Google Cloud posted $24.8 billion in Q2 2026 revenue, up 82% year-over-year, beating the $22.3 billion consensus. The headline number matters less than what it rhymes with: NVIDIA's Data Center segment grew 92% over the same period. Two companies, one renting AI compute by the hour and one selling the chips outright, are now tracing the same growth curve because they sit downstream of the same demand signal.
The backlog tells you where this goes. Google Cloud's remaining performance obligations hit $514 billion, up 385% from $106 billion a year ago, with more than $50 billion added in a single quarter. Across the four largest cloud providers, contracted demand now exceeds $2 trillion. Operating margin expanded from 20.7% to 35.6% year-over-year, closing the gap with AWS to roughly two points from nineteen a year ago. CFO Anat Ashkenazi confirmed TPU hardware sales began in Q2, but flagged the dollars were small and the bulk lands in 2027, meaning the 82% growth is almost entirely organic compute rental.
Google raised full-year 2026 capex guidance to a range of $195 billion to $205 billion and signaled capex will 'increase significantly in 2027.' Amazon, Microsoft, Meta, and Oracle are on a parallel trajectory, putting industry AI infrastructure spend well past $1 trillion next year. Read the full piece for the margin comparison table and the growth-rate convergence charts, which make the NVIDIA parallel impossible to dismiss.
[READ ORIGINAL →]