Compaction-free compressed cache for high performance multi-core system

Abstract

Compressed cache was used in shared last level cache (LLC) to increase the effective capacity. However, because of various data compression sizes, fragmentation problem of storage is inevitable in this cache design. When it happens, usually, a compaction process is invoked to make contiguous storage space. This compaction process induces extra cycle penalty and degrades the effectiveness of compressed cache design. In this paper, we propose a compaction-free compressed cache architecture which can completely eliminate the time for executing compaction. Based on this cache design, we demonstrate that our results, compared with the conventional cache, have system performance improvement by 16% and energy reduction by 16%. Compared with the work by Alameldeen et al. [1], our design has 5% more performance improvement and 3% more energy reduction. Compared with the work by Sardashti et al. [2], our design has 3% more performance improvement and 2% more energy reduction.