Advertisement

An evaluation of set-associativity in two-level caches for shared memory multiprocessors

  • Senthil Krishnamoorthy
  • Alok Choudhary
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 605)

Abstract

In this paper we study the behavior of a two level cache hierarchy for multiprocessors, where the processors share a common second level cache. Cache performance is studied as a function of set associativity of primary and second level caches, using trace driven simulations. The simulation uses numerical benchmarks compiled using a parallelizing and vectorizing compiler. Simulation study shows that a replacement scheme based on least recently used blocks gives better performance than a replacement scheme based on multilevel inclusion. Invalidations to ensure multilevel inclusion in the two level cache hierarchy do not significantly affect cache performance and can be further reduced by increasing set associativity of second level cache. Furthermore, the study shows that coherency traffic and cache interference are sensitive to set associativity of the cache. In spite of reduction in coherency traffic, improvement in performance with set associativity is largely due to increase in hit ratio of primary and second level caches.

Keywords

Cache Size Cache Performance Level Cache Cache Controller Primary Cache 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    A. Agarwal et al. APRIL: A Processor Architecture for Multiprocessing. In Proc. 17th Ann. International Symposium Computer Architecture, pages 104–114, 1990.Google Scholar
  2. [2]
    Alliant Computer Systems Corporation, FX/Series Product Summary. 1987.Google Scholar
  3. [3]
    J. Archibald and J.-L. Baer. Cache Coherence Protocols: Evaluation Using a Multiprocessor Simulation Model. ACM Transactions Computer Systems, 4(4):273–298, Nov. 1986.CrossRefGoogle Scholar
  4. [4]
    J.-L. Baer and W.-H. Wang. On the Inclusion Properties for Multi-Level Cache Hierarchies. In Proc. 15th Ann. International Symposium Computer Architecture, pages 73–80, 1988.Google Scholar
  5. [5]
    M. Carlton and A. Despain. Mutiple-Bus Shared Memory System: Aquarius Project. pages 80–83, June 1990.Google Scholar
  6. [6]
    D. Cheriton, H.A. Goosen, and P.D. Boyle. Multi-level Shared Caching Techniques for Scalability in VMP-MC. In Proc. 16th Ann. International Symposium Computer Architecture, pages 16–24, 1989.Google Scholar
  7. [7]
    A. N. Choudhary and S. Krishnamoorthy. Experimental Evaluation of Multilevel Caches for Shared Memory Multiprocessors. In 1991 Hawaii International Conference on Systems and Sciences, pages 409–420, Jan. 1991.Google Scholar
  8. [8]
    J. Goodman and P. Woest. The Wisconsin Multicube: A New Large-Scale Cache Coherent Multiprocessor. In Proc. 15th Ann. International Symposium Computer Architecture, pages 422–431, 1988.Google Scholar
  9. [9]
    A. W. Wilson Jr. Hierarchical Cache/Bus Architecture for Shared Memory Multiprocessors. In Proc. 14th Ann. International Symposium Computer Architecture, pages 244–252, 1987.Google Scholar
  10. [10]
    J.W.C.Fu and J.H.Patel. Data Prefetching in Multiprocessor Vector Cache Memories. In Proc. 18th Ann. International Symposium Computer Architecture, pages 54–65, 1991.Google Scholar
  11. [11]
    L. Kipp. Perfect Benchmarks Documentation. Technical report, Center for Supercomputing Research and Development, University of Illinois at Urbana-Champaign, Feb. 1990.Google Scholar
  12. [12]
    D. Kuck, E. Davidson, D. Lawrie, and A. Sameh. Parallel Supercomputing Today and the Cedar Approach. Science, 231:967–974, 1986.Google Scholar
  13. [13]
    D. Lenoski et al. The Directory-Based Cache Coherence Protocol for the DASH Multiprocessor. In Proc. 17th Ann. International Symposium Computer Architecture, pages 148–159, 1990.Google Scholar
  14. [14]
    PARALIN:FX/LINPACK. User's Guide.Google Scholar
  15. [15]
    S. Przybylski. Cache and Memory Hierarchy — A Performance Directed Approach. Morgan-Kauffman Publisher, 1990.Google Scholar
  16. [16]
    R.T. Short and H.M. Levy. A Simulation Study of Two-Level Caches. In Proc. 15th Ann. International Symposium Computer Architecture, pages 81–89, 1988.Google Scholar
  17. [17]
    Z. G. Vranesic, M. Stumm, D. M. Lewis, and R. White. Hector: A Hierarchically Structured Shared-Memory Multiprocessor. 24(1):72–79, 1991.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1992

Authors and Affiliations

  • Senthil Krishnamoorthy
    • 1
  • Alok Choudhary
    • 1
  1. 1.Department of Electrical and Computer EngineeringSyracuse UniversitySyracuse

Personalised recommendations