Gartner puts $3.5 trillion of AI data-center spending at risk of write-down by 2030
Gartner says more than $3.5 trillion of AI data-center spending is at risk of impairment by 2030, out of roughly $7 trillion planned. It expects more than 80% of neocloud providers to stop operating independently by then.
The Investor · Invest desk
Key figures from Gartner: a 12% return on invested capital would need $8.5 trillion in cumulative revenue. Gartner forecasts enterprise and consumer AI revenue tied to data centers of no more than $3.6 trillion through 2030. Cumulative outlay is approximately $7 trillion.
| Figure | Value | Note | Claim |
|---|---|---|---|
| Cumulative revenue needed for a 12% return on capital | $8.5 trillion | 5 | |
| Enterprise and consumer AI revenue forecast, at most | $3.6 trillion | tied to data centers, through 2030 | 6 |
| Cumulative outlay, about | $7 trillion | 4 |
What happened
- In its October 14, 2026 report, Gartner said the exposure could climb past $4 trillion by 2032.
- Gartner expects providers to invest nearly $1 trillion in 2026, with another $6 trillion projected between 2027 and 2030.
- Gartner calculates that a 12% return on invested capital on that outlay would need $8.5 trillion in cumulative revenue.
- Its forecast for enterprise and consumer AI revenue tied to data centers tops out at no more than $3.6 trillion through 2030.
- Gartner projects AI-optimized IaaS spending rising from $66 billion in 2027 to $174 billion in 2030.
Why it matters
- decision On Gartner's split, about 86% of the outlay is the 2027-2030 tranche, so the write-down depends largely on orders providers have not yet placed.
- exposure Lenders to and shareholders of neoclouds hold the exposure. A merger changes who owns the GPUs, not what they earn, so the loss falls on whoever funded them at the old valuation.
- contradiction AI-optimized IaaS spending grows about 38% a year on Gartner's numbers, yet many providers still fail to earn back their investments because hardware turns over so quickly.
- constraint Asset-light AI-native tenants limit how much of the revenue reaches infrastructure owners, according to Gartner analyst Will Sommer, so owners' returns can trail the headline revenue.
Put Gartner's two revenue figures side by side and the gap is $4.9 trillion [13], yet the impairment estimate is more than $3.5 trillion [1]. Those are different quantities (a project can miss its hurdle rate and still carry its book value), and the cryptobriefing.com account of the report does not explain how Gartner converts one into the other.
Per dollar of outlay, the hurdle asks for $1.21 of cumulative revenue and the forecast delivers $0.51 [15]. The write-down estimate, for what it is worth, is at least half of the $7 trillion [18]. Revenue and book value are different things, so the match may be an accident.
This is where we'd look to be wrong. The $3.6 trillion is a ceiling in a forecast, and the forecast would have to be 2.4 times higher to clear the 12% hurdle outright [14]. Smaller misses would shrink the write-down without removing it. The cryptobriefing.com summary is careful on the point: it says Gartner is not arguing that AI revenue won't grow, only that it won't grow fast enough to justify the spending [11]. We think the direction holds and the size is the open question, because the $3.5 trillion hangs on one revenue forecast.
On neoclouds, Gartner names cost pressure and a lack of differentiation from the large hyperscalers as the main causes of the expected loss of independence [7]. Every figure here is Gartner's forecast as relayed by one trade-press account, and none is a reported result.
What to watch
- Whether AI-optimized IaaS spending tracks Gartner's path from $66 billion in 2027 to $174 billion in 2030.
- Whether asset-light AI firms begin sharing more of their revenue with the companies that own the infrastructure.
- The first neocloud acquisitions or mergers, and whether the prices paid cover the capital invested.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence45
- Adoption
- Insufficient
- Hype gap
- Insufficient
- Incentives
- Insufficient
- Confidence50
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
Gartner warned that more than $3.5 trillion in AI infrastructure spending is at risk of impairment by the end of 2030.
- [2]
In its October 14, 2026 report, Gartner said the impairment exposure could climb past $4 trillion by 2032.
- [3]
Gartner expects AI infrastructure providers to invest nearly $1 trillion in 2026, with another $6 trillion projected between 2027 and 2030.
- [4]
The 2026 spending plus the 2027-2030 projection brings the cumulative total to approximately $7 trillion by the end of the decade.
- [5]
To earn a 12% return on invested capital on that outlay, Gartner calculates the industry would need $8.5 trillion in cumulative revenue.
- [6]
Gartner projects that enterprise and consumer AI revenue tied to data centers will top out at no more than $3.6 trillion through 2030.
- [7]
Gartner predicts more than 80% of neocloud providers will stop operating independently by 2030, pointing to cost pressures and a lack of differentiation from the large hyperscalers as the main culprits.
- [8]
Gartner projects spending on AI-optimized IaaS will rise from $66 billion in 2027 to $174 billion in 2030.
- [9]
According to the report, AI-optimized IaaS spending still won't be enough for many providers to earn back their investments, largely because hardware turns over so quickly.
- [10]
Gartner analyst Will Sommer noted that many AI-native companies run asset-light business models, which hinders revenue from reaching the companies that actually own the infrastructure.
- [11]
Gartner's forecast is not arguing that AI revenue won't grow; the argument is that it won't grow fast enough to justify what is being spent to support it.
- [12]
If more than 80% of neoclouds lose independence, the likely outcomes are acquisitions, mergers or closures.
- [13]
Gartner's revenue forecast falls $4.9 trillion short of the $8.5 trillion needed for a 12% return.
- [14]
The revenue needed for a 12% return is about 2.4 times Gartner's $3.6 trillion forecast.
- [15]
The 12% hurdle needs about $1.21 of cumulative revenue per dollar of the $7 trillion outlay, and the forecast delivers about $0.51.
- [16]
About 86% of the roughly $7 trillion outlay is the projected 2027-2030 tranche.
- [17]
AI-optimized IaaS spending grows about 2.6 times between 2027 and 2030, roughly 38% a year compounded over three years.
- [18]
The estimated write-down of more than $3.5 trillion is at least half of the roughly $7 trillion outlay.
Get this ledger inside ChatGPT. Ask about any of our stories and the answer brings its figure, each claim’s verdict and the sources. Free, no account.
Sources
1 independent publisher whose own reporting we read for this story.
- cryptobriefing.comGartner warns $3.5 trillion in AI infrastructure spending faces write-down risk by 2030
1 article · October 14, 2026
Topics and entities
Follow any of these and your For You feed starts watching them — no settings page required.
Topics
- AI Infrastructure SpendingFollow
- NeocloudsFollow
- Data Centre EconomicsFollow
Entities
- Gartner Inc.Follow
- Will SommerFollow