AI data centers turn to converted jet engines for faster power
Jet engines fill a power gap for AI campuses
Some AI data center operators are turning retired aircraft engines into natural-gas generators as they wait years for grid connections and new turbine capacity. The New York Times reported that rebuilt jet engines can allow a data center to begin operating in about 24 months, while connecting 100 megawatts to the grid in Northern Virginia takes about seven years, based on International Energy Agency estimates.
The equipment is not presented as a permanent substitute for utility-scale generation. Its appeal is speed: operators facing lengthy interconnection queues can install self-generated power rather than leave newly built computing facilities idle. Rystad Energy analyst Audun Martinsen told the Times that the priority is obtaining electricity as quickly as possible.
Operators and suppliers pursue interim generation
ProEnergy rebuilds a GE engine commonly used on Boeing jets so it can burn natural gas. State permits and satellite imagery cited by the Times show xAI installed eight converted jet-engine turbines in Mississippi to help power its Colossus 2 data center in Memphis.
OpenAI used gas turbines at its first Stargate site in Abilene, Texas in 2025. Crusoe, the Denver-based company building that campus, ordered 13 converted units from ProEnergy in April. In September, Crusoe ended a $1.25 billion agreement to buy 29 natural-gas turbines from Boom Supersonic, although a company spokesperson told TechCrunch that Crusoe still planned to use turbines.
Grid constraints can alter commercial plans. Microsoft withdrew as anchor customer for Nscale's Loughton, England, data center after the local utility said it could not provide sufficient power until the early to mid-2030s. Mitsubishi Power, a major power-generation equipment maker, has said most new power capacity is expected to arrive by 2031.
Fast deployment carries efficiency and emissions costs
Converted aircraft engines have significant limitations. The Times reported that they are less efficient and more polluting than turbines designed for conventional gas plants, and more units are needed to produce the same amount of electricity. Some experts expect efficiency and environmental considerations to become more important within a year or two.
The demand pressure sits alongside AI-driven data-center electricity demand through 2035, which examines how AI-driven electricity use could reshape data-center planning through 2035. The Trump administration is also encouraging data centers to generate their own power, adding to the incentive for on-site alternatives.
For businesses procuring AI infrastructure, a grid connection timetable should be treated as a core capacity constraint. Interim generation may accelerate deployment, but its fuel, equipment, efficiency and environmental trade-offs need to be evaluated alongside the availability of long-term utility power.