Optimizations for Energy Efficiency for Distributed Memory Programming Models

One of the primary challenges on the pathway to Exascale Computing is the 20MW power consumption envelope established by the U.S. Department of Energy's Exascale Initiative Steering Committee.  Extreme Scale Research reports indicate that the energy consumption during movement of data off-chip is orders of magnitude higher than within a chip. The direct outcome of this is a rising concern about the energy and power consumption of large-scale applications that rely on various communication libraries and parallelism constructs for distributed computing.    This work focuses on exploring multiple factors in the software stack that affect the energy consumption of large scale distributed memory applications.  As part of this talk, we present good programming practices, along with empirical evidence of their impact on program and system performance.