Pseudo-randomly interleaved memory
ISCA '91 Proceedings of the 18th annual international symposium on Computer architecture
Symbolic Boolean manipulation with ordered binary-decision diagrams
ACM Computing Surveys (CSUR)
Algebraic decision diagrams and their applications
ICCAD '93 Proceedings of the 1993 IEEE/ACM international conference on Computer-aided design
The design and performance of a conflict-avoiding cache
MICRO 30 Proceedings of the 30th annual ACM/IEEE international symposium on Microarchitecture
MediaBench: a tool for evaluating and synthesizing multimedia and communicatons systems
MICRO 30 Proceedings of the 30th annual ACM/IEEE international symposium on Microarchitecture
Investigating optimal local memory performance
Proceedings of the eighth international conference on Architectural support for programming languages and operating systems
ACM Computing Surveys (CSUR)
Improved indexing for cache miss reduction in embedded systems
Proceedings of the 40th annual Design Automation Conference
IEEE Transactions on Computers
Reducing cache misses by application-specific re-configurable indexing
Proceedings of the 2004 IEEE/ACM International conference on Computer-aided design
MiBench: A free, commercially representative embedded benchmark suite
WWC '01 Proceedings of the Workload Characterization, 2001. WWC-4. 2001 IEEE International Workshop
Application-specific reconfigurable XOR-indexing to eliminate cache conflict misses
Proceedings of the conference on Design, automation and test in Europe: Proceedings
Efficient address mapping of shared cache for on-chip many-core architecture
EuroPar'10 Proceedings of the 16th international Euro-Par conference on Parallel processing: Part I
Hi-index | 0.00 |
Stringent power and performance constraints, coupled with detailed knowledge of the target applications of a processor, allows for application-specific processor optimizations. It has been shown that application-specific reconfigurable hash functions eliminate a large number of cache conflict misses. These hash functions minimize conflicts by modifying the mapping of cache blocks to cache sets. This paper describes an algorithm to compute optimal XORfunctions, a particular type of hash functions based on XORs. Using this algorithm, we set an upper bound on the conflict reduction achievable with XOR-functions. We show that XOR-functions perform better than other reconfigurable hash functions studied in the literature such as bit-selecting functions. The XOR-functions are optimal for one particular execution of a program. However, we show that optimal XOR-functions are less sensitive to the characteristics of the execution than optimal bit-selecting hash functions. This again underlines that XOR-functions are the best known hash functions to implement reconfigurable hash functions.