Shanghai Unveils World-First Underwater Wind-Powered Intelligent Computing Cluster

Welcome to SMEOCEAN: Anchor Chain Manufacturer and Supplier

Search Content
Oct 03, 2025

A World First! Shanghai Embarks on a New Era of Underwater Wind-Driven Computing

Recently, the world's first commercial underwater data center project powered by offshore wind power officially launched off the coast of Shanghai. As a new computing power solution for the AI era, this groundbreaking project provides a new reference path for the future development of green, low-carbon computing power, and also marks a new phase of "underwater" development for global intelligent computing infrastructure. 

Underwater + Wind Power: A Breakthrough in Green Computing Power 

Located in the waters near Lingang in the Shanghai Pilot Free Trade Zone, the project plans to build a 24-megawatt deep-sea intelligent computing cluster integrating modular data units, a seawater cooling system, a renewable energy power system, and cross-border data transmission capabilities, AI inference optimization, and edge computing support. The first phase of the project, with a power capacity of 2.3 megawatts, has been designated a national green and low-carbon innovation demonstration project and is expected to officially begin operations this fall.

 

Compared to traditional land-based data centers, this underwater intelligent computing center achieves significant energy optimization—its power usage effectiveness (PUE) is kept below 1.15, cooling energy consumption is reduced by over 70%, and it leverages the marine environment to enable high-density deployment, reducing land use. This provides an alternative to the "intelligent computing land shortage" issue in urban development. 


World First! Shanghai Goes Underwater for Wind-Powered AI


In the era of inference computing power, data center layout is shifting towards intelligence and energy conservation. 

Currently, with the increasing prevalence of engineering deployment of large-scale language models, the frequency of AI inference service calls continues to rise, from medical diagnosis and voice interaction to industrial forecasting. Unlike the periodic computations during the training phase, inference tasks prioritize real-time response, stability, and concurrent processing capabilities, placing new demands on the design logic of computing infrastructure. The traditional energy-intensive and costly "GPU cluster stacking" approach is facing sustainability bottlenecks. The emergence of underwater intelligent computing centers offers a new path for low-energy, high-density, and stable operations in the "inference-first" era. Especially amidst the growing global consensus on green transformation, integrating clean energy (such as wind power) with high-computing platforms has become a key strategy for leading countries and technology companies. 

From the Hainan pilot to the Shanghai upgrade, China's green computing system is rapidly taking shape. 

It is worth noting that the launch of the Shanghai Underwater Intelligent Computing Center is not an isolated case, but rather a comprehensive upgrade following an underwater data center pilot project in southern China. As early as the end of 2022, this pilot project had achieved regular deep-sea deployment and data operation, demonstrating excellent equipment stability and supporting AI inference tasks equivalent to 30,000 high-end terminals concurrently, establishing a world-leading deep-sea data center operation paradigm. The Shanghai project builds on this foundation by further integrating a wind power supply system with a higher-density intelligent cooling design, and is considered a representative example of "Underwater Intelligent Computing 2.0."

Low Carbon + AI: Computing Infrastructure Enters a New Phase of "Green Native" 

With China's release of the "Action Plan for High-Quality Development of Computing Infrastructure (2023-2025)," which explicitly aims to reach 300 EFLOPS of computing power nationwide by 2025 and encourages the exploration of regional collaboration, green deployment, and intelligent scheduling, innovative projects like underwater data centers are becoming a strategic addition to the AI computing ecosystem. Whether for continuous inference of large AI models or for computing-intensive scenarios like 5G, the Industrial Internet of Things, and smart manufacturing, future intelligent computing centers must find the optimal solution across the three dimensions of performance, energy efficiency, and sustainability. Underwater deployment in synergy with wind power may be one of the key paths to this solution. 

Green Intelligent Computing, Heading for the Deep Sea 

The global computing landscape is undergoing a new reshaping. Large AI models are making inference the primary computing power battlefield, and in the context of climate change, green intelligent computing is an inevitable future trend. From land to seabed, and from fossil fuels to renewable energy sources like wind power, tidal power, and solar power, the transformation of intelligent computing infrastructure is no longer simply an "upgrade" but a "reconstruction."

 

China, with its pioneering implementation of underwater data centers, is taking a significant step forward in this global computing revolution.


Related News:

What Factors Affect Mooring Chain Parameter Selection in Seawater?

How do Buoys Anchored in Deep Water ?

World's First Self-Propelled Enclosed Salmon Farming Vessel "Su Hai 1" Delivered in Guangzhou


Send Message