SemiAnalysis: Reciprocating Engines May Become a Key Post-Meter Power Solution for AI Data Centers

Huo Xing Cai Jing reports that on July 21, the semiconductor and AI independent research firm SemiAnalysis stated that as demand for AI computing power grows rapidly, the power supply model for data centers is undergoing a transformation. Reciprocating engines, previously used primarily as backup power during grid outages, are increasingly being repositioned as baseline power sources operating around the clock. Reciprocating engine manufacturers have already signed contracts to supply approximately 1 GW of behind-the-meter power projects this year, with projected annual additions of over 4 GW in both 2027 and 2028. After modeling the U.S. grid’s capacity, SemiAnalysis concluded that existing grid reserves are expected to be fully consumed between 2027 and 2028, and planned utility-scale generation capacity through 2030 remains insufficient to meet the additional load demands from data centers. Combining this with its data center model, SemiAnalysis estimates that approximately 140 GW of potential data center projects have yet to sign power supply agreements and may increasingly adopt behind-the-meter power solutions to circumvent grid expansion bottlenecks. Among various behind-the-meter power technologies—including reciprocating engines, aviation-derived gas turbines, and fuel cells—SemiAnalysis expects reciprocating engines to capture the largest market share due to their lower cost, rapid deployment, modular scalability, and stronger financing capabilities. Equipment manufacturers such as Caterpillar, INNIO, and Cummins are expanding production capacity to support large-scale deployments in the coming years. As AI data centers continue to expand, on-site self-supplied power is evolving from a traditional backup system into a critical energy infrastructure, with reciprocating engines poised to become a key solution for bridging the power gap in the age of computing.