Power Bottlenecks

Because of Data Centers: AI Requires 6 Trillion Dollars in Annual Revenue

KI Geld
Facebook
X
LinkedIn
Reddit
WhatsApp

To justify investments in data centers, the AI industry must generate around 6 trillion dollars annually by 2031, according to Bain.

The AI industry must achieve annual revenues of around 6 trillion US dollars by 2031 to justify investments in data centers. Management consultancy Bain & Company came to this conclusion in a report published on September 29.

Ad

Bain expects annual spending on AI infrastructure to reach up to 1.5 trillion dollars by 2031. This includes new data centers, additional capacities, and the replacement of GPUs, memory, and networking equipment. If these investments account for roughly a quarter of industry revenue—which Bain considers ambitious but realistic based on cloud providers—a market of nearly 6 trillion dollars per year would be required.

According to Bain, the size and costs of AI data centers double approximately every 12 to 16 months. As an example, citing data from research institute Epoch AI, the report references Meta’s Prometheus data center in Ohio. In 2025, it had a capacity of 600 megawatts at an estimated cost of 24 billion dollars. By 2030, this is expected to reach 9 gigawatts and up to 200 billion dollars.

New Products as the Largest Revenue Block

New products, such as in search, advertising, autonomous systems, and physical AI, are expected to contribute the largest share at around 4.2 trillion dollars. Productivity gains in enterprises, such as in software development, sales, marketing, customer service, and IT operations, are projected to bring in 1 to 1.4 trillion dollars. Consumer services with subscriptions and advertising are estimated by Bain at 200 to 400 billion dollars.

Ad

“The debate today is fixated on worker productivity. The economics of AI infrastructure require trillions in new revenues beyond productivity gains.”

David Crawford, Chairman of Bain’s Global Technology Practice and Lead Author of the Report

Data Centers: Bottlenecks in Power and Skilled Labor

Bain cites network capacity expansion, procurement of GPUs and other components, skilled labor, and public opposition over resource consumption and noise as key challenges in building additional data centers. Governments in regions including the US, the EU, South Korea, Saudi Arabia, and the United Arab Emirates are subsidizing the expansion.

(red)

Ad

Weitere Artikel