Profiling and analysis of power consumption for virtualized systems and applications

Power consumption and energy efficiency have been a great concern in large-scale computing system such as data centers and compute clouds. Moreover, the quantity of power consumption and resources utilized for cooling in large-scale computer systems will be beyond the boundary we can control [1]. There are at least two levels at which power/energy saving are being made. The first is at the computer architecture level. Techniques have been proposed to reduce processor energy usage through multi-core processors, designing energy proportional hardware, and increasing the number of performance and sleep states. A second level of energy optimizations is at the system layer, either at a single server or across multiple servers, through energy efficient use of the underlying hardware.