Cached Two-Level Adaptive Branch Predictors with Multiple Stages

  • Colin Egan
  • Gordon Steven
  • Lucian Vintan
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2299)


In this paper, we quantify the performance of a novel family of multi-stage Two-Level Adaptive Branch Predictors. In each two- level predictor, the PHT of a conventional Two-level Adaptive Branch Predictor is replaced by a Prediction Cache. Unlike a PHT, a Prediction Cache saves only relevant branch prediction information. Furthermore, predictions are never based on uninitialised entries and interference between branches is eliminated. In the case of a Prediction Cache miss in the first stage, our two-stage predictors use a default two-bit prediction counter stored in a second stage. We demonstrate that a two- stage Cached Predictor is more accurate than a conventional two-level predictor and quantify the crucial contribution made by the second prediction stage in achieving this high accuracy. We then extend our Cached Predictor by adding a third stage and demonstrate that a Three- Stage Cached Predictor further improves the accuracy of cached predictors.


Cache Size Branch Prediction Prediction Stage History Register Branch Predictor 
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.
    Yeh, T. and Patt, Y. N.: Two-Level Adaptive Training Branch Prediction, Micro-24, Albuquerque, New Mexico, (1991), 51–61.Google Scholar
  2. 2.
    Pan, S.; So, K. and Rahmeh, J. T.: Improving the Accuracy of Dynamic Branch Prediction Using Branch Correlation, ASPLOS-V, Boston, (1992), 76–84.Google Scholar
  3. 4.
    Kessler, R. E.: The Alpha 21264 Microprocessor, IEEE Micro, (1999), 24–36.Google Scholar
  4. 5.
    Egan, C.: Dynamic Branch Prediction in High Performance Superscalar Processors, PhD thesis, University of Hertfordshire, (2000).Google Scholar
  5. 6.
    Steven, G. B.; Egan, C.; Quick, P. and Vintan, L.: A Cost Effective Cached Correlated Two-level Adaptive Branch Predictor, 18th IASTED International Conference on Applied Informatics (AI 2000), Innsbruck, (2000).Google Scholar
  6. 7.
    Yeh, T. and Patt, P.: Alternative Implementations of Two-Level Adaptive Branch Prediction, ISCA-19, Gold Coast, Australia, (1992), 124–134.Google Scholar
  7. 8.
    Yeh, T. and Patt, Y. N.: A Comparison of Dynamic Branch Predictors that use Two Levels of Branch History, ISCA-20, (1993), 257–266.Google Scholar
  8. 9.
    McFarling, S.: Combining Branch Predictors, Western Research Laboratories Technical Report TN-36, (1993).Google Scholar
  9. 10.
    Chang, P., Hao, E., Yeh. T. and Patt, Y. N.: Branch Classification: A New Mechanism for Improving Branch Predictor Performance, Micro-27, San Jose, California, (1994), 22–31.Google Scholar
  10. 11.
    Lee, C. C., Chen, I. K. and Mudge, T. N.: The Bi-Mode Branch Predictor, Micro-30, Research Triangle Park, North Carolina, (1997), 4–13.Google Scholar
  11. 12.
    Sprangle, E., Chappell, R. S., Alsup, M. and Patt, Y. N.: The Agree Predictor: A Mechanism for Reducing Negative Branch History Interference, ISCA’ 24, Denver, Colorado, (1997), 284–291.Google Scholar
  12. 13.
    Chen, I. K., Coffey, J. T. and Mudge, T. N.: Analysis of Branch Prediction via Data Compression, ASPLOS VII, (1996), 128–137.Google Scholar
  13. 14.
    Steven, G., Anguera, R., Egan, C., Steven, F. and Vintan, L.: Dynamic Branch Prediction Using Neural Networks, DSD 2001, Poland, (2001), 178–185.Google Scholar
  14. 15.
    Chang, L, Hao E. and Patt, Y. N.: Alternative Implementations of Hybrid Branch Predictors, Micro-29, Ann Arbor, Michigan, (1995), 252–257.Google Scholar
  15. 16.
    Steven, G. B., Christianson, D. B., Collins, R., Potter, R. D. and Steven, F. L.: A Superscalar Architecture to Exploit Instruction Level Parallelism,Microprocessors and Microsystems, 20 (7), (1997), 391–400.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • Colin Egan
    • 1
  • Gordon Steven
    • 1
  • Lucian Vintan
    • 2
  1. 1.University of HertfordshireHatfieldUK
  2. 2.University “Lucian Braga” of SibiuRomania

Personalised recommendations