“As I went into in last week’s premium (and per my Bastard data center model), a 100MW data center costs around $4 billion to $5 billion, with gross margins of 28% at $17.7 million a megawatt…with negative gross margins below $12.10 a megawatt.”
Our reading, the sentence that was checked:
Per the author's model, a 100MW data center costs around $4–5 billion, with 28% gross margins at $17.7 million a megawatt and negative gross margins below '$12.10 a megawatt' (as written), but only with a customer the entire time and no debt.
Assessment: Unresolved Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, Dead Money, paragraph 126.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment wyea-dead-money-42b506-a2
We recorded 50 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-28. The source may have been edited after 2026-09-29. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Observation
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No horizon
- Deadline state on 2026-10-29
- No horizon (the release’s reference date)
Conditions
- per author's model (qualifier, stated by the speaker)
- only with a customer the entire time and no debt (qualifier, stated by the speaker)
- around (qualifier, stated by the speaker)
From the speaker
- per my Bastard data center model
- a 100MW data center
- around $4 billion to $5 billion
- gross margins of 28%
- $17.7 million a megawatt
- negative gross margins below $12.10 a megawatt
- a customer the entire time
- you don’t have any debt
Our interpretation
- This is a model-based estimate, not a reported realized margin for a named operator.
- I use September 1–29, 2026 for the current cost and model assertions repeated in the post. This does not supply a missing revenue period.
- The surrounding GPU discussion suggests an equipped AI facility; shell-only construction estimates are not directly comparable.
- I preserve '$12.10 a megawatt' as written. I do not silently insert million or an annual revenue period.
- Continuous customer occupancy and no debt are scenario assumptions, not a conditional forecast requiring a subsequent event.
- Estimate denotes the author's model basis, not GAAP or audited company accounts.
Counts as supported if: Accessible model inputs and reproducible calculations yield approximately $4–5 billion in total cost and the stated margins under the specified occupancy and debt assumptions, with independent evidence supporting comparable inputs.
Counts as contradicted if: Reproducible calculations or sufficiently comparable independent measurements materially conflict with the stated costs or margins under the same units, cost coverage and scenario assumptions.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Unresolved
The record was checked, and the evidence available by the cutoff does not settle it.
Scored on a plain reading. The speaker named no measure for this claim, so it was judged on the reading an informed reader would take, stated below under our interpretation. Both reviews had to find that reading fair.
- Evidence as of
- Deadline state when assessed
- No horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
JLL's January 5, 2026 outlook separates shell-and-core construction from technology fit-out and estimates that AI technology fit-out can cost up to $25 million per MW [e9]. This component estimate does not verify the claimed $4–5 billion total cost or either gross-margin figure for the author's continuously occupied, debt-free 100MW scenario [e9]. The model claim remains unresolved.
Limitations
- The complete model and arithmetic are unavailable in the packet.
- The revenue period, gross-margin cost treatment and IT-load versus total-power basis remain unspecified.
- The literal '$12.10 a megawatt' remains unresolved; inserting 'million' would change the supplied claim.
- JLL's January outlook does not provide a matching September 1–29, 2026 measurement for the model's equipment and cost coverage.
Evidence (1 item)
Context only: JLL,
$25 million per MW
Why it matters here: JLL separates technology fit-out from shell-and-core construction. This establishes why a building-only benchmark cannot directly test an equipped AI facility, but does not verify September costs or either margin figure.
- Where
- Data center construction costs have been increasing at 7% CAGR, note below construction estimates
- Publisher
- JLL
- Kind of source
- Independent measurement
- Measures
- AI infrastructure tenants covered by JLL's industry estimate, Technology fit-out cost, excluding shell-and-core construction, USD USD millions per MW, estimate, 2026 to 2026
- Describes events of
- Public from
What would settle it
The complete dated model with explicit revenue units and period, capacity basis, equipment coverage, depreciation and gross-margin formula. Reproducible calculations and independent comparable inputs must test the stated costs and margins under continuous occupancy and no debt, with clarification of the literal $12.10 figure.
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, Dead Money
As I went into in last week’s premium (and per my Bastard data center model), a 100MW data center costs around $4 billion to $5 billion, with gross margins of 28% at $17.7 million a megawatt…with negative gross margins below $12.10 a megawatt.
- Where
- paragraph 126
- Published
- 2026-09-29
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-dead-money-42b506-a2: Current. Outcome: Unresolved. Evidence as of 2026-10-28. Public date not recorded.
Recorded events
Assessment publishedBy this site
First assessment published after both reviews and the final-text check passed.