on monitoring

Official Cloud Partner status gives providers a public stamp of approval and priority allocation of Nvidia's highest-demand chips. By making Fleet Intelligence enrollment a prerequisite for new applicants in Southeast Asia, Nvidia creates an incentive for visibility into deployments without imposing a universal tracking mandate on every buyer. Bloomberg's sources said the requirement applies to providers seeking official recognition in the region. It remains unclear whether existing Cloud Partners must enroll to retain their designation, or whether Nvidia will extend the rule beyond Southeast Asia.

What Fleet Intelligence does—and does not—provide

Fleet Intelligence gathers information about Nvidia hardware across participating companies' infrastructure, including the locations of individual AI accelerators. That data could help Nvidia trace where its chips end up after they move through supply chains. However, enrollment does not automatically give Nvidia every customer's location data. Nvidia's earlier work on software-based GPU tracking, previously described by Tom's Hardware, used hardware telemetry and network-latency measurements to estimate processor locations. The company characterized that software as a customer-installed agent for checking hardware health, integrity and inventory. Fleet Intelligence launched nearly a year ago as both a data-center management tool and a way to check chip locations.

The enforcement backdrop

The United States bars Chinese companies from buying advanced AI accelerators without a license. U.S. authorities nevertheless believe billions of dollars of restricted hardware have reached Chinese buyers through third-party networks, with Southeast Asia serving as a key transit point. In March, authorities arrested Super Micro Computer co-founder Wally Liaw on charges tied to an alleged $2.5 billion scheme to route Nvidia GPU servers into China. Separately, federal prosecutors unsealed a $300 million smuggling indictment against Earthmade Computer Inc. Malaysian customs recently seized about $12.1 million of advanced AI chips at the country's main airport while investigating a suspected smuggling network.

Megaspeed International draws congressional attention

The debate over chip tracking has centered on Singapore-based Megaspeed International, which won Cloud Partner status within a week of its founding in 2023 and has faced U.S. and Singapore probes for over a year. The New York Times reported that U.S. officials suspect a top Megaspeed executive of spying for Beijing and using close ties with Nvidia CEO Jensen Huang. Huang called his meetings with Megaspeed "completely normal," and an Nvidia spokesperson said the company is "confident that if the U.S. government believes that any person is a threat to American industry, it would let us know." Huang also stated last week that "we've studied all their deployments of our GPUs and counted every one."

Key House Republicans have asked Huang for a formal briefing. Nvidia has declined to say whether it carried out promised follow-up visits to Megaspeed's data centers. Two regional data-center operators have disputed Nvidia's account of where the chips were located. Megaspeed's main landlord recently dropped the company as a tenant.

Broader restrictions on Chinese access

Alongside the tracking push, Nvidia has taken other steps to restrict Chinese access to its hardware. The company reportedly cut half of the Asian firms on its pre-approved buyer list. Nvidia also told cloud buyers not to rent more than half of their Nvidia computing capacity to Chinese clients. "Longstanding US policy encourages every cloud to build on American infrastructure and to win business everywhere they can," an Nvidia spokesperson told the Financial Times regarding the rental cap.

Open questions

Whether the Fleet Intelligence condition will satisfy U.S. officials remains to be seen. Repeated smuggling cases could give Washington justification for tighter export rules, which would be more damaging than the cost of stricter customer screening today. The trade-off is that stricter monitoring could make Nvidia less attractive to some Southeast Asian cloud providers and complicate sales in one of the fastest-growing AI infrastructure regions. For now, the requirement is reported only for new applicants in Southeast Asia; the scope for existing partners and other regions is unresolved.