China's Ambitious Plan to Create a National Computing Power Network Faces Challenges

China's Ambitious Data Center Initiative

China is embarking on an ambitious project to create a nationwide network for selling surplus computing power from its data centers. This initiative comes in response to a significant overcapacity issue that has emerged following a rapid expansion of data center construction across the country 1

1

2

2

3

3

.

The 'Eastern Data, Western Computing' Strategy

The surge in data center development was initially driven by China's 'Eastern Data, Western Computing' strategy, launched in 2022. This plan aimed to concentrate data center facilities in western regions, where energy costs are lower, to meet the computing demands of the eastern economic zones 1

1

2

2

. The Ministry of Industry and Information Technology (MIIT) has licensed at least 7,000 computing centers as part of this initiative 2

2

.

Overcapacity and Underutilization

Despite the ambitious plans, many of these data centers are now operating well below capacity:

• Utilization rates are estimated to be around 20-30% 2

  2

• Over 100 data center projects have been cancelled in the past 18 months 1

  1

  2

  2

• State investment in data centers surged to 24.7 billion yuan ($3.4 billion) in 2024, up from 2.4 billion yuan in 2023 2

  2

The National Computing Power Network

To address these issues, Chinese authorities are now working on a plan to create a centralized cloud platform. This platform aims to:

1. Aggregate unused computing power across the country
2. Offer computing capacity as a service to users throughout China
3. Optimize existing resources and regulate future growth 1

   1

   2

   2

The Ministry of Industry and Information Technology is collaborating with state-owned telecom providers China Mobile, China Telecom, and China Unicom to develop this platform 1

1

2

2

.

Technical Challenges

The implementation of this national computing network faces several significant hurdles:

1. Latency Issues: Many western-region centers fail to meet the 20-millisecond latency standard set by the government, which is crucial for real-time applications 1

   1

   2

   2

   .

2. Hardware Diversity: Data centers use a variety of computing hardware, including Nvidia GPUs with CUDA software and Huawei's Ascend accelerators with CANN software, making integration challenging 1

   1

   3

   3

   .

3. Real-time Power Transfer: The technology for efficiently transferring computing power to users in real-time is still underdeveloped 2

   2

   .

Regulatory Measures

To address these challenges and prevent further resource wastage, China's National Development and Reform Commission (NDRC) is taking several steps:

• Conducting a comprehensive review of the sector
• Imposing new restrictions to prevent overbuilding
• Banning small-scale computing infrastructure funded by local authorities
• Setting specific thresholds for new projects, including minimum utilization requirements 1

  1

  2

  2

Future Outlook

Despite the challenges, Chinese officials remain committed to the national computing network. They believe that centralized scheduling and orchestration of computing power will support China's ambitions in AI and cloud services 1

1

. However, analysts caution that integrating diverse hardware is a complex endeavor that will take time and may not fully deliver on its promise 1

1

3

3

.

As China targets standardized interconnection of public computing power nationwide by 2028, the success of this initiative could have significant implications for the country's technological competitiveness and its race for AI supremacy with the United States 2

2

3

3

.