By This Hour AI Desk
Crusoe has abandoned a planned $1.25 billion partnership to deploy Boom Supersonic’s stationary gas turbines at artificial-intelligence data centers, ending an arrangement that had positioned Crusoe as the first customer for Boom’s new power-generation business.
The reported decision matters well beyond a single equipment order. For Crusoe, it marks a shift away from making Boom’s turbines a near-term primary power source at its projects. For Boom, it removes the launch partnership for a business intended to sell stationary versions of technology connected to its planned supersonic aircraft engine. The companies have indicated that the split is about the current fit between Crusoe’s power plans and Boom’s offering, rather than a permanent exclusion of turbines from Crusoe’s future options.
Launch order would have covered 29 turbines
The abandoned plan was reported to cover 29 of Boom’s Superpower turbines, each rated at 42 megawatts, for a stated total value of $1.25 billion. Deliveries had been expected to start in 2027. On those figures, the proposed fleet would have represented a substantial commitment to a power source that Crusoe could deploy alongside the construction of its AI computing campuses.
That commitment now appears to have been withdrawn before the expected delivery schedule began. The available account does not establish whether equipment had been manufactured, whether payments had been made, whether the arrangement contained cancellation provisions, or how the financial effects of the decision would be allocated. It also does not say whether Crusoe will replace the prospective Boom capacity with a specific alternative at any one site.
Boom’s chief executive, Blake Scholl, said in a public statement described in the report that turbines were no longer part of Crusoe’s near-term primary power mix at its Abilene locations and elsewhere. He characterized that change as making a launch partnership unsuitable for the present. The wording points to a business decision shaped by Crusoe’s immediate campus-level energy choices, not an assertion that stationary turbines lack a role in all AI infrastructure plans.
Crusoe separately confirmed that it was no longer doing business with Boom. The company said it would retain flexibility to select power solutions site by site as projects and requirements change. Its stated menu includes turbines, wind, solar, batteries and electricity from the grid. That approach leaves open the possibility of different power configurations at different campuses, rather than treating a single technology as the standard across its portfolio.
Abilene illustrates the split in Crusoe’s approach
The report identifies two Crusoe projects in Abilene, Texas, with distinct power arrangements. Crusoe’s initial 1.2-gigawatt data center there, built for Oracle and OpenAI, is powered by the grid, the company said. A gas-turbine plant at that location serves as backup power rather than the facility’s primary supply.
A separate 900-megawatt Abilene data center planned for Microsoft is expected to use on-site gas turbines. Taken together, the two projects show why Crusoe’s explanation cannot be read as a simple rejection of gas-turbine generation. The company is reported to be using, or planning to use, turbines differently depending on the site. In one case they are described as backup; in the other, they are planned as on-site power.
That distinction is central to the end of the Boom arrangement. Scholl’s explanation concerned Crusoe’s near-term primary power mix. Crusoe’s own position says turbines remain among the technologies it may select. Both accounts therefore point in the same direction: the partnership has ended, but turbines have not disappeared from Crusoe’s stated set of options.
Still, the available information does not identify the criteria Crusoe uses to decide among grid supply, on-site turbines, renewable generation and batteries. It does not specify whether those decisions turn on timing, location, capacity, financing, contractual needs, or other operational factors. Nor does it detail the energy configuration of campuses beyond the Abilene projects discussed in the report. Those omissions limit any conclusion about Crusoe’s broader infrastructure strategy.
Boom loses a prominent first customer
For Boom, the loss carries particular significance because Crusoe had been presented as the first customer for Superpower. The company is best known for developing Overture, a proposed supersonic passenger jet. It later introduced the stationary-power business around a turbine that, according to the report, shares much of its underlying component base with Symphony, the engine being developed for Overture.
The arrangement with Crusoe therefore linked two ambitions: supplying power equipment to large computing facilities and using the stationary business as a commercial path connected to Boom’s aviation-engine work. A launch customer can help define an early product’s expected deployment, scale and market case. Without Crusoe’s order, Boom will need to show that its stated pipeline of other sites can provide an alternative route for Superpower.
Boom said it expects to deliver roughly 250 megawatts of Superpower capacity to other sites in 2027 and is targeting one gigawatt in 2028. Those are company targets, not completed deliveries, and the available reporting does not identify the sites, customers, contracts, financing or production arrangements behind them. The statement indicates Boom is continuing to pursue the business despite the collapse of the Crusoe partnership, but it does not independently demonstrate that the capacity targets will be met.
The report also says Boom had raised $300 million in the prior year, with the stationary-power effort intended to support commercialization of that business. It does not provide sufficient detail to determine how the loss of Crusoe’s planned purchase changes Boom’s funding needs, development timetable, manufacturing plans or work on Overture. Describing the cancellation as a setback is reasonable in the narrow sense that Boom has lost its named launch partner; determining the wider financial consequences would require information that is not available here.
A change in fit, not a complete account of why
Neither company’s reported explanation supplies a detailed account of the negotiations that led to the split. Crusoe said Boom had been a good partner but was not the right fit at present. Boom expressed continued support for Crusoe and left open the prospect of working together if turbines become part of Crusoe’s primary mix in the future. The language from both sides is notably forward-looking and does not allege a dispute over performance, delivery capability or the underlying turbine design.
That restraint matters. It would be inaccurate to infer from the cancellation that the Superpower turbine failed a technical test, that Crusoe has abandoned on-site generation, or that either company has ruled out future cooperation. None of those conclusions is supported by the material available. The confirmed point is narrower: the planned commercial relationship has ended and the specific $1.25 billion order is no longer going forward.
The timing is also only partly clear. The source account places the public acknowledgment after Crusoe’s recent fundraising, but it does not establish when the companies reached their decision, when Crusoe changed its power mix, or whether the change followed a revision to any particular Abilene project. It is similarly unclear whether the companies had a binding purchase agreement, a preliminary commercial arrangement, or another form of commitment. The reported dollar value and turbine count describe the intended plan, but not its legal structure.
For customers seeking AI computing capacity, the immediate practical question is whether Crusoe can bring individual campuses online with the energy arrangements it has chosen. The account provides no indication that the end of the Boom deal has changed the operating status of the grid-powered 1.2-gigawatt Abilene facility or the plans for the separate 900-megawatt Microsoft facility. It would be speculative to claim disruption to either project from the information provided.
For Boom, the next visible test will be whether it names or serves the other sites included in its projected 2027 capacity and advances toward its 2028 target. For Crusoe, the relevant measure will be the energy choices it makes for future campuses and whether turbines regain a primary role at any of them. Both questions remain open.
This report is based on a single secondary account and statements attributed to the companies. The reported end of the partnership, the value and scope of the proposed order, and the companies’ accounts of their energy plans have not been independently corroborated.
For further context on this subject, see AI Error Reportedly Brought U.S. Operation Against Chinese Vessel Close to Launch.
Reporting notes
What is confirmed: Both companies are reported to have confirmed the commercial relationship has ended, while Crusoe says turbines remain among its possible site-specific energy options.
Why this matters: The decision changes Crusoe’s near-term power plans and deprives Boom’s stationary-power business of its initial flagship customer.
What remains unclear: The reporting does not establish the contract structure, cancellation terms, replacement power plan, or effect on Boom’s finances and production schedule. This report is based on one source and has not been independently corroborated.
