The sudden resignation of Jensen Huang, co-founder and CEO of Nvidia, has cast a shadow over more than $3 billion (23 billion yuan) worth of artificial intelligence computing projects across China. These critical projects, designed around Nvidia’s advanced H100, H200, and H800 processors, now face significant risks of delays, supply chain disruptions, or expensive redesigns.
Huang's departure comes amidst heightened scrutiny of Nvidia, reportedly linked to a scandal involving a Chinese businesswoman suspected of intelligence ties to Beijing. While an independent investigation by Nvidia's board has found no evidence of company secrets or US national security compromises, the timing exacerbates concerns.
Impact on Chinese AI Development
A comprehensive review of over 150 Chinese procurement notices, investment plans, and corporate disclosures revealed the extensive reliance on Nvidia hardware. These projects span major technology hubs along China's wealthy coast and extend into remote western provinces, indicating a broad demand for high-end AI chips despite stringent US export controls.
Many facilities are already operational or under construction, while others remain in the planning phases. For instance, a data center in northwestern Gansu province was slated to install thousands of Nvidia chips by year-end, though their arrival remains unconfirmed. The estimated $3.2 billion figure encompasses not just the cost of Nvidia chips but also associated infrastructure like buildings and power systems.
Nvidia's Stance and Market Reality
Nvidia has consistently stated that US export controls have effectively barred it from China's AI data-center market, excluding Chinese data-center computing revenue from its financial forecasts. Jensen Huang himself had been a vocal advocate for easing restrictions, claiming last year that Nvidia's market share in China had plummeted from 95% to zero.
However, an IDC report contradicted this, estimating that Nvidia still commanded 34% of Chinese AI-accelerator shipments in the first half of 2026. This suggests that despite Washington's efforts to limit Beijing's access to advanced AI chips, Nvidia technology remains deeply embedded in projects involving local government-backed ventures, universities, and regional computing centers.
Future Uncertainty Amid Geopolitical Tensions
Huang's exit could significantly diminish Nvidia's lobbying power as US lawmakers continue to push for tighter export controls. Washington began restricting advanced Nvidia processors in 2022 and further tightened controls in January, requiring case-by-case review of H200 export licenses for eligible Chinese customers.
The identified Chinese projects have not been directly linked to Megaspeed or the specific allegations against Huang, and Megaspeed has denied illegally diverting Nvidia chips to China. Nevertheless, the unfolding situation underscores the complex interplay between global technology supply chains, corporate leadership, and international geopolitical tensions.