Analysis of a Shared Way Set Associative Cache
Resumo
A new cache memory organization called Shared Way Set Associative (SWSA) is analyzed in this work. The SWSA cache uses two memory banks with different sizes. A placement policy that uses bit map indexing is given for SWSA caches as well as an expression for the associativity. Different LRU related replacement policies are analyzed. The performance of SWSA caches was measured for the SPEC95 benchmarks using trace driven simulations. It was found that SWSA caches of various associatives usually perform better than two-way set associative caches of equivalent size. An analysis of the results using the D3S model is done. Also SWSA caches are compared with victim caches. An additional benefit of SWSA caches is that non power of two total cache sizes may be used. Hence, when SWSA caches are used, a very precise adjustment of the total cache size needed for a given workload is possible. For the SPEC95 benchmarks, savings of 33 percent for the total cache size are obtained with SWSA caches relative to two-way set associative caches that produce nearly the same miss rate.
Palavras-chave:
Cache, associativity, replacement policy
Referências
Belady, L. A.. A Study of Replacement Algorithms for a Virtual Storage Computer. IBM Systems Journal, Vol 5, N2, 1966.
Belady, L. A.; Nelson R. A.; Shidler, G. S.. An Anomaly in Space-Time Characteristics of Certain Programs Running in a Paging Environment. Communications of the ACM, p. 349-353, December 1969.
Hamkalo, J. L.; Cernuschi-Frías, B.. A Taxonomy for Cache Memory Misses.. In: Proceedings of the 11th Symposium on Computer Architecture and High Performance Computing. Natal, Brazil, 1999, p. 67-73.
Hill, M. D.. A Case for Direct-Mapped Caches.. IEEE Computer, Vol 21, n.12, p. 25-40, 1998.
Hill, M. D.; Smith, A. J.. Evaluating Associativity in CPU Caches. IEEE Trans. on Computer, C-38, n.12, p. 1612-1630, 1995.
Jouppi, P. N.. Improving Direct-Mapped Cache Performance by the Addition of a Small Fully-Associative Cache and Prefetch Buffer. In: Proceedings of the ACM INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE, 1990. p. 364-373.
Patterson, D. A.; Hennessy, J. L.. Computer Architecture: A Quantitative Approach, San Mateo, California: Morgan Kaufmann Publishers, 1995.
Smith, A. J. Cache Memories. ACM Computing Surveys, p. 473-530. Sept. 1982.
Srivastava, A.; Eustace, A. ATOM: A System for Building Customized Program Analysis Tools. In: Proceedings of the ACM CONFERENCE ON PROGRAMMING LANGUAGE DESIGN AND IMPLEMENTATION, 1994. p. 196-205.
Belady, L. A.; Nelson R. A.; Shidler, G. S.. An Anomaly in Space-Time Characteristics of Certain Programs Running in a Paging Environment. Communications of the ACM, p. 349-353, December 1969.
Hamkalo, J. L.; Cernuschi-Frías, B.. A Taxonomy for Cache Memory Misses.. In: Proceedings of the 11th Symposium on Computer Architecture and High Performance Computing. Natal, Brazil, 1999, p. 67-73.
Hill, M. D.. A Case for Direct-Mapped Caches.. IEEE Computer, Vol 21, n.12, p. 25-40, 1998.
Hill, M. D.; Smith, A. J.. Evaluating Associativity in CPU Caches. IEEE Trans. on Computer, C-38, n.12, p. 1612-1630, 1995.
Jouppi, P. N.. Improving Direct-Mapped Cache Performance by the Addition of a Small Fully-Associative Cache and Prefetch Buffer. In: Proceedings of the ACM INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE, 1990. p. 364-373.
Patterson, D. A.; Hennessy, J. L.. Computer Architecture: A Quantitative Approach, San Mateo, California: Morgan Kaufmann Publishers, 1995.
Smith, A. J. Cache Memories. ACM Computing Surveys, p. 473-530. Sept. 1982.
Srivastava, A.; Eustace, A. ATOM: A System for Building Customized Program Analysis Tools. In: Proceedings of the ACM CONFERENCE ON PROGRAMMING LANGUAGE DESIGN AND IMPLEMENTATION, 1994. p. 196-205.
Publicado
24/10/2000
Como Citar
HAMKALO, José Luis; CERNUSCHI-FRÍAS, Bruno.
Analysis of a Shared Way Set Associative Cache. In: INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE AND HIGH PERFORMANCE COMPUTING (SBAC-PAD), 12. , 2000, São Pedro/SP.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2000
.
p. 57-64.
DOI: https://doi.org/10.5753/sbac-pad.2000.41204.
