Cloud Computing Watch: Google Cloud's Growth Rate Show That Enterprise AI Workload Wins, Not Raw Infrastructure Scale

The CODEW Cloud Computing Watch | September 1, 2026

Google Cloud's Growth Rate Just Lapped Both of Its Rivals — and the Bill for AI Spend Is Catching Up With Everyone Else

Executive Summary

Two Numbers That Define Cloud Right Now: 82% and 72%

Google Cloud's most recent quarter posted 82% year-over-year revenue growth, reaching $24.8 billion and tripling operating income, with backlog swelling to $514 billion — an acceleration from 63% growth just one quarter earlier and a pace that now outstrips both Azure's roughly 40% and AWS's roughly 28%. It's the clearest evidence yet that enterprise AI workload wins, not raw infrastructure scale, are now the variable moving hyperscaler growth rates.

The other number worth sitting with: 72% of global companies exceeded their allocated cloud budgets in the last fiscal year, according to the FinOps Foundation's sixth annual survey of 1,192 practitioners representing more than $83 billion in annual cloud spend. The same survey found the share of FinOps teams actively managing AI-specific spend has gone from 31% to 98% in two years — the fastest adoption curve of any cost category in enterprise IT. Put together, the two numbers describe the same market from opposite sides: hyperscalers are winning AI workloads at an accelerating pace, and the enterprises buying that capacity are still building the cost discipline to keep up with what they're spending.

01 — The Cloud Lead

Google Cloud's Acceleration Is Reshaping the Hyperscaler Pecking Order

Google Cloud's jump from 63% to 82% year-over-year growth in a single reporting period is not a rounding-error blip — it's the sharpest quarter-over-quarter acceleration any of the three major hyperscalers has posted this cycle. Combined with a backlog that's swelled to $514 billion, the numbers suggest enterprises signing new, large multi-year AI commitments are increasingly choosing Google Cloud over its rivals, not just adding it as a secondary provider inside an existing multicloud setup.

AWS remains the largest provider by absolute revenue — still roughly double Google Cloud's quarterly total — and growing 28% off a much larger base adds more raw dollars than Google Cloud's 82% off a smaller one. But the trend line matters as much as the snapshot: analysis of net-new annual recurring revenue additions since GPT-4's launch has repeatedly shown Microsoft Azure and Google Cloud outpacing AWS in the specific category of new enterprise AI workload wins, a pattern this quarter's numbers reinforce rather than reverse.

02 — The FinOps Reckoning Behind Every Cloud Growth Number

Every dollar of hyperscaler growth described above is landing on an enterprise cost-management function that, by its own admission, is still catching up. The FinOps Foundation's 2026 survey found 72% of global companies exceeded their cloud budgets in the last fiscal year, and 44% still report limited visibility into their cloud expenditure despite using native or third-party cost-management tools — figures that haven't meaningfully improved despite years of FinOps investment, according to the survey's own authors.

The gap is specifically an AI problem, not a general cloud-cost problem. CloudZero's State of AI Costs research found average monthly enterprise AI spend climbing from roughly $63,000 in 2024 to a projected $85,500 in 2025, a 36% jump, and only 53% of enterprises report having extended their FinOps frameworks to actually govern GPU and AI-pipeline costs specifically. IDC's FutureScape 2026 warns large organizations face up to a 30% rise in underestimated AI infrastructure costs by 2027 — driven, IDC argues, by systematic under-forecasting rather than reckless spending, which is a meaningfully different problem to solve.

03 — The Neocloud Utilization Gap Nobody's Fully Closed Yet

The specialist GPU-cloud tier this series has tracked continues to show the same pattern: enormous contracted capacity that hasn't yet turned into delivered, billable infrastructure. CoreWeave's most recent quarter reported 1.5 gigawatts of active power against more than 4.2 gigawatts of contracted power — meaning roughly two-thirds of its committed future capacity remains unbuilt. Nebius, meanwhile, posted 684% year-over-year revenue growth with capacity described as "sold out" and reportedly four or more customers competing for every available GPU tranche. Neither company has fully closed what the industry calls the "MFU gap" — the difference between compute capacity and actual utilization, which commonly runs 30-40% even at well-run facilities — and that gap, more than raw GPU supply, may be the more fixable near-term constraint on how much of this contracted capacity actually reaches paying customers this year.

Three Cloud Signals

1. The hyperscaler growth-rate hierarchy has flipped. Google Cloud's acceleration to 82% growth, ahead of Azure's ~40% and AWS's ~28%, shows enterprise AI workload wins — not raw scale — are now the variable determining which hyperscaler grows fastest in any given quarter.
2. Enterprise cost discipline is structurally behind enterprise AI adoption. With 98% of FinOps teams now managing AI spend but only 53% having extended governance to GPU and pipeline costs specifically, the tooling and forecasting gap — not carelessness — is the primary driver of the 72% budget-overrun rate.
3. Contracted neocloud capacity and delivered capacity remain two very different numbers. CoreWeave's 1.5GW active versus 4.2GW contracted gap is a reminder that headline capacity commitments across the sector overstate near-term available compute until power and utilization catch up.

The CODEW Take

Google Cloud's 82% growth number is the headline, but the more durable story this week is the widening gap between how fast enterprises are committing to AI cloud spend and how well-equipped they are to govern it. A hyperscaler posting record acceleration and a FinOps community that's gone from managing AI costs at 31% of organizations to 98% in two years are two sides of the same adoption curve — and the fact that 72% of companies still blew their cloud budgets last year, even with that governance maturity racing to catch up, suggests the spend is outrunning the discipline built to control it, not the other way around.

The neocloud utilization gap adds a second layer to the same story: even where capacity has technically been contracted, a meaningful share of it isn't yet delivering billable compute. What to watch next: whether Google Cloud's growth rate holds or reverts toward Azure's pace over its next two reporting cycles, and whether FinOps tooling vendors can close the AI cost-governance gap before another full budget cycle passes.

---

Editorial Note: Cloud Computing Watch is The CODEW's recurring intelligence series tracking cloud infrastructure, hyperscaler economics, and the specialist providers reshaping how AI compute gets bought and sold.