Paper Info
Title
HyComp: a hybrid cache compression method for selection of data-type-specific compression methods
Abstract
Proposed cache compression schemes make design-time assumptions on value locality to reduce decompression latency. For example, some schemes assume that common values are spatially close whereas other schemes assume that null blocks are common. Most schemes, however, assume that value locality is best exploited by fixed-size data types (e.g., 32-bit integers). This assumption falls short when other data types, such as floating-point numbers, are common. This paper makes two contributions. First, HyComp -- a hybrid cache compression scheme -- selects the best-performing compression scheme, based on heuristics that predict data types. Data types considered are pointers, integers, floating-point numbers and the special (and trivial) case of null blocks. Second, this paper contributes with a compression method that exploits value locality in data types with predefined semantic value fields, e.g., as in the exponent and the mantissa in floating-point numbers. We show that HyComp, augmented with the proposed floating-point-number compression method, offers superior performance in comparison with prior art.
Year
DOI
Venue
2015
10.1145/2830772.2830823
Proceedings of the 48th International Symposium on Microarchitecture
Keywords
Field
DocType
huffman coding
Locality of reference,CPU cache,Computer science,Cache,Parallel computing,Cache algorithms,Real-time computing,Data type,Data compression,Image compression,Lossless compression
Conference
ISBN
Citations 
PageRank 
978-1-5090-6601-8
9
0.44
References 
Authors
21
3
Name
Order
Citations
PageRank
Angelos Arelakis1441.62
fredrik dahlgren290.44
Per Stenström33048234.09