Plans are under way to hyperscale the data centre resource to embrace the demands of AI. But do we have enough resources – power, products and people – to supply this new fleet?
The world’s five biggest hyperscalers plan to shell out almost $700bn this year on AI-related capital expenditure.
Alongside forecasts from Amazon Web Services (AWS), Google, Microsoft, Meta and Oracle, there are standalone plans, including the four-year $500bn Project Stargate, specialist investors such as CoreWeave and Lambda, and – mercurial to a fault – Elon Musk and SpaceX/xAI. But will they all be able to spend the money?
David Cahn, a partner at leading technology investor Sequoia Capital, US, believes that 2026 could fall short of expectations. “My prediction for 2026 is that it will be a tale of two AIs. On the one hand, it will be a year of delays – first in data centre buildouts, many of which will fall behind schedule, and second in the AGI [artificial general intelligence] timeline,” he warns.
The data...