Holistic Control Cooling of a 6G Edge Node using Jalusi-based Airflow control prototype

Sammanfattning

This report investigates power demand, cooling efficiency, and associated operational costs in a rack-scale data center test environment, representative of future edge computing scenarios. As power densities increase in modern compute infrastructure, maintaining thermal stability without excessive energy use has become a critical challenge. The study focuses on the implementation of a jalusi-based airflow control system—an adjustable blanking panel placed in front of server racks—to modulate Δρ, the differential pressure, between the hot and cold aisles of the server module, and influence cooling dynamics. By analyzing temperature differentials between supply and exhaust air in conjunction with detailed power usage data, the report evaluates how variations in server load and airflow restriction impact cooling performance and energy use. The findings are visualized through bar graphs and accompanied by tables detailing energy and cooling costs. The report highlights the correlation between load variations, ΔT values, and energy efficiency, offering insights for optimizing cooling strategies to reduce operational costs. The analysis provides recommendations for improving energy efficiency and cooling effectiveness in high-density computing environments. The implementation of the jalusi-based airflow control system resulted in a measurable reduction in Power Usage Effectiveness (PUE), reaching a value of 1.04—indicating improved energy efficiency compared to experiments conducted without blanking panels. This was achieved by lowering fan power demand relative to server load.