Bandwidth Adaptive Cache Coherence Optimizations for Chip Multiprocessors

  • Abdullah KayiEmail author
  • Olivier Serres
  • Tarek El-Ghazawi


Chip Multiprocessors (CMPs) have different technological parameters and physical constraints than earlier multi-processor systems, which should be taken into consideration when designing cache coherence protocols. Also, contemporary cache coherence protocols use invalidate schemes that are known to generate a high number of coherence misses. This is especially true under producer-consumer sharing patterns that can become a performance bottleneck as the number of cores increases. This paper presents two mechanisms to design efficient and scalable cache coherence protocols for CMPs. First, we propose an adaptive hybrid protocol to reduce coherence misses observed in write-invalidate based protocols. The proposed protocol is based on a write-invalidate scheme. However, adaptively, it can push updates to potential consumers based on observed producer-consumer sharing patterns. Secondly, we extend this adaptive protocol with an interconnection resource aware mechanism. Experimental evaluations, conducted on a tiled-CMP via full-system simulation, were used to assess the performance from our proposed dynamic hybrid protocols. Performance analysis is presented on a set of scientific applications from the SPLASH-2 and NAS parallel benchmark suites. Results showed that the proposed mechanisms reduce cache-to-cache sharing misses up to 48 % and speed up application performance up to 34 %. In addition, the proposed interconnection resource aware mechanism is proven to perform well under varying interconnection utilizations.


Cache coherence Multi-core Bandwidth adaptive 



Authors would like to thank Dan Gibson from Google, formerly at University of Wisconsin Multifacet group, for his help and suggestions on our implementations in GEMS simulation infrastructure. Authors also would like to thank Arctic Region Supercomputing Center (ARSC) for their support in this research.


