Bloomberg’s investigation revealed that China is building around 40 data centers in Xinjiang powered by tens of thousands of NVIDIA chips despite US export bans, aiming to develop a competitive domestic AI industry. However, challenges such as securing reliable power sources and developing indigenous high-performance chips remain significant obstacles to China’s AI ambitions.
Earlier this year, Bloomberg reporters visited EW, a remote county in Xinjiang, China, as part of an investigation into China’s rapidly expanding data center infrastructure. They discovered that several Chinese companies are constructing around 40 data centers across the desert region. These facilities are planned to be powered by tens of thousands of NVIDIA chips, despite the fact that the United States imposed an export ban on such chips to China in 2022. This development highlights China’s determination to build a robust domestic AI industry.
The primary goal behind these data centers is to develop local computing power capable of competing with major US tech giants like OpenAI, Alphabet, Amazon, and Meta. By establishing this infrastructure, China aims to foster its own AI models and technological advancements, reducing reliance on foreign technology and asserting itself as a global leader in artificial intelligence. The scale of these projects underscores Beijing’s ambition to rival or even surpass the US in AI capabilities.
However, there are significant uncertainties surrounding this ambitious plan. One major question is whether China can create a cheap and reliable power source to support these massive data centers, which require enormous amounts of energy. Additionally, acquiring the necessary computer chips for AI development remains a challenge, especially given the US export restrictions on advanced semiconductor technology. These factors could potentially hinder the pace and success of China’s AI ambitions.
Another critical issue is whether China can manufacture its own high-performance chips independently, without relying on US companies like NVIDIA. Developing domestic semiconductor technology at the required scale and sophistication is a complex and resource-intensive endeavor. Success in this area would be crucial for China to sustain its AI development efforts and reduce vulnerability to international trade restrictions and geopolitical pressures.
Despite these challenges, what is clear is the extent of China’s commitment to advancing its AI capabilities. The massive investment in data centers and infrastructure in remote regions like Xinjiang reflects a strategic push to build a competitive AI ecosystem. This effort signals China’s long-term vision to become a dominant player in the global AI landscape, reshaping the technological balance between China and the United States.