Ensuring Efficiency of Exascale Supercomputer Centers
Владимир Валентинович Воеводин, Vadim Voevodin · 2015
In this paper, we describe a set of systems aimed to ensure efficiency of supercomputer centers at every level of concern. This includes application and cluster performance monitoring and analysis, user and resource management, supercomputer reliability and viability. Each system is addressed to solve one particular field of research, but they are designed to be interrelated allowing for more holistic and deep efficiency analysis overall.