China Bans Foreign AI Chips in State-Funded Data Centers

Nov 6, 2025
China has ordered that state-funded data centers use only domestically made AI chips, effectively barring foreign suppliers such as Nvidia, AMD, and Intel from participating in these projects. The move reinforces Beijing's efforts to reduce reliance on foreign technology and boost local semiconductor production.

The Chinese Government has issued new guidance requiring data centers receiving state funding to use only domestically made artificial intelligence chips. Projects that are less than 30% complete must remove any installed foreign chips or cancel purchase plans, while those further along will be reviewed individually.

The directive affects major U.S. chipmakers, including NVIDIA, Advanced Micro Devices, Inc. (AMD), and Intel Corporation. It covers Nvidia’s H20 processors, the most advanced chips permitted for sale to China under U.S. export rules, as well as higher-end models such as the B200 and H200, which are restricted but still circulate through unofficial channels.

AI data center projects in China have received more than $100 billion in state funding since 2021, meaning most facilities are likely affected. Some projects have already been suspended, including one in a northwestern province that planned to use Nvidia chips. The policy could significantly reduce foreign chipmakers’ market share in China, while strengthening domestic suppliers such as Huawei, Cambricon, MetaX, Moore Threads, and Enflame.

The move aligns with Beijing’s broader strategy to increase chip self-sufficiency amid ongoing U.S. export controls. While it may boost local production, the restriction could also widen the gap between Chinese and U.S. data center computing capabilities, as American firms continue large-scale deployments using advanced Nvidia hardware.

We hope you enjoyed this article

Consider subscribing to one of our newsletters like AI Policy Brief, Silicon Brief or Daily AI Brief.

Free newsletter

AI Policy Brief

Weekly report on AI regulations, safety standards, government policies, and compliance requirements worldwide.

Whitepaper

Tensordyne Napier: What If One Rack Could Do the Work of Nine?

Tensordyne

This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.

Read more
Free, six days a week

Daily AI Brief: the AI news that matters, in your inbox.