Big Tech compute spend jumps to $700B

Microsoft, Alphabet, Amazon and Meta have pushed projected 2026 AI infrastructure spending to about $700 billion, based on company guidance and investor commentary published through May 20. The figure is not a formal industry total from the companies themselves; it is an aggregation drawn from capital-spending plans disclosed around recent earnings and summarized by market commentators. Motley Fool said Alphabet, Meta, Microsoft and Amazon were collectively expected to spend nearly $700 billion on AI data-center infrastructure this year, while Yahoo Finance put the figure closer to $725 billion after the latest updates. (finance.yahoo.com) That rise reflects a sharp step-up from earlier expectations. (fool.com) Motley Fool wrote on May 19 that Microsoft, Alphabet and Amazon alone were tracking to roughly $570 billion in 2026 capital spending, before adding Meta’s separate spending plans to the broader hyperscaler picture. Yahoo Finance reported that the high end of estimates had been around $670 billion before the latest quarterly reports pushed the tally higher. ### Where does the $700 billion number come from? Motley Fool’s March 16 analysis said Alphabet, Meta Platforms, Microsoft and Amazon had “collectively guided toward nearly $700 billion” in 2026 capital expenditures for AI data-center build-outs. Yahoo Finance, writing after the latest quarterly reports on April 29, said the same four companies were now closer to $725 billion. (fool.com) The company-by-company numbers explain the jump. Microsoft told investors it expects calendar 2026 capital spending of about $190 billion, including roughly $25 billion tied to higher memory and component costs, according to both Motley Fool and Yahoo Finance. Amazon said in January its 2026 capital expenditures would approach $200 billion and later said that plan remained largely unchanged, Yahoo Finance reported. (fool.com) Alphabet told investors it expects full-year capex of $180 billion to $190 billion, and Meta raised its 2026 capex forecast to $125 billion to $145 billion. ### Why are these companies spending at this scale? Meta said its higher capex forecast reflected “expectations for higher component pricing this year and, to a lesser extent, additional data center costs to support future year capacity,” according to Yahoo Finance. (fool.com) Microsoft said it remained confident in the return on its investments because of “higher demand signals and increasing product usage,” Yahoo Finance reported. Amazon, Microsoft and Alphabet have also tied spending to cloud and AI growth. Motley Fool said Microsoft’s AI business was running at a $37 billion annual revenue run rate, up 123% year over year. (fool.com) The same report said Google Cloud revenue grew 63% in the latest quarter, while Amazon Web Services rose 28%, its fastest pace in nearly four years. ### Why does Meta’s number stand out? Meta’s 2026 capex range of $125 billion to $145 billion is the clearest sign that AI infrastructure spending is no longer confined to the traditional cloud providers. Yahoo Finance said Meta raised that forecast by $10 billion at both ends after its latest results. (finance.yahoo.com) The user’s source briefing also pointed to a roughly $27 billion Meta data-center investment as an emblem of how concentrated these projects have become. I could verify that Meta entered a $27 billion financing arrangement tied to a Louisiana data-center project in Reuters-republished coverage from October 2025, but that deal predates the current 2026 spending guidance and sits outside the latest earnings cycle. (fool.com) ### What does this mean for compute supply? Yahoo Finance said the updated spending plans showed hyperscalers were “not done raising their investment ambitions,” and another Yahoo Finance item said the combined commitment was driving demand for AI chips and infrastructure while power and cooling were becoming constraints. (finance.yahoo.com) That framing comes from market commentary, not from a joint statement by the companies. For operators that depend on rented AI capacity, the immediate fact is that the biggest buyers are still expanding. Nvidia’s May 20 earnings are the next scheduled checkpoint for investors looking for fresh data on chip demand, component pricing and supply conditions. (newsbreak.com) (fool.com) (finance.yahoo.com)