The Crusoe Boom turbine deal is no longer moving forward, ending a launch partnership that covered 29 stationary Superpower turbines with 1.21 gigawatts of planned capacity and a value above $1.25 billion. The important consequence is not that Crusoe has abandoned onsite generation; it is that AI data-centre power plans are becoming site-specific portfolios rather than fixed technology bets.
The first disclosure came from Boom founder and chief executive Blake Scholl on 25 September, followed by a direct Crusoe confirmation reported by TechCrunch. Boom’s official product fact sheet documents the original order. Because independent coverage beyond TechCrunch is limited and later summaries trace back to the same disclosure, this package uses a narrow primary-plus-one editorial exception and does not treat rewrites as additional corroboration.
What changed in the Crusoe Boom turbine deal
Crusoe had been named as the launch customer for Boom’s Superpower stationary gas turbine. The plan called for 29 units, each rated at 42 megawatts, with deliveries expected from 2027. Boom positioned the order as more than $1.25 billion of launch demand and as a bridge between its stationary-power business and the engine technology behind its planned Overture aircraft.
Scholl said turbines are no longer part of Crusoe’s near-term primary power mix at Abilene and related sites, so a joint launch no longer made sense. Crusoe told TechCrunch that it remains flexible across turbines, wind, solar, batteries and grid power, but that the Boom partnership is not the right fit today. Neither company described the change as a dispute, and no termination fee was disclosed.
That distinction matters. The verified event is the end of a specific launch arrangement, not a claim that Crusoe will never use gas turbines or that Boom has stopped developing Superpower. Boom says other customers remain in its pipeline and gave separate delivery targets; those targets are company projections, not completed orders verified by this package.
Why AI infrastructure power plans keep moving
A large AI campus cannot choose electricity on headline capacity alone. It must line up construction timing, interconnection, fuel supply, permits, cooling, backup architecture and customer delivery dates. A turbine that is attractive for one location may be late, expensive or unnecessary at another if grid capacity or a different onsite option becomes available.
Crusoe’s own projects illustrate the portfolio logic. TechCrunch reported that the first 1.2-gigawatt Abilene campus uses grid power with gas turbines for backup, while another planned campus for Microsoft is expected to use onsite turbines. The source did not say Boom hardware would power the latter, so this article does not infer a replacement supplier.
For Boom, losing the named launch customer removes a public proof point. A first customer can validate engineering requirements, factory sequencing and lender confidence before broad deployment. The company now has to demonstrate that its claimed pipeline converts into contracted sites on the timetable it has described.
The financing consequence is asymmetric
Crusoe recently announced a large financing round and says it has a broad contracted project base. That gives it flexibility to match power technologies to sites. Boom’s stationary-power business, by contrast, was presented partly as a commercial engine that could support the longer Overture programme. The same termination therefore has different strategic weight for each company.
The package does not assume the full $1.25 billion was recognised revenue, paid consideration or a cancellation loss. Boom’s fact sheet described an order book value, but the companies have not published payment milestones, cancellation terms or accounting treatment. Those details should remain open until a filing or direct company record answers them.
The Crusoe Boom turbine deal ended because the equipment no longer fit Crusoe’s near-term primary-power mix; the larger signal is that AI campuses will select power site by site, making headline equipment orders vulnerable to schedule and design changes.
What operators and investors should watch
The first test is whether Boom names replacement customers and converts its projected 2027 deliveries into identifiable projects. The second is manufacturing progress: the Superpower product shares engineering roots with the Symphony aviation engine, but stationary deployment still requires its own certification, factory, service and reliability evidence.
For Crusoe, watch the generation mix attached to each new campus rather than a company-wide slogan. Grid contracts, turbine orders, battery capacity and renewable power-purchase agreements should be evaluated against the construction schedule and the computing customer’s commissioning plan.
India’s data-centre market faces the same constraint at a different scale. Developers must secure dependable power, land, water and transmission while serving customers that may change chip and rack requirements quickly. The lesson from this termination is contractual: align long-lead equipment commitments with site-specific milestones and preserve alternatives when interconnection assumptions change.
Verified facts and withheld assumptions
| Item | Verified position |
|---|---|
| Original order | 29 Superpower turbines |
| Unit rating | 42 MW |
| Planned total | About 1.21 GW |
| Headline value | More than $1.25 billion |
| Status | Launch partnership no longer proceeding |
| Not disclosed | Termination fees, payments made, replacement supplier |
For background, Lapaas Voice’s Crusoe funding analysis explains the balance-sheet expansion behind its campus pipeline. The AI Energy Management Alliance report covers flexible-load approaches, while the California data-centre laws analysis shows how grid and cost obligations can reshape site economics.
The next contractual signal will be whether the parties formally restate the old order in their own investor or product materials. Until that happens, readers should separate the verified end of the launch partnership from assumptions about refunds, penalties or factory utilisation. Lapaas Voice will treat any later filing as a dated update rather than resetting this disclosure’s 25 September freshness.
Frequently asked questions
Did Crusoe cancel all turbine plans?
No. Crusoe said the Boom partnership is not the right fit today and described a mixed strategy that can still include turbines at other sites.
How large was the original order?
Boom’s fact sheet said 29 turbines at 42 megawatts each, totalling about 1.21 gigawatts and valued above $1.25 billion.
Why did the partnership end?
Boom’s CEO said its turbines were no longer part of Crusoe’s near-term primary-power mix, so the launch arrangement no longer made sense.
What remains unknown?
The companies have not disclosed termination fees, payments already made, replacement suppliers or the detailed economics behind the change.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
