Towards a high performance neural branch predictor
Authors:
- L.N. VintanLVL.N. Vintan
- M. IridonMIM. Iridon
- 22 January 2003
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 2, 868-873
Abstract
The main aim of this short paper is to propose a new branch prediction approach called by us "neural branch prediction". We developed a first neural predictor model based on a simple neural learning algorithm, known as Learning Vector Quantization algorithm. Based on a trace driven simulation method we investigated the influences of the learning step and training processes. Also we compared the neural predictor with a powerful classical predictor and we establish that they result in close performances. Therefore, we conclude that in the nearest future it might be necessary to model and simulate other more powerful neural adaptive predictors, based on more complex neural networks architectures or even time series concepts, in order to obtain better prediction accuracies compared with the previous known schemes.Keywords
This publication has 7 references indexed in Scilit:
- The impact of cache organisation on the instruction issue rate of a superscalar processor1999
- Target prediction for indirect jumps1997
- A superscalar architecture to exploit instruction level parallelismMicroprocessors and Microsystems, 1997
- Using hybrid branch predictors to improve branch prediction accuracy in the presence of context switches1996
- Neural Network Learning and Expert Systems1993
- Improving the accuracy of dynamic branch prediction using branch correlation1992
- Alternative implementations of two-level adaptive branch prediction1992