Chaotic attractor prediction for server run-time energy consumption

Authors:
Adam Lewis;Jim Simon;Nian-Feng Tzeng
Affiliations:
Center for Advanced Computer Studies, University of Louisiana, Lafayette, Louisiana;Center for Advanced Computer Studies, University of Louisiana, Lafayette, Louisiana;Center for Advanced Computer Studies, University of Louisiana, Lafayette, Louisiana
Venue:
HotPower'10 Proceedings of the 2010 international conference on Power aware computing and systems
Year:
2010

Citing 8
Cited 2

Time Series Analysis: Forecasting and Control

Time Series Analysis: Forecasting and Control
Chaos and Time-Series Analysis

Chaos and Time-Series Analysis
SPEC CPU2006 benchmark descriptions

ACM SIGARCH Computer Architecture News
Models and metrics for energy-efficient computer systems

Models and metrics for energy-efficient computer systems
Proactive temperature balancing for low cost thermal management in MPSoCs

Proceedings of the 2008 IEEE/ACM International Conference on Computer-Aided Design
Thread criticality predictors for dynamic performance, power, and resource management in chip multiprocessors

Proceedings of the 36th annual international symposium on Computer architecture
Run-time energy consumption estimation based on workload in server systems

HotPower'08 Proceedings of the 2008 conference on Power aware computing and systems
Sleepless in seattle no longer

USENIXATC'10 Proceedings of the 2010 USENIX conference on USENIX annual technical conference

Runtime energy consumption estimation for server workloads based on chaotic time-series approximation

ACM Transactions on Architecture and Code Optimization (TACO)
OptiPlace: Designing Cloud Management with Flexible Power Models through Constraint Programing

UCC '13 Proceedings of the 2013 IEEE/ACM 6th International Conference on Utility and Cloud Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper proposes a chaotic time series model of server system-wide energy consumption to capture the dynamics present in observed sensor readings of underlying physical systems. Based on the chaotic model, we have developed a real-time predictor that estimates actual server energy consumption according to its overall thermal envelope. This chaotic time series regression model relates processor power, bus activity, and system ambient temperatures for real-time prediction of power consumption during job execution to enable run-time control of their thermal impacts. An experimental case study compares our Chaotic Attractor Predictor (CAP) against previous prediction models constructed according to other statistical methods. Our CAP is found to be accurate within an average error of 2% (or 7%) and the worst case error of 7% (or 20%) for the AMD Opteron processor (or for the Intel Nehalem processor), based on executing a set of SPEC CPU2006 benchmarks.