Empirical power consumption characterization and energy aware scheduling in data centers

Energy-efficient management is key to reduce operational cost and environmental contamination in modern data centers. Energy renewable energy utilization are strategies optimize consumption high-performance computing. In any case, understanding the power behavior of physical servers datacenter fundamental implement energy-aware policies effectively. These should deal with possible performance degradation applications ensure quality service. This manuscript presents an empirical evaluation for scientific computing multicore systems. Three types studied, single combined executions on Intel AMD servers, evaluating overall each application. The main results indicate that has a strong dependency type Additional analysis shows best load server regarding efficiency depends applications, decreasing heavily loaded situations. allow formulating models characterize according consumption, efficiency, resource sharing, which provide useful information scheduling policies. Several evaluated using proposed model over realistic workloads. Results confirm maximize host efficiency.