Energy consumption-based performance tuning of software and applications using Particle Swarm Optimization

Software development is increasing amidst of various emerging concerns for new technological trends, namely, grids, clouds, and HPC. However, Software developers of such technologies, to be more specific, are concerned about the performance aspects of their code - for instances, the developers are worried about the memory leakage, pipeline stalls, cache misses, and so forth. Recently, energy consumption analysis and tuning of software/applications have enabled a wide research spectrum among HPC research community. This research is crucial for developing an eco-friendly compute machines. In this scenario, our paper reveals a methodology which does energy consumption-based tuning of software and applications when Particle Swarm Optimization (PSO) algorithm was used in EnergyAnalyzer. EnergyAnalyzer is an online-based energy analysis tool which is a Department of Science and Technology, India, funded ongoing project. The research study was carried out in HPCCLoud Research Laboratory of our premise which comprises of a HP ProLiant 48 core compute machine.