The Compute Bill Comes Due
Useful-life assumptions change reported infrastructure costs. They need to be tested alongside utilization, cash spending and the financing behind demand.

Editor's note, September 20, 2026: Links to the primary filings and announcements behind the figures have been added, and the passage on useful-life assumptions now names the companies and years involved and notes that Amazon shortened a subset of its servers from six years back to five in 2025. The argument is unchanged. Further review September 20, 2026: Removed unsourced utilization and comparable-capex claims; specified the depreciation example and distinguished an accounting effect from evidence of misstatement.
Somewhere in the footnotes of every hyperscaler’s 10-K sits a number that matters more than the headline capex figure everyone quotes on earnings calls: the useful life assumption applied to servers and networking equipment. A few years ago, most of that gear was depreciated over three years; Microsoft, for one, carried servers at three years until mid-2020. Several large cloud providers have since stretched the assumption to five or six: Microsoft and Alphabet moved servers to six years in 2023, Meta to 5.5 years in 2025, and Amazon to six years in 2024 — though Amazon then shortened a subset of its servers back to five years in 2025, citing the pace of AI hardware development. Nothing about the physics of a GPU changed. What changed is that stretching the depreciation schedule quietly moves cost out of the current income statement and into the future, raising reported margins relative to a shorter-life assumption; whether that better reflects economic use requires evidence. That accounting choice, more than any model release, is the real story of the AI build-out — because it is the mechanism by which an industry has convinced itself that compute is a durable asset rather than a rapidly decaying one.
The scale of the spending is not in dispute. Our corporate spending comparison records actual quarterly cash purchases for Microsoft, Alphabet, Amazon and Meta. It separates lease payments and company-wide spending from AI-only claims; combining headline guidance without reconciling those definitions would give a misleading total. That capital is being raised and deployed on the premise that the resulting compute will generate revenue for long enough, and at high enough utilization, to earn back its cost. Whether that premise holds turns on three separate variables that get collapsed into one another in most public discussion: how fast the hardware actually degrades in economic (not physical) terms, how much of the installed base is actually busy at any given moment, and how much of the capex is being justified by revenue that is itself circular — vendor financing dressed up as demand.
The depreciation problem is a demand problem in disguise
Start with the hardware. A GPU cluster does not become useless in three years; it keeps computing. What erodes is its competitive value relative to the next generation, which arrives on a cadence that has, if anything, compressed rather than lengthened. A training cluster optimized two years ago for a given architecture is now running workloads at a meaningfully worse cost-per-token than newer silicon, which means the economically rational move is often to retire it early for frontier training and relegate it to inference or lower-margin work — a second life, but at a discount. This is precisely the dynamic that shaped the fiber-optic buildout of the late 1990s: the assets did not disappear, but their claim on future revenue was worth a fraction of what the original financing assumed, and the mismatch showed up first in write-downs, then in restructurings.
The accounting choice to lengthen useful-life assumptions is defensible in isolation — reasonable people can disagree about whether a GPU is a three-year or five-year asset — but it has a specific effect on reported earnings that is worth being explicit about. For a new asset with zero residual value and straight-line depreciation, five years rather than three reduces annual expense by 40 percent. Revising the remaining life of an existing asset is a different calculation. The depreciation scenario makes the new-asset assumptions explicit. On capital bases in the tens of billions of dollars, that is not a rounding error; it is the difference between a segment reporting a healthy operating margin and one reporting a thin one. None of this is fraudulent. It is disclosed, audited, and consistent with GAAP. But it means that a chunk of the profitability story investors are currently pricing into AI infrastructure names is an assumption about asset life rather than a fact about cash generation, and assumptions can be revised — usually downward, usually after the market has already re-rated the stock on the way up.
Utilization is the number nobody wants to publish
The second variable, utilization, is harder to pin down because almost nobody discloses it cleanly. Hyperscalers report aggregate capacity and aggregate revenue, not the fraction of provisioned GPU-hours that actually sell. We do not have a consistent, sourced utilization series for these companies. Capacity reserved between runs or for demand peaks is a possible source of lower returns, but this article does not estimate its share or claim an observed industry-wide rate.
The accounting life, economic life and utilization rate are separate assumptions. A financing case needs to survive plausible changes in each.
This matters because the entire justification for the build-out rests on an assumption of near-full utilization eventually arriving, financed today by capital markets willing to underwrite the gap. If demand for AI services grows more slowly than the capacity being installed — a downside scenario, not a probability estimate established here — the excess capacity does not vanish. It sits on the books, depreciating on an optimistic schedule, waiting for demand that may arrive a generation late, by which point the hardware underneath it is no longer competitive.
The circularity nobody wants to price
The third variable is the one that has drawn the most scrutiny in trade press but still gets underweighted in valuation models: how much of the revenue justifying today’s capex is itself financed by the same capital doing the spending. Chip vendors have taken equity stakes in, or extended credit facilities to, some of the very companies buying their chips. Cloud providers have signed multi-year compute commitments with AI labs that are, in some cases, partially capitalized by investment from those same cloud providers. None of this is new to capital-intensive industries — vendor financing has a long history in telecom and semiconductors — but it does mean that a portion of the demand signal currently justifying continued capex is not independent, third-party demand. It is capital recycling through a short loop, and loops like that tend to unwind quickly once any single participant needs the cash back for something else.
None of this argues that the compute build-out is unjustified in aggregate — the workloads are real, and some fraction of the capacity will find durable, high-margin use. It argues for skepticism about the specific accounting and financing choices currently smoothing the picture, because those choices are the mechanism by which a genuinely uncertain bet on future demand gets reported today as a comfortably profitable one. Depreciation schedules are a claim about the future. This one is being tested in real time, and the bill, when it comes, arrives at the balance sheet rather than the model.
Sources
- Microsoft Form 10-K, fiscal 2021 (server useful life raised from three to four years)
- Microsoft Form 10-K, fiscal 2023 (server useful life raised from four to six years)
- Alphabet Form 10-K, fiscal 2023 (server useful life raised from four to six years)
- Meta Platforms Form 10-K, fiscal 2025 (server useful life raised to 5.5 years)
- Amazon Form 10-K, fiscal 2025 (subset of servers shortened from six to five years)
- Sherwood News, combined 2026 capex guidance of Alphabet, Amazon, Microsoft and Meta (Apr. 30, 2026)
- OpenAI and NVIDIA Announce Strategic Partnership to Deploy 10 Gigawatts of NVIDIA Systems (Sept. 22, 2025)
- Amazon announces $5B Anthropic investment, up to $20B more (Amazon, 2026)