SoftBank Teams Up with Tech Giants to Build $500 Billion Super AI Data Center

Super Project: A $500 Billion AI Computing Ambition

SoftBank Group is turning an unprecedented AI infrastructure vision into reality. According to Sina Finance, SoftBank's planned mega AI data center project in Ohio, USA, is advancing rapidly, with total investment potentially reaching $500 billion. Once completed, it will become one of the world's largest artificial intelligence data centers.

The figure is staggering — $500 billion exceeds the annual technology budgets of most nations and signals that AI infrastructure construction has officially entered the era of "super projects." The facility's total power capacity could reach 10 gigawatts (GW), a scale that surpasses the combined AI data center operations of Google, Amazon, Microsoft, and all other major tech companies.

Tech Giants Race to Secure Space as the Computing Power Battle Intensifies

One of the project's most notable highlights is its powerful appeal to the world's top technology companies. Reports indicate that Google and Microsoft are already in discussions with SoftBank about securing space, while Nvidia, Meta, and OpenAI have also shown strong interest in locking down critical computing capacity at the facility.

This convergence of industry heavyweights is hardly surprising. As the parameter scales of large language models continue to climb, the demand for computing power in AI training and inference is growing exponentially. Whether it is Google's Gemini, Microsoft-backed OpenAI models, or Meta's Llama open-source ecosystem, all face enormous computing power shortfalls. Securing large-scale, centralized computing resources in advance has become a key strategic advantage in the AI race.

For Nvidia, this super data center represents massive GPU chip orders. As the undisputed leader of the current AI chip market, Nvidia's H100, B200, and other chip series are in short supply. A 10-gigawatt-class data center would consume millions of high-end AI chips — a commercial prospect Nvidia is undoubtedly eager to see materialize.

What Does 10 Gigawatts Mean?

To understand the astonishing scale of this data center, one needs an intuitive grasp of what 10 gigawatts of power capacity represents. Ten gigawatts is roughly equivalent to the generating capacity of about 10 large nuclear power plants — enough to power a mid-sized city of several million people. Currently, the world's largest individual data centers typically have power capacities in the hundreds of megawatts range, and the 10-gigawatt plan raises this benchmark by an entire order of magnitude.

Such enormous power demands have also drawn public attention to energy supply and environmental impact. Ensuring a stable, clean power supply will be one of the critical factors determining whether the project can proceed smoothly. In recent years, companies like Microsoft and Google have turned to nuclear energy, geothermal, and other clean energy sources to meet data center power needs. SoftBank's Ohio project will most likely also feature innovative energy arrangements.

SoftBank's AI Strategic Blueprint

For SoftBank Group founder Masayoshi Son, this super data center is a core pillar of his "all-in on AI" strategy. After experiencing the ups and downs of the Vision Fund's investment portfolio, Son has firmly set his sights on the artificial intelligence track. From its controlling stake in ARM to massive investments in AI infrastructure, SoftBank is building a complete AI ecosystem spanning chips, computing power, and applications.

Notably, SoftBank had previously announced plans to invest $100 billion in the U.S. AI sector over the next four years. If the $500 billion data center project includes joint investments from partners, it would mean SoftBank is leveraging even greater capital in an attempt to seize a dominant position in the global AI infrastructure race.

The Global AI Infrastructure Race Accelerates

SoftBank's "super project" reflects the fact that global AI infrastructure construction is entering an unprecedented period of acceleration. Beyond SoftBank, multiple forces are deploying simultaneously:

• Microsoft has announced it will invest over $80 billion in AI data center construction in fiscal year 2025;
• Amazon AWS continues to expand its data center footprint worldwide;
• Google is accelerating the deployment of next-generation AI computing infrastructure in the United States, Southeast Asia, and other regions;
• Several large AI data center projects backed by Middle Eastern sovereign wealth funds are also progressing at full speed.

AI computing infrastructure has become the core battleground for national technology competition and corporate strategic rivalry. Whoever controls computing power controls the discourse of the AI era.

Outlook: Opportunities and Challenges Coexist

The construction prospects for SoftBank's Ohio super data center are exciting, but the project also faces significant challenges. Beyond the sustained commitment of enormous capital, power supply assurance, land approvals, environmental compliance, and supply chain management must all be addressed one by one. Furthermore, the pace of technological iteration in the AI industry is extremely fast, and ensuring that the infrastructure's design can accommodate technological evolution over the coming years or even decades is a question that must be answered.

Regardless, when the world's most elite technology companies are all vying to secure space in the same data center, it sends a powerful signal: the trend toward large-scale, centralized AI infrastructure is irreversible, and this computing power arms race has only just begun.