Implementation and comparison of bi-modal dynamic branch prediction with static branch prediction schemes

Abstract

In a parallel processor, the pipeline cannot fetch the conditional instructions with the next clock cycle, leading to a pipeline stall. So, conditional instructions create a problem in the pipeline because the proper path can only be known after the branch execution. To accurately predict branches, a significant predictor is proposed for the prediction of the conditional branch instruction. Firstly, the always taken prediction scheme and always not-taken prediction scheme are applied to the trace files using a static branch prediction scheme. Further, a bi-modal dynamic branch prediction scheme is proposed, which uses the saturating counter to provide more accuracy rate than the other prediction schemes. The result suggests that the proposed bi-modal dynamic branch prediction scheme provides a higher accuracy rate than always taken and always not-taken static branch prediction by the rate of 9.82% and 6.12%, respectively. Moreover, the proposed bi-modal dynamic branch prediction limits the misprediction rate by 10.33% and 8.16% compared with the always taken and always not-taken static branch prediction scheme.