  1. 1.
    NAS Parallel Benchmarks: OpenMP version developed by the Omni group
  2. 2.
    Acacio, M., González, J., García, J., Duato, J.: Owner prediction for accelerating cache-to-cache transfer misses in a cc-NUMA architecture. In: Proceedings of the 2002 ACM/IEEE Conference on Supercomputing, pp. 1–12. IEEE Computer Society Press Los Alamitos, CA, USA (2002)Google Scholar
  3. 3.
    Acacio, M.E., González, J., García, J.M., Duato, J.: The use of prediction for accelerating upgrade misses in cc-NUMA multiprocessors. In: IEEE PACT, pp. 155–164. IEEE Computer Society (2002)Google Scholar
  4. 4.
    Alam, S.R., Barrett, R.F., Kuehn, J.A., Roth, P.C., Vetter, J.S.: Characterization of scientific workloads on systems with multi-core processors. In: IISWC, pp. 225–236. IEEE (2006)Google Scholar
  5. 5.
    Anderson, C., Karlin, A.R.: Two adaptive hybrid cache coherency protocols. In: International Symposium on High-Performance Computer Architecture (HPCA), pp. 303–313 (1996)Google Scholar
  6. 6.
    Cheng, L., Carter, J.B.: Extending cc-numa systems to support write update optimizations. In: SC ’08: Proceedings of the 2008 ACM/IEEE Conference on Supercomputing, p. 30. IEEE/ACM (2008)Google Scholar
  7. 7.
    Chu, M., Ravindran, R., Mahlke, S.: Data access partitioning for fine-grain parallelism on multicore architectures. In: MICRO ’07: Proceedings of the 40th Annual IEEE/ACM International Symposium on Microarchitecture, pp. 369–380. IEEE Computer Society, Washington, DC, USA (2007). doi: 10.1109/MICRO.2007.11
  8. 8.
    Cox, A.L., Fowler, R.J.: Adaptive cache coherency for detecting migratory shared data. In: International Symposium on Computer Architecture (ISCA), pp. 98–108 (1993)Google Scholar
  9. 9.
    Dahlgren, F.: Boosting the performance of hybrid snooping cache protocols. In: ISCA ’95: Proceedings of the 22nd Annual International Symposium on Computer Architecture, pp. 60–69. ACM, New York, NY, USA (1995). doi: 10.1145/223982.223998
  10. 10.
    Dahlgren, F., Stenström, P.: Reducing the write traffic for a hybrid cache protocol. In: International Conference on Parallel Processing (ICPP), pp. 166–173 (1994)Google Scholar
  11. 11.
    Eggers, S.J., Katz, R.H.: Evaluating the performance of four snooping cache coherency protocols. SIGARCH Comput. Archit. News 17(3), 2–15 (1989). doi: 10.1145/74926.74927
  12. 12.
    Eisley, N., Peh, L.S., Shang, L.: In-network cache coherence. In: MICRO 39: Proceedings of the 39th Annual IEEE/ACM International Symposium on Microarchitecture, pp. 321–332. IEEE Computer Society, Washington, DC, USA (2006). doi: 10.1109/MICRO.2006.27
  13. 13.
    Eisley, N., Peh, L.S., Shang, L.: Leveraging on-chip networks for data cache migration in chip multiprocessors. In: PACT ’08, pp. 197–207. ACM, New York, NY, USA (2008). doi: 10.1145/1454115.1454144
  14. 14.
    Fensch, C., Cintra, M.: An OS-based alternative to full hardware coherence on tiled CMPs. In: 14th International Symposium on High Performance Computer Architecture (HPCA), pp. 355–366. IEEE (2008). doi: 10.1109/HPCA.2008.4658652
  15. 15.
    Geer, D.: Industry trends: chip makers turn to multicore processors. IEEE Comput. 38(5), 11–13 (2005)CrossRefGoogle Scholar
  16. 16.
    Ghosh, D., Carter, J.B., III, H.D.: Perceptron-based coherence predictors. In: Proceedings of 2nd Workshop on Chip Multiprocessor Memory Systems and Interconnects (CMP-MSI), in Conjunction with ISCA 2008 (2008)Google Scholar
  17. 17.
    Gorder, P.F.: Multicore processors for science and engineering. Comput. Sci. Eng. 9(2), 3–7 (2007). doi: 10.1109/MCSE.2007.35
  18. 18.
    Grahn, H.K., Stenström, P.: Evaluation of a competitive-update cache coherence protocol with migratory data detection. J. Parallel Distrib Comput 39, 39–42 (1996)CrossRefGoogle Scholar
  19. 19.
    Karlin, A.R., Manasse, M.S., Rudolph, L., Sleator, D.D.: Competitive snoopy caching. Algorithmica 3, 77–119 (1988)CrossRefMathSciNetGoogle Scholar
  20. 20.
    Kaxiras, S., Goodman, J.R.: Improving cc-NUMA performance using instruction-based prediction. In: International Symposium on High-Performance Computer Architecture (HPCA), pp. 161 (1999)Google Scholar
  21. 21.
    Kaxiras, S., Young, C.: Coherence communication prediction in shared-memory multiprocessors. In: International Symposium on High-Performance Computer Architecture (HPCA), pp. 156–167 (2000)Google Scholar
