Sustaining Performance While Reducing Energy Consumption: A Control Theory Approach

Rutten, Éric; Cerf, Sophie; Bleuse, Raphaël; Reis, Valentin; Perarnau, Swann

doi:10.1007/978-3-030-85665-6_21

Cited by 6 publications

(13 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In Computer Science and Software Engineering, Autonomic Computing [2] is addressing this concern, proposing software architectures with feedback loops involving a variety of decision mechanisms (e.g., programmatic, based on AI and Machine Learning, scheduling, constraint programming). A particularly interesting approach involves Control Theory [3]- [5] with applications to HPC e.g., [6]- [8].…”

Section: A Need Of Control Theory For High Performance Computingmentioning

confidence: 99%

Model-Free Control for Resource Harvesting in Computing Grids

Guilloteau

Robu

Join³

et al. 2022

2022 IEEE Conference on Control Technology and Applications (CCTA)

View full text Add to dashboard Cite

Cloud and High-Performance Computing (HPC) systems are increasingly facing issues of dynamic variability, in particular w.r.t. performance and power consumption. They are becoming less predictable, and therefore demand more runtime management by feedback loops. In this work, we describe results addressing autonomic administration in HPC systems through a control theoretical approach. We more specifically consider the need for controllers that can adapt to variations a long timee in the behavior of controlled systems, but also to being reused on different systems and processors. We therefore explore the application of Model-Free Control (MFC) in the context of resource harvesting in a Computing Grid, by regulating the injection of flexible jobs while limiting perturbation of the prioritary applications.

show abstract

Section: A Need Of Control Theory For High Performance Computingmentioning

confidence: 99%

Model-Free Control for Resource Harvesting in Computing Grids

Guilloteau

Robu

Join³

et al. 2022

2022 IEEE Conference on Control Technology and Applications (CCTA)

View full text Add to dashboard Cite

show abstract

“…The present work builds on and improves on the energy regulation for sober HPC systems proposed by [14]. That previous work defined a nonlinear model with first-order dynamics and a proportional-integral (PI) controller relying on RAPL as a power actuator to decrease energy consumption while sustaining an appropriate level of performance.…”

Section: B Contributionsmentioning

confidence: 99%

“…• Identification of the limitation of the model and the available control in [14] • Design of an adaptive controller that is robust to different clusters and to the applications environment • A novel initialization method called two-runs, well suited for short and repeated controller executions • Extensive experimental evaluation of the proposed controller on the data center testbed. The remainder of this paper is organized as follows.…”

Section: B Contributionsmentioning

confidence: 99%

“…Modern processors are able to adapt the speed of computation to save energy. The original approach described in [14] and refined in this work takes root in the observation that during non-compute-bound phases, processors can be slowed down without affecting the performance of the application. A typical HPC system architecture is as follows.…”

Section: A Hpc System: Application and Architecturementioning

confidence: 99%

“…We are now interested in analyzing the control problem regarding possible disturbances, noises, and nonlinearities. First, the literature emphasized that the RAPL actuator is not accurate [16], and this error increases as the control signal grows [14]. Figure 3 gives the static characteristic of a specific HPC application [14].…”

Section: System Analysismentioning

confidence: 99%

See 2 more Smart Citations

Adaptive Power Control for Sober High-Performance Computing

Hawila

Cerf

Bleuse

et al. 2022

2022 IEEE Conference on Control Technology and Applications (CCTA)

Self Cite

View full text Add to dashboard Cite

Soberness-in terms of electrical power-of data centers and high-performance computing systems is becoming an important design issue, as the global energy consumption of information technologies is rising at considerable levels. This issue is all the more complex as these systems are increasingly heterogeneous and variable in their behavior, for example, w.r.t. performance and power consumption, and less predictable, thus demanding runtime management and feedback control.This paper addresses the problem of the control of the power allocated to processors and hence their energy consumption and performance. The use of feedback control allows the energy consumption to be reduced by decreasing the speed without losing performance, by exploiting periods where read/write operations slow the progress. Previous works present limitations regarding both modeling (nonlinear models with numerous parameters) and control performance (mainly instability caused by platform variations). We develop a novel adaptive control that is robust to the variety of execution platforms while maintaining the existing global goals of energy management. We evaluate-on a real system using the Grid'5000 testbed-the robustness of the control to changes in initial parameters and to disturbances, and we compare it with the previous proportionalintegral (PI) control. Our adaptive control approach allows for up to 25% energy savings.

show abstract