Performance Analysis of Weighted Victim Cache Replacement Policy
Kartik Patel, Virendra Kumar Yadav, Uday Karra, Bheemappa Halavar · 2024
Efficient cache replacement policies are crucial for enhancing the performance of modern computing systems, particularly in Last Level Caches (LLC). Traditional policies like LRU fail to consider important factors such as reuse distance and access frequency, leading to missed optimization opportunities. This paper introduces the Weighted Victim Cache Replacement Policy (WVCRP), which combines both frequency and reuse distance in a weighted approach to minimize cache misses. The proposed method is evaluated on in-house simulator for initial evaluation against the set of generated inputs with varying cache sizes, result shows WVCRP outperformed LRU which result in reducing cache misses. Further proposed algorithm will be evaluated on different benchmarks with the integration into cache simulator(champsim) in future for detailed analysis.