Focused Prefetching: Performance Oriented Prefetching Based On Commit Stalls

Loads that miss in L1 or L2 caches and waiting for their data at the head of the ROB cause signicant slow down in the form of commit stalls. We identify that most of these commit stalls are caused by a small set of loads, referred to as LIMCOS (Loads Incurring Majority of COmmit Stalls). We propose simple history-based classiers that track com-mit stalls suered by loads to help us identify this small set of loads. We study an application of these classiers to prefetching. The classiers are used to train the prefetcher to focus on the misses suered by LIMCOS. This, referred to as focused prefetching, results in a 9.8 % gain in IPC over naive GHB based delta correlation prefetcher along with a 20.3 % reduc-tion in memory trac for a set of 17 memory-intensive SPEC2000 benchmarks. Another important impact of fo-cused prefetching is a 61 % improvement in the accuracy of prefetches. We demonstrate that the proposed classi-cation criterion performs better than other existing criteria like criticality and delinquent loads. Also we show that the criterion of focusing on commit stalls is robust enough across cache levels and can be applied to any prefetcher without any modi cations to the prefetcher