Presenter

Michael Thomas has collaborated with three-dozen Top500 organizations on HPC data center capacity planning, design, and optimization initiatives. Mike understands that the quality of the mechanical, electrical, structural and building automation infrastructure can be a roadblock or enabler for HPC deployments, mission execution, and competitive advantage. Mike advocates early stakeholder alignment, business case development, and total cost of ownership (TCO) analysis to deliver machine room infrastructure that is more cost effective, energy efficient, resilient, secure, scalable, and sustainable. In addition to collaborating with LLNL, Mike’s engagements include NCSA’s Blue Waters Petascale Computing Facility design and Argonne National Laboratory’s Aurora supercomputer infrastructure commissioning. Mike has participated in SC since 2006 and co-presented the “Data Center Planning and Design for HPC Folks” tutorials in 2016 and 2017. Mike has presented at the PRACE Symposium, HP CAST, NCSA’s Annual Meeting, University of Iowa’s IT Tech Forum, and SUNY’s Technology Conference.