Runtime estimation of performance–power in CMPs under QoS constraints

One of the main challenges in data center systems is operating under certain Quality of Service (QoS) while minimizing power consumption. Increasingly, data centers are exploring and adopting heterogeneous server architectures with different power and performance trade-offs. This not only requires careful understanding of the application behavior across multiple architectures at runtime so as to enable meeting power and performance requirements but also an understanding of individual and aggregated behaviour of application and server level performance and power metrics