What people are saying - Write a review
We haven't found any reviews in the usual places.
Analysis of Existing Address Prediction Techniques
4 other sections not shown
8KB direct-mapped addr unchanged address calculation latency address transformation function address transforming calculation ancestor loads associated average bandwidth base register benchmarks branch prediction bundle-based address predictor bzip2 cache miss chapter conditional branches context predictor control flow cycle data cache decoupled address dependent load destination register direct-mapped cache dynamic loads fetch Figure fixed access latency function identifier function lookup table global data gzip hardware hash hashing function hybrid address predictor hybrid predictor hysteresis counter Id traffic implementing input value instance instruction address instruction window load balancing load instruction load latency load's address memory dependencies memory disambiguation memory hierarchy misprediction number of unique partitioned cache pipeline potential precomputation prediction accuracy prediction lookup table prefetching processor reduce result scheme second level cache set-associative static load static memory instructions static stores store buffer stride predictor stride-based superscalar transformation dependence graph transformation graph unique cache lines unpartitioned update latency variable access latency wire delay