Practical power consumption estimation for real life HPC applications
Now cloud computing is rapidly growing as an alternative to traditional computing architecture. However, it is based on models like cluster computing in general. Thus, improving the energy consumption of the cluster system is the basis for the green cloud. In order to reach exascale computing, more and more efforts are made to improve the energy consumption and efficiency in high performance computing systems. As the de facto standard for designing parallel applications in cluster environment, the Message Passing Interface has been widely used in high performance computing. Therefore, getting the energy consumption information of MPI applications is critical for improving the energy efficiency of cluster systems. By creating a distributed measuring framework which can collect all nodes' energy consumption without the aid of power meters, it is possible to get the detailed energy information of an MPI application. In this work, we present MEMT, a software framework that eases the energy collection in cluster environment. Using this tool, it is viable to find out parameters that affect an MPI program's energy efficiency and to build the models for execution time and energy consumption. Based on the pre-built models, energy saving strategy can be designed. The use of this tool is tested in a cluster.