A new energy aware performance metric
- 314 Downloads
Energy aware algorithms are the wave of the future. The development of exascale systems made it clear that extrapolations of current technologies, algorithmic practices and performance metrics are simply inadequate. The community reacted by introducing the FLOPS/WATT metric in order to promote energy awareness. In this work we take a step forward and argue what one should aim for is the total reduction of the spent energy in conjunction with minimization of time to solution. Thus, we propose to use f(time to solution)⋅energy (FTTSE) as the performance metric, where f(⋅) is an application dependent function of time. In this paper, we introduce our ideas and showcase them with a recently developed framework for solving large dense linear systems.
KeywordsEnergy aware Performance metrics Analysis Fault tolerance
Unable to display preview. Download preview PDF.
- 1.Bekas C, Curioni A, Fedulova I (November 2009) Low cost high performance uncertainty quantification. Workshop on high performance computational finance, SC09, Portland, OR, USA Google Scholar
- 4.Colella P (2004) Defining software requirements for scientific computing Google Scholar
- 6.IBM Red Book (2009) IBM system Blue Gene solution: performance analysis tools. http://www.redbooks.ibm.com/abstracts/redp4256.html. Last update 6 May 2010
- 7.Kogge P (2009) The road to exascale: hardware and software challenges panel, SC09, Nov. 14–19, 2009, Portland, Oregon, USA. Available at: www.exascale.org/mediawiki/images/6/6e/Sc09-exa-panel-kogge.pdf. Accessed 2009
- 9.The Green 500. www.green500.org. Accessed 2010
- 10.Whaley RC, Dongarra J (1998) Automatically tuned linear algebra software. SuperComputing 1998: High performance networking and computing Google Scholar