China Launches World's First Commercial Offshore Wind-Powered Underwater Data Center
Key Takeaways
- ▸China's underwater data center operates at commercial scale with ~2,000 GPU servers sealed beneath the ocean, using passive seawater cooling instead of traditional chillers
- ▸Achieves a PUE rating of 1.15 versus industry average of 1.5, demonstrating significant energy efficiency gains from natural ocean temperature stability
- ▸Directly integrates offshore wind farms for renewable energy, reducing dependence on conventional grid-based electricity and addressing AI infrastructure's growing power demands
Summary
China has begun commercial operations at an underwater data center housing nearly 2,000 servers approximately 35 meters below the ocean surface near Shanghai. The $226 million facility, developed jointly by HiCloud Technology and Chinese authorities, represents a novel approach to addressing the cooling and energy demands of AI infrastructure scaling. The installation combines offshore wind generation with subsea computing infrastructure, using naturally stable ocean temperatures for passive seawater cooling instead of traditional industrial chillers. Server modules from China Telecom and LinkWise handle AI workloads, 5G services, and large-scale data annotation operations across the 24-megawatt facility. The underwater data center achieved a Power Usage Effectiveness (PUE) rating of 1.15—significantly lower than the industry average of 1.5—indicating that electricity supports computing tasks more efficiently when cooling overhead is minimized. This project reflects China's broader strategy to integrate renewable energy generation directly into digital infrastructure while reducing national power grid pressures driven by AI expansion.
- Hardware sealed in pressure-resistant modules with remote monitoring minimizes maintenance complexity, though long-term durability of subsea infrastructure remains an operational concern
- Positions underwater computing as a potential solution to global data center cooling bottlenecks that increasingly constrain AI model deployment and training
Editorial Opinion
This project exemplifies how geographic and engineering innovation can tackle the energy bottleneck in AI infrastructure—seawater cooling and offshore wind integration are elegant solutions to real scaling constraints. However, the practical advantages over land-based facilities may be overstated; underwater deployments trade instant hardware accessibility and maintenance agility for marginal efficiency gains. Real impact depends on whether similar projects can operate reliably at scale and whether the environmental costs (submarine cables, pressure sealing materials, restricted expansion) offset the energy savings.
