Google DeepMind запустила систему обучения с подкреплением для оптимизации энергопотребления дата-центров
The reinforcement learning system developed by Google DeepMind can dynamically control data center cooling and workload scheduling, learning optimal strategies based on historical telemetry data, and operated under human supervision in a closed loop with safety constraints.
Initial deployment achieved double-digit reductions in cooling energy consumption while maintaining reliability and thermal limits.
The system also introduces grid-aware scheduling, shifting flexible AI tasks to periods of low grid stress or high renewable energy availability, serving as a case study for RL application in large-scale industrial systems.
Google states it is willing to share experiences with operators.
Privacy and environmental advocates are closely monitoring, and the industry is working to offset the energy consumption growth caused by AI workloads.
Initial deployment achieved double-digit reductions in cooling energy consumption while maintaining reliability and thermal limits.
The system also introduces grid-aware scheduling, shifting flexible AI tasks to periods of low grid stress or high renewable energy availability, serving as a case study for RL application in large-scale industrial systems.
Google states it is willing to share experiences with operators.
Privacy and environmental advocates are closely monitoring, and the industry is working to offset the energy consumption growth caused by AI workloads.