Quality Evaluation Strategies for Approximate Computing in Embedded Systems

  • Olaf NeugebauerEmail author
  • Peter Marwedel
  • Roland Kühn
  • Michael Engel
Conference paper
Part of the IFIP Advances in Information and Communication Technology book series (IFIPAICT, volume 499)


The quest for increased performance at lower energy consumption rates, especially in embedded systems used in smart systems, has reached physical limits that can no longer be exploited using traditional optimization techniques. One popular way to achieve additional gains is to intentionally perform inaccurate computations. Our framework eases evaluation, analysis and comparison of approximation techniques in terms of energy consumption, run time, quality and user-defined criteria. Applied to a set of benchmarks, we obtain valuable insights into related side effects, including increased file sizes indicating that a careless utilization of approximate computing threatens its usefulness.


Approximate computing Energy consumption Quality metrics 



The authors like to thank the German Research Foundation (DFG) for supporting part of this work within the Collaborative Research Center SFB 876 “Providing Information by Resource-Constrained Data Analysis”, projects A3 and B2,


  1. 1.
    Akturk, I., Khatamifard, K., Karpuzcu, U.R.: On quantification of accuracy loss in approximate computing. In: Proceedings of WDDD (2015)Google Scholar
  2. 2.
    Baek, W., Chilimbi, T.M.: Green: A framework for supporting energy-conscious programming using controlled approximation. In: Proceedings of PLDI (2010)Google Scholar
  3. 3.
    Bienia, C., Kumar, S., Singh, J.P., Li, K.: The PARSEC benchmark suite: Characterization and architectural implications. In: Proceedings of PACT (2008)Google Scholar
  4. 4.
    Gupta, V., Mohapatra, D., Park, S.P., Raghunathan, A., Roy, K.: IMPACT: IMPrecise adders for low-power approximate computing. In: Proceedings of ISLPED (2011)Google Scholar
  5. 5.
    Image Processing Place, December 2016.
  6. 6.
    Independent JPEG Group, December 2016.
  7. 7.
    Kedem, Z., Mooney, V.J., Muntimadugu, K.K., Palem, K.V., Devarasetty, A., Parasuramuni, P.D.: Optimizing energy to minimize errors in dataflow graphs using approximate adders. In: Proceedings of CASES (2010)Google Scholar
  8. 8.
    Lin, W., Jay Kuo, C.C.: Perceptual visual quality metrics: A survey. J. Vis. Commun. Image Represent. 22(4), 297–312 (2011)CrossRefGoogle Scholar
  9. 9.
    Liu, J., Jaiyen, B., Veras, R., Mutlu, O.: RAIDR: Retention-aware intelligent DRAM refresh. In: Proceedings of ISCA, Vol. 40 (2012)Google Scholar
  10. 10.
    Misailovic, S., Sidiroglou, S., Hoffmann, H., Rinard, M.: Quality of service profiling. In: Proceedings of ICSE, vol. 1 (2010)Google Scholar
  11. 11.
    Mittal, S.: A survey of techniques for approximate computing. ACM Comput. 48(4), 62 (2016)Google Scholar
  12. 12.
    Neugebauer, O., Libuschewski, P., Engel, M., Müller, H., Marwedel, P.: Plasmon-based virus detection on heterogeneous embedded systems. In: Proceedings of SCOPES (2015)Google Scholar
  13. 13.
    Renganarayana, L., Srinivasan, V., Nair, R., Prener, D.: Programming with relaxed synchronization. In: Proceedings of RACES (2012)Google Scholar
  14. 14.
    Sampson, A., Dietl, W., Fortuna, E., Gnanapragasam, D., Ceze, L.: EnerJ: approximate data types for safe and general low-power computation. In: Proceedings of PLDI (2011)Google Scholar
  15. 15.
    Sidiroglou, S., Misailovic, S., Hoffmann, H., Rinard, M.: Managing performance vs. accuracy trade-offs with loop perforation. In: Proceedings of ESCE/FSE (2011)Google Scholar
  16. 16.
    Wang, Z., Bovik, A.C.: A universal image quality index. IEEE Sig. Process. Lett. 9(3), 81–84 (2002)CrossRefGoogle Scholar
  17. 17.
    Wang, Z., Bovik, A.C., Sheikh, H.R., Simoncelli, E.P.: Image quality assessment: From error visibility to structural similarity. IEEE Trans. Image Process. 13(4), 600–612 (2004)CrossRefGoogle Scholar
  18. 18.
    Xiph.Org Foundation, December 2016.

Copyright information

© IFIP International Federation for Information Processing 2017

Authors and Affiliations

  • Olaf Neugebauer
    • 1
    Email author
  • Peter Marwedel
    • 1
  • Roland Kühn
    • 1
  • Michael Engel
    • 2
  1. 1.TU Dortmund UniversityDortmundGermany
  2. 2.Coburg University of Applied Sciences and ArtsCoburgGermany

Personalised recommendations