Christina Delimitrou, a newly tenured associate professor at MIT, and her research group apply machine learning to make large-scale data centers more efficient, secure and reliable, according to an MIT profile published Oct. 8, 2026. The work redesigns cloud computing systems and server architectures so operators can extract more computing power from existing hardware, which the profile frames as a way to limit the power burned to meet demand.

Delimitrou, a member of the Computer Science and Artificial Intelligence Laboratory, said most large systems she and her mentor examined during her Stanford graduate work were running at only about 15 percent capacity. Her group's tools include Seer, which uses deep learning to anticipate problems in web applications, and Ditto, which mimics proprietary applications so academic teams can study systems they cannot access directly. The profile says she is also working on making AI tools more interpretable, and she stresses that AI results still need auditing.

The article describes her path from northern Greece to the National Technical University of Athens, Stanford and Cornell before she joined MIT's Department of Electrical Engineering and Computer Science in 2022. It treats the efficiency and sustainability benefits as research aims, not demonstrated commercial results.