Nvidia Finds Banned Blackwell Ultra Chips in China Telecom Network

Data center operators in the United States may not see an immediate change in their daily work, but the discovery of restricted hardware in China confirms that global chip controls remain porous. This week, Israeli security firm Lava announced it found 32 Nvidia Blackwell Ultra B300 chips inside the network of China Telecom. The state-owned telecommunications provider is a major player in the Chinese market, making the presence of these specific processors a notable event for tech watchers.

The B300 chips are currently barred from export to China under U.S. regulations. Typically, these high-performance AI GPUs are only sold to approved buyers who agree not to resell them to sanctioned entities. The question now is how these units ended up in a Chinese data center despite these strict rules. Previous reports from 2025 indicated that Chinese firms smuggled approximately $1 billion worth of Nvidia AI chips into the country in just three months, with B200 models being the primary target at the time.

Nvidia has responded to the findings by downplaying their significance. A company spokesperson stated that the diversion of controlled products is less than half of one percent based on recent court cases. The spokesperson described this volume as a drop in the bucket compared to the vast amount of domestic compute power China already possesses. Nvidia noted that with tens of millions of GPUs in service in approved countries, the current report demonstrates that smuggling is not a major factor in the market.

Lava identified the chips using Nvidia’s own DCGM Exporter tool, a software designed for monitoring server health and performance. While intended for internal use, this tool can sometimes be accessed from outside the data center network without requiring a password. The exact location of the chips within China Telecom’s infrastructure remains unclear, and it is unknown how they were physically moved into the server racks. Nvidia is currently working on a tool to help locate its chips but has not yet released it.

The B300 is a highly sought-after component, announced in March of last year. It offers 50% more memory and FP4 compute capability than its predecessor, the B200. Nvidia marketed the chip as built for the age of reasoning, citing advanced language models like DeepSeek R1. The chip was originally expected to ship by the end of 2025 and quickly became a target for smugglers. In August of this year, nine people were indicted in Taiwan for illegally exporting B300 GPUs to China, involving a firm with whitelisted status.

The high demand for these chips has driven up prices for older hardware as well. Nvidia A100 servers, which are five years old, have seen their prices triple and can now cost up to $82,000. The pressure has led some buyers to retool gaming GPUs for AI workloads where possible. This latest discovery adds to the ongoing scrutiny of how Nvidia’s silicon moves through global supply chains despite export controls.

The situation highlights the complex landscape of international tech trade. While Nvidia insists the current leak is a non-factor, the presence of B300 chips in a state-owned network underscores the challenges of enforcing export restrictions. As AI infrastructure continues to expand, the race to secure these powerful processors shows no sign of slowing down.