Improving Branch Prediction Performance by Removing Temporally Close Aliases

摘要

In this paper, we propose a new branch prediction technique to significantly reduce the more likely destructive aliases by caching up recurring aliasing branches that occur within a small temporal locality with one another. Our pre-simulation analysis on traces supports such a claim showing that aliasing branches with a small temporal locality account for majority of all aliases. We further show that such aliases are more likely to lead to destructive prediction result. Our technique, incorporated with a small additional "alias table" with its size much smaller than the existing prediction table normally used, is capable of eliminating most of aliases, especially the highly performance-degrading repetitive "local" aliases.