Bain says AI needs $6 trillion a year in revenue by 2031 to justify its data centres

Started by Darren, Today at 07:08 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Topic: Bain says AI needs $6 trillion a year in revenue by 2031 to justify its data centres   Views(Read 64 times)
Active members in this topic:
Darren(1)

Darren

Bain & Company has put a number on the question everyone keeps asking about the AI boom. According to its latest report, the industry will need around $6 trillion in annual revenue by 2031 to justify all the data centres being built. Existing AI products are only expected to bring in $1.2 to $1.8 trillion a year by then. That leaves a gap of roughly $4.2 trillion that nobody has yet worked out how to fill

The spending side is enormous. Bain expects annual infrastructure spending could hit $1.5 trillion by 2031, covering new data centres, chips, memory and networking. The five biggest spenders, Microsoft, Google, Amazon, Meta and Oracle, are projected to put $780 billion into capital expenditure this year alone, almost five times what they spent three years ago. The largest data centres are doubling in size and cost every 12 to 16 months

Bain's model has $5 to $6.5 trillion being spent through 2030 to build almost 150 gigawatts of new capacity, which would nearly triple global AI capacity in five years. Power supply, chip availability, skilled labour and planning permission are all listed as bottlenecks. Anyone who followed our threads on underwater data centres and India's data centre boom will recognise those problems

So where could the missing money come from? Bain suggests ad supported chatbots could bring in $100 to $200 billion, autonomous vehicles and industrial automation around $400 billion and physical AI, such as robots, digital twins and simulations, about $900 billion. The rest would have to come from applications that do not exist yet. David Crawford, who chairs Bain's technology practice, said the economics demand trillions in new revenue beyond productivity gains

This fits neatly with the $10 trillion buildout piece we discussed last week and the huge compute commitments in Anthropic's prospectus. Either AI becomes one of the biggest industries in history very quickly, or a lot of investors are going to be disappointed. Which way do you think it goes?


Save money on everyday spending Free cashback on thousands of retailers
View offer