Nvidia B300s Found in Chinese State Network

Israeli security firm Lava has identified 32 Nvidia Blackwell Ultra B300 graphics processing units operating within the infrastructure of China Telecom, a state-owned telecommunications provider. The discovery is significant because these specific chips are currently barred from export to China under U.S. regulations. By finding the hardware inside a major state entity, researchers have highlighted the ongoing difficulty in enforcing export controls on high-performance artificial intelligence components, raising questions about how restricted technology continues to reach the hands of sanctioned buyers despite official whitelisting protocols.

The presence of these chips follows a pattern of smuggling that has persisted since 2025. During that year, reports indicated that approximately one billion dollars worth of Nvidia AI chips were smuggled into China over a three-month period. At that time, the B200 model was the primary target, but distributors were already preparing for the arrival of the B300 series. The B300, announced in March of last year, offers substantial performance improvements over its predecessor, including 50% more memory and enhanced FP4 compute capabilities. Nvidia positioned the chip as a tool for advanced reasoning models, making it highly desirable for data center operators seeking to maintain competitive advantages in AI development.

Lava did not determine exactly how the B300 units entered China Telecom’s server racks, nor could they pinpoint their specific location within the company's broader network. However, the firm confirmed their presence using Nvidia’s DCGM Exporter tool. This software is designed for internal use by data center operators to monitor server health and performance. While typically contained within the data center environment, the tool can sometimes be accessed from outside the network, and in some cases, it does not require password authentication. Nvidia has acknowledged working on a new tool to help locate its chips but has not yet released it, leaving a gap in visibility for both the company and its customers.

In response to the findings, Nvidia downplayed the incident, stating that it does not deny the presence of the chips but considers the volume insignificant. A company spokesperson noted that recent court cases have shown the diversion of controlled U.S. products to be less than half of one percent. The spokesperson described this leakage as a drop in the bucket compared to the massive amount of domestic compute power China already possesses. With tens of millions of GPUs in service in approved countries, Nvidia maintains that smuggling remains a non-factor, citing efforts by governments on both sides of the globe to block unauthorized transfers.

The enforcement of these controls has become a complex operation, often involving intermediary firms. In August 2026, nine individuals in Taiwan were indicted for the illegal export of B300 GPUs to China. This legal action occurred shortly after Supermicro confirmed it had fired several employees following an internal investigation into smuggling activities. The cases revealed a sophisticated process that utilized a Taiwanese server trading firm called Flying Tiget Tech, which held whitelisted status for purchasing GPUs. This status allowed the firm to buy restricted chips legally, which were then diverted to unauthorized parties, illustrating the loopholes that continue to exist in the supply chain.

The demand for these high-end chips has also driven up the cost of older hardware. Nvidia’s A100 servers, which are five years old, have seen their prices triple, with units now costing up to $82,000. This scarcity has led some organizations to retool gaming GPUs to handle AI workloads as a temporary workaround. The situation underscores the critical role Nvidia’s silicon plays in the global AI infrastructure, where the line between legal domestic use and illegal export remains increasingly blurred by market pressures and creative logistical solutions.

Looking ahead, the industry faces a continued balance between restricting technology transfer and maintaining global supply chain efficiency. While Nvidia argues that the impact of smuggled chips is minimal, the discovery within a state-owned network suggests that the enforcement mechanisms are not perfect. As new generations of chips are released, the pressure to secure these assets will likely increase, with both governments and private companies seeking better tools to track and verify the location of their most valuable hardware.