Clustered indexing for branch predictors

ct 8 As a result of resource limitations, state in branch predictors is frequently shared between uncorrelated branches. This interference 9 can significantly limit prediction accuracy. In current predictor designs, the branches sharing prediction information are determined 10 by their branch addresses and thus branch groups are arbitrarily chosen during compilation. This feasibility study explores a more ana11 lytic and systematic approach to classify branches into clusters with similar behavioral characteristics. We present several ways to incor12 porate this cluster information as an additional information source in branch predictors. 13 Our profile-based results demonstrate that cluster information is useful in various branch prediction schemes. When clustered indexing 14 is applied, the same performance can be obtained with 2–8 times less hardware budget. For small predictor budgets, clustered indexing is 15 very cost-effective, e.g., the misprediction rate in an 8 Kib gshare...

• Type: Journal
• Year: 2007
• Where: MAM
• Authors: Veerle Desmet, Hans Vandierendonck, Koen De Bosschere
• Keywords: Branch Predictors, Cluster Information, Hardware Budget, MAM 2007