A Cache Replacement Policy for Improving the Performance of Last Level Cache in Processors
0
조회
0
다운로드
3
인용
0
참고문헌
키워드
초록
Reduction in cache miss rates continues to be an important issue in processor design, especially at the last level cache(LLC). The Least Recently Used(LRU) replacement policy has been widely adapted in processors for the past decades. The LRU replacement policy represents the cache blocks in a set as LRU stack. So, it only use the recency information. And it is expensive to implement in hardware. Also in some workload, it raises the thrashing, i.e., blocks with high reuse evicting each other from the cache. We propose an new replacement policy to solve problems of LRU replacement policy. In the proposed policy, a incoming block is inserted at the bottom of the set. To exploit the frequency information, a hit block exchanges its position with its adjacent block above in the set. A victim block can be chosen from bottom. Selecting the bottom block as the victim makes the proposed policy to protect the cache from polluting by less frequently used blocks. It also proposes to divide the blocks in a cache set into groups to resolve the thrashing. The victim block is selected alternately from each group. The proposed policy reduces the average MPKI(Miss Per Kilo Instructions) of the baseline 1MB 16-way LLC cache using LRU replacement policy by 12.9%. It reduces the storage requirement by 47% compared to LRU replacement policy.
Abstract
Reduction in cache miss rates continues to be an important issue in processor design, especially at the last level cache(LLC). The Least Recently Used(LRU) replacement policy has been widely adapted in processors for the past decades. The LRU replacement policy represents the cache blocks in a set as LRU stack. So, it only use the recency information. And it is expensive to implement in hardware. Also in some workload, it raises the thrashing, i.e., blocks with high reuse evicting each other from the cache. We propose an new replacement policy to solve problems of LRU replacement policy. In the proposed policy, a incoming block is inserted at the bottom of the set. To exploit the frequency information, a hit block exchanges its position with its adjacent block above in the set. A victim block can be chosen from bottom. Selecting the bottom block as the victim makes the proposed policy to protect the cache from polluting by less frequently used blocks. It also proposes to divide the blocks in a cache set into groups to resolve the thrashing. The victim block is selected alternately from each group. The proposed policy reduces the average MPKI(Miss Per Kilo Instructions) of the baseline 1MB 16-way LLC cache using LRU replacement policy by 12.9%. It reduces the storage requirement by 47% compared to LRU replacement policy.
참고문헌 (12)
- M. K. Qureshi, "Adaptive insertion policies for high-performance caching", In ISCA-07, 35(2), pp.381~391 (2007)
- H. Dybdahl, "An LRU-based replacement algorithm augmented with frequency of access in shared chip-multiprocessor caches", MEDEA-06, pp.45~52 (2006)
- S. Jiang, "LIRS: An efficient low inter-reference recency set replacement policy to improve buffer cache performance", In Proceedings of the ACM SIGMETRICS Conference on Measurement and Modeling of Computer Systems, 30(1), pp.31~42 (2002)
- W. Wong, "Modified LRU policies for improving second-level cache behavior", In HPCA-2000, pp.49~60 (2000)
- D. A. Jiménez, "Insertion and promotion for tree-based pseudoLRU last-level caches", MICRO-13, pp.284~296 (2013)
- M. K. Qureshi, "A case for MLP-aware cache replacement", In ISCA-06, 34(2), pp.167~178 (2006)
- R. Subramanian, "Adaptive caches: Effective shaping of cache behavior to workloads", In MICRO-06, pp.385~396 (2006)
- S. Gupta, "Adaptive Cache Bypassing for Inclusive Last Level Caches", In IPDPS-2013, pp.1243~1254 (2013)
- J. T. Robinson, "Data cache management using frequency-based replacement", SIGMETRICS '90, 18(1), pp.134~142 (1990)
- D. Lee, "LRFU : A Spectrum of Policies that Subsumes LRU and LFU Policies", IEEE Trans. on Computer, 50(12), pp.1352~1361 (2001)
- The simplescalar tool set, "D. Burger", Univ. of Wisconsin-Madison, ~, pp.#1342 (1997)
- "SimPoint home page"[URL]