AI Data Centers Harness Repurposed Jet Engines to Accelerate Power Generation

Repurposed Jet Engines Powering AI Data Centers

According to НВ — Техно: To meet the surging power demands of artificial intelligence, several data centers have turned to a novel solution: using refurbished jet engines to generate electricity. This approach significantly cuts down the time required to become operational, circumventing the multi-year delays often encountered when connecting to traditional power grids. Industry leaders like Elon Musk’s xAI and OpenAI have already adopted this innovative technology.

These converted engines, adapted to run on natural gas, enable data centers to be up and running in roughly two years, according to Rystad Energy analyst Audun Martinsen. In contrast, adding 100 MW of grid capacity in Northern Virginia can take up to seven years, as reported by the International Energy Agency.

Companies Implementing Jet Engine Power Solutions

xAI has deployed eight gas turbines at its Colossus 2 data center in Memphis, Mississippi. These turbines are supplied by ProEnergy, a firm specializing in refurbishing GE engines originally designed for Boeing aircraft. Meanwhile, OpenAI’s Stargate project in Abilene, Texas, also utilizes gas turbines, with the startup Crusoe handling campus construction.

Earlier this year, Crusoe ordered 13 refurbished units from ProEnergy. However, in September, the company canceled a $1.25 billion contract for 29 gas turbines from Boom Supersonic, a developer of supersonic passenger jets that also offers modified engines for data center power generation.

While repurposed jet engines enable faster data center deployment, they come with trade-offs. - The New York Times

These engines are less efficient and produce higher emissions compared to standard gas turbines used in conventional power plants. More units are required to generate the same amount of electricity. During the Trump administration, there was a push for data center operators to self-generate power, reflecting broader concerns over energy independence.

The use of converted aviation engines highlights a creative pivot in addressing the urgent energy needs of AI infrastructure. This strategy illustrates how technologies outside the traditional energy sector are being adapted to meet immediate demands. However, environmental impact and operational efficiency remain key considerations. Given the global emphasis on reducing emissions and transitioning to cleaner energy sources, jet engine power may serve as a temporary fix for data centers needing rapid deployment rather than a long-term solution.