Determine energy-saving potential in wait-states of large-scale parallel programs

Knobloch, Michael; Mohr, Bernd; Minartz, Timo

doi:10.1007/s00450-011-0196-7

Cited by 14 publications

(9 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In such cases, the power consumption of idle cores or processors could be avoided by either switching them on/off or clocking the CPU frequencies to low power states [3]. Knobloch et al [7] and Flin et al [4] have studied the effects of the wait times of real world MPI applications and their respective energy consumption details.…”

Section: Related Workmentioning

confidence: 99%

Threshold acceptance algorithm based energy tuning of scientific applications using EnergyAnalyzer

Benedict

2014

Proceedings of the 7th India Software Engineering Conference

View full text Add to dashboard Cite

Energy tuning of scientific applications is emerging as an important research agenda among HPC application developer and performance analysis tool developer communities as it is a critical and ponderous task to deal with, especially when exa-scale computing machines were concerned. The energy tuning research has deepened its thrust among researchers due to the necessity of addressing an energy consumption issue of applications which is considered as one of the global challenge. Although a few efforts were carried out in the direction of automatic energy tuning of scientific applications, the working researchers encounter challenges, such as, finding an optimal solution within a short time frame among a wide set of optimization search space, generating fully automatic optimized versions of scientific applications, considering the heterogeneous nature of future generation compute machine architectures, and so forth. This paper discusses about an online based energy consumption tuning mechanism of EnergyAnalyzer tool using Threshold Acceptance (TA) algorithm. In addition, the paper demonstrates the speedy process of finding a near-optimal optimized code versions of a scientific application among a huge set of optimization search space using EnergyAnalyzer. EnergyAnalyzer is an online-based energy consumption analysis tool which is under development at the HPCCLoud Research laboratory of our premise. The proposed algorithm and the energy tuning mechanism could be of an interest to energy conscious scientific application developers and tool developers.

show abstract

Section: Related Workmentioning

confidence: 99%

Threshold acceptance algorithm based energy tuning of scientific applications using EnergyAnalyzer

Benedict

2014

Proceedings of the 7th India Software Engineering Conference

View full text Add to dashboard Cite

show abstract

“…Knobloch et al [28] have analyzed the external power information in Vampir to see that a parallel application has shown high power consumption in MPI routines due to active waiting, i.e. polling with highest frequency whether the communication partner is ready.…”

Section: Tracing-based Toolsmentioning

confidence: 99%

Tools and Methods for Measuring and Tuning the Energy Efficiency of HPC Systems

Schöne

Treibig²,

Dolz

et al. 2014

Scientific Programming

Self Cite

View full text Add to dashboard Cite

Energy costs nowadays represent a significant share of the total costs of ownership of High Performance Computing (HPC) systems. In this paper we provide an overview on different aspects of energy efficiency measurement and optimization. This includes metrics that define energy efficiency and a description of common power and energy measurement tools. We discuss performance measurement and analysis suites that use these tools and provide users the possibility to analyze energy efficiency weaknesses in their code. We also demonstrate how the obtained power and performance data can be used to locate inefficient resource usage or to create a model to predict optimal operation points. We further present interfaces in these suites that allow an automated tuning for energy efficiency and how these interfaces are used. We finally discuss how a hard power limit will change our view on energy efficient HPC in the future.

show abstract

“…Knobloch et al [18] analyze power performanc e of MPI applicati ons. Their study is specially focused on energy saving opportunities that originate from busy-wait states that are shown to be highly power-inefficient.…”

Section: Related Workmentioning

confidence: 99%

A novel black-box simulation model methodology for predicting performance and energy consumption in commodity storage devices

Prada

Calderón

García

et al. 2013

Simulation Modelling Practice and Theory

View full text Add to dashboard Cite

Traditional approache s for storage devices simulation have been based on detailed and analytic models. However, analytic models are difficult to obtain and detailed models require a high computational cost which may be not af fordable for large scale simulations (e.g. detailed data center simulation s). In current systems like large clusters, grids, or clouds, performance and energy studies are critical, and fast simulation s take an impo rtant role on them.A different approach is the black-box statistical modeling, where the storage device, its interface, and the interconnection mechanisms are modeled as a single stochastic process, defining the request response time as a random variable with an unknown distribution. A random variate generator can be built and integrated into a bigger simulation model. This approach allows to generate a simulation model for both real and synthetic complex workloads.This article describes a novel methodology that aims to build fast simulation models for storage devices. Our method uses as starting point a workload and produc es a random variate generator which can be easily integrated into large scale simulation models. A comparison between our variate generator and the widely known simulation tool DiskSim, shows that our variate generator is faster, and can be as accurate as DiskSim for both performa nce and energy consumption predictions.

show abstract

Determine energy-saving potential in wait-states of large-scale parallel programs

Cited by 14 publications

References 16 publications

Threshold acceptance algorithm based energy tuning of scientific applications using EnergyAnalyzer

Threshold acceptance algorithm based energy tuning of scientific applications using EnergyAnalyzer

Tools and Methods for Measuring and Tuning the Energy Efficiency of HPC Systems

A novel black-box simulation model methodology for predicting performance and energy consumption in commodity storage devices

Contact Info

Product

Resources

About