On the Importance of Nonlinear Modeling in Computer Performance Prediction

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8207)


Computers are nonlinear dynamical systems that exhibit complex and sometimes even chaotic behavior. The low-level performance models used in the computer systems community, however, are linear. This paper is an exploration of that disconnect: when linear models are adequate for predicting computer performance and when they are not. Specifically, we build linear and nonlinear models of the processor load of an Intel i7-based computer as it executes a range of different programs. We then use those models to predict the processor loads forward in time and compare those forecasts to the true continuations of the time series.


Multiple Linear Regression Multiple Linear Regression Model Computer Performance Iterate Function System Prediction Horizon 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Alexander, Z., Bradley, E., Garland, J., Meiss, J.: Iterated function system models in data analysis: Detection and separation. CHAOS 22(2) (April 2012)Google Scholar
  2. 2.
    Alexander, Z., Mytkowicz, T., Diwan, A., Bradley, E.: Measurement and dynamical analysis of computer performance data. In: Cohen, P.R., Adams, N.M., Berthold, M.R. (eds.) IDA 2010. LNCS, vol. 6065, pp. 18–29. Springer, Heidelberg (2010)CrossRefGoogle Scholar
  3. 3.
    Bass, T.: The Eudaemonic Pie. Penguin, New York (1992)Google Scholar
  4. 4.
    Browne, S., Deane, C., Ho, G., Mucci, P.: PAPI: A portable interface to hardware performance counters. In: Proceedings of Department of the Defense HPCMP Users Group Conference (1999)Google Scholar
  5. 5.
    Casdagli, M., Eubank, S. (eds.): Nonlinear Modeling and Forecasting. Addison Wesley (1992)Google Scholar
  6. 6.
    Cavazos, J., Fursin, G., Agakov, F., Bonilla, E., O’Boyle, M.F.P., Temam, O.: Rapidly selecting good compiler optimizations using performance counters. In: Proceedings of the International Symposium on Code Generation and Optimization (2007)Google Scholar
  7. 7.
    Desikan, R., Burger, D., Keckler, S.W.: Measuring experimental error in microprocessor simulation. In: Proceedings of the International Symposium on Computer Architecture (2001)Google Scholar
  8. 8.
    Faraway, J.J.: Practical regression and ANOVA in R (2002),
  9. 9.
    Fraser, A., Swinney, H.: Independent coordinates for strange attractors from mutual information. Physical Review A 33(2), 1134–1140 (1986)MathSciNetCrossRefzbMATHGoogle Scholar
  10. 10.
    Garland, J., Bradley, E.: Predicting computer performance dynamics. In: Gama, J., Bradley, E., Hollmén, J. (eds.) IDA 2011. LNCS, vol. 7014, pp. 173–184. Springer, Heidelberg (2011)CrossRefGoogle Scholar
  11. 11.
    Grassberger, P., Hegger, R., Kantz, H., Schaffrath, C., Schreiber, T.: On noise reduction methods for chaotic data. Chaos 3, 127 (1993)MathSciNetCrossRefzbMATHGoogle Scholar
  12. 12.
    Grechanik, M., Fu, C., Xie, Q.: Automatically finding performance problems with feedback-directed learning software testing. In: Proceedings of the International Conference on Software Engineering (2012)Google Scholar
  13. 13.
    Jain, R.: The Art of Computer Systems Performance Analysis: Techniques for Experimental Design, Measurement, Simulation, and Modeling, 2nd edn. John Wiley & Sons (1991)Google Scholar
  14. 14.
    Kantz, H., Schreiber, T.: Nonlinear Time Series Analysis. Cambridge University Press (1997)Google Scholar
  15. 15.
    Kennel, M., Isabelle, S.: Method to distinguish possible chaos from colored noise and to determine embedding parameters. Phys. Rev. A 46, 3111 (1992)CrossRefGoogle Scholar
  16. 16.
    Kennel, M.B., Brown, R., Abarbanel, H.D.I.: Determining minimum embedding dimension using a geometrical construction. Physical Review A 45, 3403–3411 (1992)CrossRefGoogle Scholar
  17. 17.
    Liu, J., Wu, S., Zidek, J.: On segmented multivariate regression. Statistica Sinica 7, 497–525 (1997)MathSciNetzbMATHGoogle Scholar
  18. 18.
    Lorenz, E.N.: Atmospheric predictability as revealed by naturally occurring analogues. Journal of the Atmospheric Sciences 26, 636–646 (1969)CrossRefGoogle Scholar
  19. 19.
    Makhoul, J.: Linear prediction: A tutorial review. Proceedings of the IEEE 63(4), 561–580 (1975)CrossRefGoogle Scholar
  20. 20.
    Martinez, J.F., Ipek, E.: Dynamic multicore resource management: A machine learning approach. IEEE Micro 29(5), 8–17 (2009)CrossRefzbMATHGoogle Scholar
  21. 21.
    McGee, V.E., Carleton, W.T.: Piecewise regression. Journal of the American Statistical Association 65(331), 1109–1124 (1970)CrossRefGoogle Scholar
  22. 22.
    Moseley, T., Kihm, J.L., Connors, D.A., Grunwald, D.: Methods for modeling resource contention on simultaneous multithreading processors. In: Proceedings of the International Conference on Computer Design (2005)Google Scholar
  23. 23.
    Myktowicz, T., Diwan, A., Bradley, E.: Computers are dynamical systems. Chaos 19, 033124 (2009), doi:10.1063/1.3187791Google Scholar
  24. 24.
    Mytkowicz, T.: Supporting experiments in computer systems research. Ph.D. thesis, University of Colorado (November 2010)Google Scholar
  25. 25.
    Packard, N., Crutchfield, J., Farmer, J., Shaw, R.: Geometry from a time series. Physical Review Letters 45, 712 (1980)CrossRefGoogle Scholar
  26. 26.
    Sauer, T., Yorke, J., Casdagli, M.: Embedology. Journal of Statistical Physics 65, 579–616 (1991)MathSciNetCrossRefzbMATHGoogle Scholar
  27. 27.
    Takens, F.: Detecting strange attractors in fluid turbulence. In: Rand, D., Young, L.S. (eds.) Dynamical Systems and Turbulence, Springer, Berlin (1981)Google Scholar
  28. 28.
    Weigend, A., Gershenfeld, N. (eds.): Time Series Prediction: Forecasting the Future and Understanding the Past. Santa Fe Institute (1993)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  1. 1.University of ColoradoBoulderUSA

Personalised recommendations