Abstract | ||
---|---|---|
Historically, Markovian predictors have been very successful in predicting branch out- comes. 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. |
Year | Venue | Keywords |
---|---|---|
2007 | J. Instruction-Level Parallelism | neural network |
Field | DocType | Volume |
Markov process,Computer science,Prediction by partial matching,Artificial intelligence,Artificial neural network,Branch predictor | Journal | 9 |
Citations | PageRank | References |
1 | 0.36 | 1 |
Authors | ||
5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ram Srinivasan | 1 | 1 | 0.36 |
Eitan Frachtenberg | 2 | 1060 | 85.08 |
Olaf Lubeck | 3 | 157 | 19.11 |
Scott Pakin | 4 | 1098 | 134.55 |
Jeanine Cook | 5 | 29 | 4.67 |