Neuro-PPM Branch Prediction

DMCA

Neuro-PPM Branch Prediction

Cached

Download Links

by Ram Srinivasan , Eitan Frachtenberg , Olaf Lubeck , Scott Pakin , Jeanine Cook

BibTeX

@MISC{Srinivasan_neuro-ppmbranch,
    author = {Ram Srinivasan and Eitan Frachtenberg and Olaf Lubeck and Scott Pakin and Jeanine Cook},
    title = {Neuro-PPM Branch Prediction},
    year = {}
}

OpenURL

Abstract

Historically, Markovian predictors have been very successful in predicting branch outcomes. In this work we propose a hybrid scheme that employs two Prediction by Partial Matching (PPM) Markovian predictors, one that predicts based on local branch histories and one based on global branch histories. The two independent predictions are combined using a neural network. On the CBP-2 traces the proposed scheme achieves over twice the prediction accuracy of the gshare predictor. 1.

Keyphrases

neuro-ppm branch prediction markovian predictor prediction accuracy branch outcome local branch history partial matching proposed scheme hybrid scheme independent prediction gshare predictor neural network global branch history

DMCA