Blast from the Past: Least Expected Use (LEU) Cache Replacement with Statistical History
Cache replacement policies typically use some form of statistics on past access behavior. As a common limitation, however, the extent of the history being recorded is limited to either just the data in cache or, more recently, a larger but still finite-length window of accesses, because the cost of keeping a long history can easily outweigh its benefit.
This paper presents a statistical method to keep track of instruction pointer-based access reuse intervals of arbitrary length and uses this information to identify the Least Expected Use (LEU) blocks for replacement. LEU uses dynamic sampling supported by novel hardware that maintains a state to record arbitrarily long reuse intervals. LEU is evaluated using the Cache Replacement Championship simulator, tested on PolyBench and SPEC, and compared with five policies including a recent technique that approximates optimal caching using a fixed-length history. By maintaining statistics for an arbitrary history, LEU outperforms previous techniques for a broad range of scientific kernels, whose data reuses are longer than those in traces traditionally used in computer architecture studies.
Sun 18 JunDisplayed time zone: Eastern Time (US & Canada) change
16:00 - 17:20
|Blast from the Past: Least Expected Use (LEU) Cache Replacement with Statistical History|
Sayak Chakraborti University of Rochester, Zhizhou (Chris) Zhang Uber Technologies, Noah Bertram Cornell University, Sandhya Dwarkadas University of Rochester, Chen Ding University of RochesterDOI
|OMRGx: Programmable and Transparent Out-of-Core Graph Partitioning and Processing|
|ZipKV: In-Memory Key-Value Store with Built-In Data Compression|
Linsen Ma Rensselaer Polytechnic Institute, Rui Xie Rensselaer Polytechnic Institute, Tong Zhang Rensselaer Polytechnic InstituteDOI
|Flexible and Effective Object Tiering for Heterogeneous Memory Systems|
Brandon Kammerdiener University of Tennessee, Jeffrey Zachariah McMichael University of Tennessee, Michael Jantz University of Tennessee, Kshitij Doshi Intel Corporation, Terry Jones Oak Ridge National LaboratoryDOI