Dynamic Voltage and Frequency Scaling for Scientific Applications

Hsu, Chung-Hsing; Kremer, Ulrich

doi:10.1007/3-540-35767-x_6

Cited by 4 publications

(2 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Load imbalances between processes and threads provide significant opportunities to decrease energy consumption. DVFS has been applied to improve the EE of MPI programs by lowering the frequency of light-loaded processors (Freeh et al, 2005a(Freeh et al, , 2005bGe et al, 2005;Hsu and Feng, 2005;Hsu and Kremer, 2003a;Kappiah et al, 2005;Li et al, 2010a;Rountree et al, 2007Rountree et al, , 2009Rountree et al, , 2011. Dynamic concurrency throttling (DCT) controls the number of active CPU cores for multithreaded programs, including OpenMP programs, at runtime (Curtis-Maury et al, 2006a, 2006bFreeh et al, 2007a;Grant and Afsahi, 2006;Li and Martinez, 2006;Li et al, 2010b;Suleman et al, 2008).…”

Section: Taxonomy Of Energy-efficient Methods For Parallel Computingmentioning

confidence: 99%

“…DVFS can be applied at different levels, such as a whole program or a function call. Using the Wattch CPU energy and performance simulator (Brooks et al, 2000), Hsu and Kremer (2003a) investigate the opportunity of applying DVFS to save energy for highly optimized scientific codes by taking advantage of memory stalls. In particular, using five SPECfp95 benchmark applications, they demonstrate that energy consumption can be reduced up to 60% with a performance penalty of 9.58% or less.…”

Section: Dvfs-based Power-saving Methods For Mpi Applicationsmentioning

confidence: 99%

See 1 more Smart Citation

A survey on software methods to improve the energy efficiency of parallel computing

Jin

Supinski

Abramson

et al. 2016

The International Journal of High Performance Computing Applica

View full text Add to dashboard Cite

Energy consumption is one of the top challenges for achieving the next generation of supercomputing. Codesign of hardware and software is critical for improving energy efficiency (EE) for future large-scale systems. Many architectural power-saving techniques have been developed, and most hardware components are approaching physical limits. Accordingly, parallel computing software, including both applications and systems, should exploit power-saving hardware innovations and manage efficient energy use. In addition, new power-aware parallel computing methods are essential to decrease energy usage further. This article surveys software-based methods that aim to improve EE for parallel computing. It reviews the methods that exploit the characteristics of parallel scientific applications, including load imbalance and mixed precision of floating-point (FP) calculations, to improve EE. In addition, this article summarizes widely used methods to improve power usage at different granularities, such as the whole system and per application. In particular, it describes the most important techniques to measure and to achieve energy-efficient usage of various parallel computing facilities, including processors, memories, and networks. Overall, this article reviews the state-of-the-art of energy-efficient methods for parallel computing to motivate researchers to achieve optimal parallel computing under a power budget constraint.

show abstract

Section: Taxonomy Of Energy-efficient Methods For Parallel Computingmentioning

confidence: 99%

Section: Dvfs-based Power-saving Methods For Mpi Applicationsmentioning

confidence: 99%

A survey on software methods to improve the energy efficiency of parallel computing

Jin

Supinski

Abramson

et al. 2016

The International Journal of High Performance Computing Applica

View full text Add to dashboard Cite

show abstract

Pattern Description for the Energy-Efficient Code Generation

So¹,

Ko²

2009

2009 International Conference on New Trends in Information and Service Science

View full text Add to dashboard Cite

Together with a new advent of embedded processor developed to support specific application area, and it evolution, a new study of software development to support the embedded processor and its commercial use has been revitalized. In this paper, specially, In a mobile device that is built-in embedded processor, software management is as important as hardware management for the limited energy. In this paper, we suggest that the code generation technique considering the energy dissipation through the verified retargetable compiler backend tool, EXPRESSION. For this goals, we describes the efficient code generation patterns and showed the variable performance results.

show abstract