In cutting-edge Microsoft data centres, racks of chips used to train AI models sit idle. “The biggest issue we are now having is not a compute glut, but it’s power,” said Microsoft’s chief executive Satya Nadella during a recent podcast interview. The topic has been top of mind in a year when big tech “hyperscalers” — Amazon, Google, Meta and Microsoft — have set out plans to spend more than $400bn in capital expenditure. That outlay, mainly on data centres, has triggered market fears of an AI-fuelled bubble. Big tech groups have been undeterred, arguing that demand outstrips supply. Hyperscalers are seeking to power their own workloads, but also supply Anthropic, OpenAI, xAI and other AI developers looking to train large language models (LLMs). These models underpin a host of tools that promise to transform industries. Big tech stocks have risen as a result, but if computing supply is constrained by a lack of power, the AI “bubble” could deflate.

Full visual story : A lack of electricity for new data centres could deflate the AI ‘bubble.’