The explosion of cloud services has turned data centers into silent environmental villains. They now consume up to 1‑2 % of global electricity, producing billions of tons of CO₂ each year. As demand skyrockets, the sector’s carbon footprint threatens climate goals, forcing researchers to hunt for smarter, greener operating models.
Associate Professor Christina Delimitrou is tackling this challenge by applying artificial intelligence directly to workload management. Her AI‑driven framework predicts traffic spikes, dynamically allocates resources, and shuts down idle servers without sacrificing performance. Early tests show a 20‑30 % reduction in power draw while maintaining latency standards.
Deploying such intelligent orchestration at scale could reshape the entire cloud ecosystem. Operators would enjoy lower electricity bills and meet regulatory pressure, while the planet benefits from a tangible carbon cut. If Delimitrou’s research gains industry traction, AI may become the key lever that transforms data centers from eco‑hazards into model sustainability champions and inspire broader green tech adoption globally.
Source: Read original article