Fast Technology News, November 7 — NVIDIA is tightening requirements for Southeast Asian cloud service providers applying for official partner status. Applicants are now required to enroll in its Fleet Intelligence monitoring system to track the deployment of artificial intelligence chips.

After reviewing relevant reports, Fast Technology learned that joining the monitoring system has been listed as a prerequisite for applying for NVIDIA cloud partner status. This qualification not only brings public recognition from NVIDIA but also gives cloud service providers priority access to scarce chips.

Fleet Intelligence was initially designed mainly for data center management, collecting GPU deployment and operational data to help operators understand equipment usage. According to the report, the system can also collect geolocation information from individual accelerators to track where chips are deployed after delivery.

This change comes against the backdrop of a U.S. investigation into AI chips subject to export restrictions being transshipped to China via Southeast Asia. NVIDIA hopes to use partner qualification as leverage to encourage more customers to join the monitoring program.

Currently, the requirement applies to cloud service providers in Southeast Asia applying for official qualification. In other cases, the monitoring program remains voluntary. Once users join the system, they are also not required by default to share location data with NVIDIA or the U.S. government.

NVIDIA has not yet clarified whether existing partners must enroll in order to retain their qualification, or whether this requirement will expand to other regions.

In previous software documentation, NVIDIA stated that this service requires customers to install it themselves and can monitor GPU utilization, power consumption, temperature, configuration, and errors. It provides read-only data and cannot modify GPU configuration or operating status. The company also stated that its GPUs do not contain hardware tracking devices, remote kill switches, or backdoors.

Singapore-based company Megaspeed, which is under investigation, is one example in the related controversy. Jensen Huang recently said that dealings with the company were completely normal and that NVIDIA had studied its GPU deployment and counted the devices one by one. Megaspeed has also denied any wrongdoing.

Regarding the new requirements for partner qualification, NVIDIA did not respond to requests for comment in the report. Whether existing partners need to complete registration retroactively remains to be clarified.