Operational Data Analytics: Optimizing the National Energy Research Scientific Computing Center Cooling Systems

In 2017/2018, the Energy Efficient HPC Working Group (EE HPC WG) Dashboard Team conducted an analysis that assessed the current use of information dashboards for operational facility management in major supercomputing centers around the globe, resulting in the formalization of a process now referred to as Operational Data Analytics (ODA). Subsequent to surveys of multiple HPC facilities, the EE HPC WG determined that case studies were needed to help HPC facilities evaluate the value of implementing ODA practices. This paper provides a summary of the successful use of an ODA approach being used by the National Energy Research Scientific Computing Center (NERSC) at Lawrence Berkeley National Laboratory (Berkeley Lab) in meeting organizational energy efficiency performance goals for the NERSC HPC cooling systems.