Mapping and Frequency Joint Optimization for Energy Efficient Execution of Multiple Applications on Multicore Systems

International audienceRun-time resource managers are essential componentsto optimize energy consumption in cluster-based multicorearchitectures. However, with the ever increasing numberof functionalities supported by these architectures, it is alsonecessary to optimize the usage of processing resources whileguaranteeing that applications’ timing constraints are met. Inthis paper, we present a new run-time management strategythat includes both processing resource allocation and frequencytuning to optimize clusters energy consumption when multipleapplications are executed concurrently. The proposed run-timeallocation process minimizes the number of used processing coreswhile meeting the latency constraint of each application. Thisapproach offers a good trade-off between efficiency and complexity.The achieved energy saving has been demonstrated throughvarious case-studies with different sets of active applications.Results show an improvement of energy saving respectively of36% and up to 206% when compared to the literature