Value
prediction techniques have been widely-researched in recent years to predict
the results of register writing instructions in CPUs. In this project, we
apply these techniques to a new domain: trace file compression. Since the
files we are interested in contain sequences of instruction addresses and their
corresponding load values, value predictors are ideal for this purpose
because they are known to yield a high percentage
of correct predictions. We compress the trace files by encoding the
sequences of binary values in the files with predictor numbers. To
decompress, we run the compressed file through the same set of predictors in
the order specified in the file to recreate the original values. Our
approach delivers substantially better compression rates than existing
commercial software such as Winzip and gzip.