LRFU: A Spectrum of Policies that Subsumes the Least Recently Used and Least Frequently Used Policies

By Donghee Lee, Jongmoo Choi, Jong-Hun Kim, Sam H. Noh, S. Min, Yookun Cho, Chong-Sang Kim. IEEE Trans. Computers, 2001

641 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Distributed Hierarchical GPU Parameter Server for Massive Scale Deep Learning Ads Systems (2020, methodology): “Many caching policies have been developed for storage systems, such as the LRU-3 K (O’Neil et al., 1993), DBMIN (Chou & DeWitt, 1986), LRFU(Lee et al., 2001), and Semantic Caching (Dar et al., 1996).”
SHiP: Signature-based Hit Predictor for high performance caching (2011, methodology): “While we cannot cover all innovations in cache replacement research [1, 2, 3, 6, 10, 11, 14, 16, 18, 19, 20, 24, 26, 27, 28, 30, 31], we summarize prior art that closely resembles SHiP.”
H2O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Models (2023, positive): “Conventional approaches such as Least Recently Used and Least Frequently Used (O’neil et al., 1993; Lee et al., 2001) prioritize the recency and frequency of data access.”
A unified approach to the performance analysis of caching systems (2013, background): “Other g eneralizations/extensions of LRU known as LRU-2Q, LRUk and LRFU have been proposed in [20], [31] and [32] respective ly.”