hyperscaler
In plain wordsOne of the few giant technology companies — such as Amazon, Microsoft, Google and Meta — that run vast networks of data centers and spend most of the money that builds AI computing.
A company that operates computing infrastructure at 'hyperscale' — fleets of very large data centers designed to add servers, storage and networking smoothly as demand grows — usually as a public cloud provider. In AI discussions the word mainly denotes Amazon Web Services, Microsoft Azure and Google Cloud, often with Meta and sometimes Oracle, whose combined capital spending on AI data centersAI data center, n.: A building or campus packed with AI chips, power equipment and cooling, built to train and run AI models. The biggest use as much electricity as a city., chips and power became a central indicator of the AI boom in 2024–2026. Hyperscalers design their own accelerators (such as Google's TPUsTPU, abbr.: Google's own family of chips built specifically for AI, used to train and run its Gemini models and rented to other companies through its cloud. and Amazon's Trainium), supply computing to AI laboratories in exchange for investment stakes, and increasingly contract extra capacity from neocloudsneocloud, n.: A newer kind of cloud company that mainly rents out powerful AI chips, rather than the full range of computing services that giants like Amazon or Microsoft sell..
Chip stocks rallied after all four hyperscalers raised their capex guidance.
Origin
From 'hyperscale', used in the data-center industry for architectures that scale out across very large numbers of servers; 'hyperscaler' for the largest cloud operators was established by the 2010s and became everyday financial vocabulary during the AI investment surge of the 2020s.
Related
- AI data centerAI data center, n.: A building or campus packed with AI chips, power equipment and cooling, built to train and run AI models. The biggest use as much electricity as a city.
- neocloudneocloud, n.: A newer kind of cloud company that mainly rents out powerful AI chips, rather than the full range of computing services that giants like Amazon or Microsoft sell.
- computecompute, n.: The raw computing power used to train and run AI, usually measured by how many specialized chips are used and for how long. With data and algorithms, it is a main driver of AI progress.
- AI bubbleAI bubble, n.: The worry that investors are pouring far more money into AI companies, chips and data centers than AI will earn back soon, so that prices could crash as in the dot-com bust of 2000.
- circular dealcircular deal, n.: A business arrangement in which an AI chip or cloud supplier invests in, or lends to, a customer that then spends the money buying the supplier's own products.
- StargateStargate, prop. n.: A giant venture announced in 2025 by OpenAI, SoftBank and Oracle to build data centers in the United States for training and running AI, with a pledged $500 billion.
Sources
- Wikipedia contributors. Hyperscale computing. Wikipedia (accessed October 2026). en.wikipedia.org/wiki/Hyperscale_computing