More than $3 billion worth of Chinese artificial intelligence projects remain dependent on Nvidia processors despite US export restrictions and the recent resignation of Nvidia CEO Jensen Huang amid regulatory scrutiny.
- Over $3 billion in Chinese AI projects identified using restricted Nvidia processors
- Nvidia CEO Jensen Huang resigned amid a scandal linked to US-China intelligence concerns
- US export controls have drastically reduced Nvidia’s direct presence in China’s AI data centers
What happened
A Reuters investigation uncovered more than $3 billion in Chinese artificial intelligence data-center projects built around Nvidia's restricted AI processors, including models H100, H200, and H800. These projects span China’s wealthy coastal areas to remote western provinces and involve local governments, universities, and regional computing centers. Despite US export controls enacted since 2022, these deployments demonstrate the entrenched role of Nvidia technology within China’s AI infrastructure.
This revelation follows the resignation of Nvidia’s co-founder and CEO, Jensen Huang, amid allegations connected to his ties with a Chinese businesswoman suspected by US authorities of intelligence links. Huang had been an advocate for easing export restrictions, and Nvidia now excludes China data-center computing revenue from forecasts due to regulatory uncertainty. Nevertheless, Nvidia technology persists widely in Chinese AI efforts, highlighting a complex enforcement landscape.
Why it matters
Washington’s export controls aim to stem China’s access to cutting-edge AI processors, critical for advancing artificial intelligence capabilities and data-center growth. Yet, the persistence of substantial projects reliant on Nvidia chips raises questions about the effectiveness and enforcement of these restrictions. Beijing’s AI ambitions continue to depend heavily on American technology, despite official limitations, creating tension between policy intentions and on-the-ground realities.
The exit of Jensen Huang weakens Nvidia’s lobbying power at a time when US lawmakers are pressing for tighter controls. This change may increase uncertainties and operational challenges for Chinese projects using Nvidia hardware, possibly leading to delays, redesigns with domestic chips, or supply disruptions. The situation highlights a broader dilemma for US policy: balancing technological containment with the complexities of global supply chains and market demand in China.
What to watch next
Observers should monitor how US authorities enforce export restrictions going forward, particularly regarding who ultimately uses Nvidia-powered AI computing versus who purchases the chips superficially. Enhanced scrutiny might focus on secondary channels and overall use cases instead of just initial buyers. Developments tied to Nvidia’s governance and corporate responses following Huang's resignation will also be critical in shaping future market dynamics in China.
The progress and viability of existing Chinese AI data-center projects built around Nvidia technology remain uncertain. Some facilities under construction could face supply interruptions or need to pivot to Chinese alternatives. Tracking investment flows, installation completions, and government involvement in regional AI hubs will provide insight into how China adapts to altered access to Nvidia’s hardware and whether domestic chip makers gain ground.