  22. 22.
    Kayi, A., Kornkven, E., El-Ghazawi, T., Al-Bahra, S., Newby, G.: Performance evaluation of clusters with ccNUMA nodes: a case study. In: HPCC ’08, pp. 320–327 (2008)Google Scholar
  23. 23.
    Lai, A.C., Falsafi, B.: Memory sharing predictor: The key to a speculative coherent dsm. In: ISCA ’99: Proceedings of the 26th, Annual International Symposium on Computer Architecture, pp. 172–183 (1999)Google Scholar
  24. 24.
    Leventhal, S., Franklin, M.: Perceptron based consumer prediction in shared-memory multiprocessors. In: ICCD 2006: International Conference on, Computer Design, pp. 148–154 (2006). doi: 10.1109/ICCD.2006.4380808
  25. 25.
    Magnusson, P.S., Christensson, M., Eskilson, J., Forsgren, D., Hållberg, G., Högberg, J., Larsson, F., Moestedt, A., Werner, B.: Simics: a full system simulation platform. IEEE Comput. 35(2), 50–58 (2002)CrossRefGoogle Scholar
  26. 26.
    Martin, M.M.K.: Formal verification and its impact on the snooping versus directory protocol debate. In: ICCD 2005: International Conference on Computer Design, pp. 543–449. IEEE Computer Society (2005)Google Scholar
  27. 27.
    Martin, M.M.K., Harper, P.J., Sorin, D.J., Hill, M.D., Wood, D.A.: Using destination-set prediction to improve the latency/bandwidth tradeoff in shared-memory multiprocessors. In: International Symposium on Computer Architecture (ISCA), pp. 206–217. IEEE Computer Society (2003)Google Scholar
  28. 28.
    Martin, M.M.K., Sorin, D.J., Hill, M.D., Wood, D.A.: Bandwidth adaptive snooping. In: International Symposium on High-Performance Computer Architecture (HPCA), pp. 251–262 (2002)Google Scholar
  29. 29.
    Marty, M.R., Bingham, J.D., Hill, M.D., Hu, A.J., Martin, M.M.K., Wood, D.A.: Improving multiple-cmp systems using token coherence. In: ISCA ’05: Proceedings of the 32nd Annual International Symposium on Computer Architecture, pp. 328–339. IEEE Computer Society (2005)Google Scholar
  30. 30.
    Moore, G.E.: Cramming more components onto integrated circuits. Electronics 38(8), 114–117 (1965)Google Scholar
  31. 31.
    Mukherjee, S.S., Hill, M.D.: Using prediction to accelerate coherence protocols. In: International Symposium on Computer Architecture (ISCA), pp. 179–190 (1998)Google Scholar
  32. 32.
    Nilsson, H., Stenström, P.: An adaptive update-based cache coherence protocol for reduction of miss rate and traffic. In: Proceedings of Parallel Architectures and Languages Europe (PARLE), pp. 363–374. Springer (1994)Google Scholar
  33. 33.
    Nilsson, J., Landin, A., Stenström, P.: The coherence predictor cache: a resource-efficient and accurate coherence prediction infrastructure. In: IPDPS ’03: Proceedings of the International Parallel and Distributed Processing Symposium, p. 10. IEEE Computer Society (2003)Google Scholar
  34. 34.
    Raghavan, A., Blundell, C., Martin, M.M.K.: Token tenure: patching token counting using directory-based cache coherence. In: MICRO, pp. 47–58. IEEE Computer Society (2008)Google Scholar
  35. 35.
    Raynaud, A., Zhang, Z., Torrellas, J.: Distance-adaptive update protocols for scalable shared-memory multiprocessors. In: HPCA ’96: Proceedings of the Second International Symposium on High-Performance Computer, Architecture, pp. 323–334 (1996). doi: 10.1109/HPCA.1996.501197
  36. 36.
    Stenström, P., Brorsson, M., Sandberg, L.: An adaptive cache coherence protocol optimized for migratory sharing. In: International Symposium on Computer Architecture (ISCA), pp. 109–118 (1993)Google Scholar
  37. 37.
    Woo, S.C., Ohara, M., Torrie, E., Singh, J.P., Gupta, A.: The SPLASH-2 Programs: characterization and methodological considerations. In: ISCA ’95, pp. 24–36 (1995)Google Scholar
  38. 38.
    Yeh, T.Y., Patt, Y.N.: Alternative implementations of two-level adaptive branch prediction. In: International Symposium on Computer Architecture (ISCA), pp. 124–134 (1992)Google Scholar

Copyright information

© Springer Science+Business Media New York 2013

Authors and Affiliations

  • Abdullah Kayi
    • 1
    Email author
  • Olivier Serres
    • 2
  • Tarek El-Ghazawi
    • 2
  1. 1.Intel PTDHillsboroUSA
  2. 2.The George Washington UniversityWashingtonUSA

Personalised recommendations