One of the research areas of EECE Associate Professor, Cris Ababei, is energy efficiency of computer systems across all levels: starting from multicore processor level up to the cloud level. His research on this topic started with the help of an award from National Science Foundation (NSF); to study methods to improve lifetime reliability and to reduce power consumption of future multicore processors with tens, hundreds, or even thousand cores per chip. As part of this project, Dr. Ababei and his team developed novel dynamic power and reliability-management techniques based on dynamic voltage and frequency scaling and application remapping as well as deep machine learning (ML) methods. This research was taken to the next level, namely focusing on datacenter or cloud levels, which has been funded in part by another award from NSF. At the cloud level, they developed innovative ML based hierarchical scheduling algorithms to improve the energy delay product of increasing number of jobs run on today's complex datacenters.