The Community for Technology Leaders
2011 International Green Computing Conference and Workshops (2011)
Orlando, FL
July 25, 2011 to July 28, 2011
ISBN: 978-1-4577-1222-7
pp: 1-6
H. Anzt , Inst. for Appl. & Numerical Math. 4, Karlsruhe Inst. of Technol., Karlsruhe, Germany
V. Heuveline , Inst. for Appl. & Numerical Math. 4, Karlsruhe Inst. of Technol., Karlsruhe, Germany
J. I. Aliaga , Depto. de Ing. y Cienc. de Comput., Univ. Jaume I, Castellon, Spain
M. Castillo , Depto. de Ing. y Cienc. de Comput., Univ. Jaume I, Castellon, Spain
J. C. Fernandez , Depto. de Ing. y Cienc. de Comput., Univ. Jaume I, Castellon, Spain
R. Mayo , Depto. de Ing. y Cienc. de Comput., Univ. Jaume I, Castellon, Spain
E. S. Quintana-Orti , Depto. de Ing. y Cienc. de Comput., Univ. Jaume I, Castellon, Spain
ABSTRACT
Energy efficiency is a major concern in modern high-performance-computing. Still, few studies provide a deep insight into the power consumption of scientific applications. Especially for algorithms running on hybrid platforms equipped with hardware accelerators, like graphics processors, a detailed energy analysis is essential to identify the most costly parts, and to evaluate possible improvement strategies. In this paper we analyze the computational and power performance of iterative linear solvers applied to sparse systems arising in several scientific applications. We also study the gains yield by dynamic voltage/frequency scaling (DVFS), and illustrate that this technique alone cannot to reduce the energy cost to a considerable amount for iterative linear solvers. We then apply techniques that set the (multi-core processor in the) host system to a low-consuming state for the time that the GPU is executing. Our experiments conclusively reveal how the combination of these two techniques deliver a notable reduction of energy consumption without a noticeable impact on computational performance.
INDEX TERMS
GPU, power consumption optimisation, power consumption analysis, iterative solution, sparse linear system, multicore platforms, manycore platforms, energy efficiency, hybrid platforms, hardware accelerators, graphics processors, energy analysis, dynamic voltage/frequency scaling, DVFS
CITATION

V. Heuveline et al., "Analysis and optimization of power consumption in the iterative solution of sparse linear systems on multi-core and many-core platforms," 2011 International Green Computing Conference and Workshops(IGCC), Orlando, FL, 2011, pp. 1-6.
doi:10.1109/IGCC.2011.6008594
165 ms
(Ver 3.3 (11022016))