WiP abstract: Intelligent power- and performance-aware tradeoffs for multicore servers in cloud data centers

Cloud service providers are seeking new solutions for reducing the energy consumption in their data centers. Our work in developing middleware that is integrated with a Virtual Machine Manager (VMM) to lower the energy consumption while maximizing the performance of or minimizing the service level agreement (SLA) violations for applications. To that end our solution leverages techniques such as Dynamic voltage/frequency scaling (DVFS) and server consolidation through live migration of virtual machines from an overloaded host machine to a less-loaded one.