BIB-VERSION:: CS-TR-v2.0
          ID:: STAN//CSL-TR-92-553
       ENTRY:: November 02, 1994
ORGANIZATION:: Stanford University, Computer Systems Laboratory
       TITLE:: BRANCH PREDICTION USING LARGE SELF HISTORY
        TYPE:: Technical Report
      AUTHOR:: Johnson, John D.
        DATE:: December 1992
       PAGES:: 31
    ABSTRACT:: Branch prediction is the main method of providing speculative
               opportunities for new high performance processors, therefore
               the accuracy of branch prediction is becoming very important.
               Motivated by this desire to achieve high levels of branch
               prediction, this study examines methods of using up to 24
               bits branch direction history to determine the probable
               outcome of the next execution of a conditional branch. Using
               profiling to train a prediction logic function achieves an
               average branch prediction accuracy of up to 96.9% for the six
               benchmarks used in this study.
       NOTES:: [Adminitrivia V1/Prg/19941102]
         END:: STAN//CSL-TR-92-553