Crusoe scraps $1.25 B turbine deal with Boom Supersonic, pivots to mixed‑energy data centers

Crusoe, the Denver‑based AI data‑center operator that recently closed a $3.9 billion funding round, announced that it will no longer buy Boom Supersonic’s Superpower turbines. The decision ends a $1.25 billion contract for 29 units that would have supplied up to 1.2 GW of on‑site power.

Key points

• Crusoe canceled a $1.25 billion order for 29 × 42‑MW Superpower turbines.
• Boom Supersonic will still deliver about 250 MW of turbines to other customers in 2027.
• Crusoe’s Abilene campus currently runs on grid power with a backup gas turbine.
• The company is building a 900 MW site for Microsoft that will use on‑site gas turbines.
• Both firms remain amicable, with future collaboration possible if Crusoe’s power mix changes.

What happened

In a Friday post on X, Boom Supersonic CEO Blake Scholl confirmed that the launch partnership with Crusoe had dissolved. He wrote, "The TL/DR is that turbines are no longer part of Crusoe’s near term primary power mix at Abilene/etc., so a launch partnership just didn’t make sense." Boom plans to ship roughly 250 MW of Superpower turbines to other sites next year and aims for 1 GW by 2028.

Crusoe responded by email, noting that its strategy emphasizes flexibility: "We build AI factories from the power up, and we’re bringing new campuses online across the country, powered by innovative energy sources," said spokesperson Andrew Schmitt. He added that while Boom has been a good partner, the turbine deal no longer fits Crusoe’s immediate needs. The company’s flagship 1.2‑GW Abilene campus for Oracle and OpenAI remains grid‑powered, supplemented by a backup gas turbine, while a separate 900‑MW Microsoft project will rely on on‑site gas turbines.

The backstory

Crusoe launched in 2018 as a bitcoin mining operation that captured excess natural‑gas flare from oil fields. Over the past few years, the firm pivoted to AI workloads, scaling its data‑center footprint and attracting marquee customers such as OpenAI and Microsoft. Boom Supersonic, known for its Overture supersonic jet, spun off a stationary‑power business in 2023, repurposing the jet’s Symphony engine into a 42‑MW gas‑turbine product called Superpower.

The two Denver companies signed a deal that would have made Crusoe the first customer for the new turbine line, with deliveries slated to start in 2027. Boom raised $300 million in 2025 to accelerate the commercial launch, counting the Crusoe contract as a cornerstone.

Why the partnership fell apart

Crusoe’s rapid campus expansion demands a power mix that can be deployed on varied timelines and regulatory environments. Grid access, renewable credits, and local emissions rules differ across sites, making a single turbine solution less attractive. The company’s statement that turbines are “not part of Crusoe’s near term primary power mix” suggests that cost, lead time, or performance metrics did not align with its rollout schedule.

Boom’s reliance on the Crusoe order to validate its Superpower line now faces a revenue gap. The firm’s plan to ship 250 MW to other customers mitigates the shortfall, but the loss of a high‑profile AI data‑center client reduces market credibility and may slow the path to the 1 GW target.

Implications for the energy‑tech market

The split highlights a broader tension between emerging AI workloads and nascent stationary‑power technologies. Companies that can bundle power generation with AI compute must meet stringent uptime, cost, and carbon‑reduction criteria. A single‑source turbine model may struggle against diversified portfolios that blend grid, solar, wind, battery storage, and modular gas generators.

For investors, the episode signals that capital‑intensive turbine projects need multiple anchor customers before scaling. Boom’s ability to secure additional contracts will determine whether its turbine line can achieve economies of scale or remain a niche offering.

What builders and businesses should do

• Map power‑source options to site‑specific cost and regulatory constraints before committing to a single technology.
• Keep a diversified energy portfolio that includes grid, renewables, and modular generators to hedge against supply‑chain delays.
• Evaluate turbine economics against alternative on‑site solutions such as combined‑cycle gas plants or battery‑backed solar.
• Track partner announcements closely; a shift in one firm’s strategy can ripple through the supply chain.
• Build flexibility into data‑center designs to accommodate future power‑mix changes without major retrofits.

The bottom line

Crusoe’s abandonment of the Boom turbine deal highlights the need for adaptable energy strategies as AI compute expands. Watch for Boom’s next customer announcements in early 2027 and for Crusoe’s power‑mix decisions as it scales new campuses across the United States.