Crusoe Abandons $1.25 Billion Plan to Power AI Data Centers with Boom Supersonic Turbines

Overview

AI infrastructure company Crusoe has walked away from a $1.25 billion partnership that would have used Boom Supersonic's turbines to power its AI data centers. According to Boom Supersonic CEO Blake Scholl, Boom's newly developed stationary power generation units are no longer part of Crusoe's near-term plans.

This decision marks a significant shift in what had been a closely watched collaboration between the two companies, and highlights the complexity and uncertainty surrounding energy supply choices as AI compute capacity continues to expand.

Boom's Foray into Power Generation

Boom Supersonic is best known for developing supersonic commercial aircraft, with deep engineering expertise in high-performance aviation turbine engines. Adapting that aviation-grade turbine technology into stationary power generation units represents Boom's attempt to extend its business beyond aerospace — leveraging existing turbine engineering capabilities to break into the rapidly growing data center power market.

AI data centers are experiencing explosive growth in power demand. Compared to traditional grid connections, which often face lengthy approval and construction timelines, distributed generation solutions like gas turbines offer flexible deployment and stable power supply, making them an attractive transitional option for many AI infrastructure companies. Boom's stationary generation units were squarely aimed at this opportunity.

Boom Supersonic's flagship project is the "Overture" supersonic airliner, powered by its in-house "Symphony" engine — a pure turbojet with no turbofan, designed for efficient combustion at high-altitude, high-speed cruise conditions. Aviation turbine engines and industrial gas turbines (used for stationary power generation) share the same thermodynamic principles, but their engineering priorities differ significantly: aviation engines optimize for power-to-weight ratio, while ground-based power generation turbines prioritize long-duration continuous operation, low maintenance costs, and high thermal efficiency. Adapting an aviation turbine into a stationary power unit typically requires redesigning the cooling system, lubrication system, and control logic, as well as specialized optimization for grid frequency stability — a technical migration path that is far from straightforward. The industrial gas turbines from established players like GE and Siemens are the product of decades of dedicated engineering, which represents the primary competitive barrier Boom faces as a new entrant.

Why the Plan Was Shelved

Blake Scholl stated explicitly that these stationary power units are no longer part of Crusoe's near-term plans. The original reporting did not disclose the specific reasons behind Crusoe's decision, which could involve cost considerations, technology maturity, delivery timelines, or a broader shift in Crusoe's energy procurement strategy.

For a company like Crusoe, which is focused on AI infrastructure, reliability, cost-effectiveness, and scalability of power solutions are the core decision-making factors. When a partnership worth $1.25 billion is put on hold, it typically signals that after a comprehensive evaluation, the company concluded the solution failed to meet its expectations for cost efficiency or supply certainty.

Implications for the AI Energy Landscape

This episode reflects the increasingly prominent energy bottleneck underlying the AI compute race. As large model training and inference continue to scale, the power gap facing data centers has become an industry-wide challenge, spurring a wave of innovative power solutions — from gas turbines to nuclear, geothermal, and beyond.

However, cross-domain technology transitions are rarely smooth. While aviation turbines and stationary power generation share technical commonalities, they differ significantly in operating conditions, maintenance frameworks, and commercialization pathways. Crusoe's exit serves as a reminder to the industry that emerging power solutions must still pass both market and engineering scrutiny before achieving true large-scale deployment.

For observers at the intersection of AI infrastructure and energy, the outcomes of such partnerships will serve as important benchmarks for assessing the commercial viability of non-traditional power solutions.

The power crisis facing today's AI data centers has structural roots: a large GPU cluster can consume hundreds of megawatts — comparable to the industrial power load of a mid-sized city — while traditional grid expansion can take anywhere from three to seven years from application to commissioning. This supply-demand time gap has fueled a boom in behind-the-meter generation, with gas turbines, small modular reactors (SMRs), geothermal power, and even the restart of retired nuclear plants all under exploration. Tech giants like Microsoft, Google, and Amazon have already signed long-term nuclear power purchase agreements, while AI infrastructure specialists like Crusoe had shown strong interest in fast-deployment options like gas turbines due to their need for quicker build-out timelines. The collapse of this deal suggests that even amid urgent demand, buyers remain cautious about novel power technologies that lack large-scale validation — and that technology maturity and delivery certainty are threshold requirements that outweigh price considerations.