China Commits $295 Billion to Nationwide AI Data Center Network
Beijing’s biggest bet on AI infrastructure yet
On June 9, 2026, Bloomberg reported that China is preparing to spend approximately 2 trillion yuan ($295 billion) over the next five years building a nationwide network of interconnected AI data centers. The initiative, drafted by the National Development and Reform Commission (NDRC), represents the largest state-backed AI infrastructure program ever attempted.
State-owned telecom giants China Mobile and China Telecom will operate much of the infrastructure and oversee connectivity between facilities. The plan envisions a computing grid that links data centers across the country, allowing AI workloads to be distributed based on available capacity, energy costs, and cooling requirements.
The strategic logic
The initiative serves two purposes. First, it addresses China’s acute shortage of AI compute capacity. Despite advances in domestic chip development (Huawei’s Ascend 910C, Biren’s BR100), China still lags behind the US in total available AI compute by an estimated 4:1 ratio, according to a March 2026 analysis by SemiAnalysis.
Second, it reduces dependence on US chips. By building massive scale around domestic accelerators — even if they’re individually less powerful than NVIDIA’s latest — China can achieve aggregate compute parity through sheer volume. The NDRC blueprint specifically calls for “maximizing utilization of domestic AI chips” in the new data centers.
What the money buys
The 2 trillion yuan breaks down roughly as follows, according to sources familiar with the draft plan:
- Facility construction and land acquisition: ~800 billion yuan
- Domestic AI chips and servers: ~600 billion yuan
- Networking and interconnect infrastructure: ~300 billion yuan
- Cooling and power systems: ~200 billion yuan
- Software, operations, and talent: ~100 billion yuan
The plan prioritizes regions with cheap electricity and cool climates — Inner Mongolia, Gansu, Guizhou, and Ningxia — echoing the “Eastern Data, Western Computing” initiative launched in 2022 but at dramatically larger scale.
Impact on the AI chip market
If executed, the plan would create massive demand for domestic AI chips. Huawei’s Ascend series, Cambricon’s MLU chips, and Biren’s BR processors would be the primary beneficiaries. NVIDIA, which has been restricted from selling its most advanced chips to China since October 2023, would see its China data center revenue continue to erode.
However, skeptics note that China has announced large infrastructure programs before, and execution timelines often slip. The “Eastern Data, Western Computing” initiative, launched with much fanfare in 2022, has only achieved about 60% of its targeted capacity by mid-2026, according to CCID Consulting.
Global implications
The scale of China’s AI infrastructure investment is reshaping the global compute landscape. If the plan proceeds on schedule, China’s total AI compute capacity could approach US levels by 2030, fundamentally altering the competitive dynamics of AI development.
For US chip companies, the message is blunt: the China market for advanced AI chips is shrinking, not growing. NVIDIA’s CFO Colette Kress acknowledged on the company’s May 29 earnings call that “China data center revenue will remain at current levels or decline” through 2027.
Sources
- Bloomberg, “China Prepares $295 Billion Plan to Fund Nationwide AI Buildout,” June 9, 2026
- SemiAnalysis, China AI Compute Assessment, March 2026
- NVIDIA Q1 FY2027 Earnings Call, May 29, 2026
- CCID Consulting, “Eastern Data Western Computing Progress Report,” May 2026