For a number of reasons, improving energy efficiency is a critical need in the quest for exascale computing, according to a recent blog post from HiPEAC, the European Network on High Performance and Embedded Architecture and Compilation.
The post explores why energy efficiency is a driving factor and introduces the PowerStack collaboration, a part of the Exascale Computing Project (ECP) that is endeavoring to design a holistic, extensible power management framework. Insights are provided by some of the members of the collaboration’s core committee, which is composed of experts from the Energy Efficient High Performance Computing Working Group, ECP, Intel, Lawrence Livermore National Laboratory (LLNL), Leibniz Supercomputing Centre, RIKEN, Technical University of Munich, and the University of Tokyo.
Tapasya Patki, a computer scientist at the Center for Applied Science Computing and principal investigator of the ECP Power Steering project at LLNL, is among the committee members explaining the challenges.
Let’s Talk Exascale podcast episode 11, featuring Tapasya Patki: Efficiently Using Power and Optimizing Performance of Scientific Applications at Exascale