한국통신학회논문지 (The Journal of Korean Institute of Communications and Information Sciences)

한국통신학회 (The Korean Institute of Commucations and Information Sciences)
권호 표지
DOI QR코드

DOI QR Code

시스템 성능 및 버스 트래픽에 대한 트랜잭셔널 메모리의 충돌 관리 정책 영향 분석

Analysis of the Influence of the Conflict Management Policy of the Transactional Memory on the System Performance and Bus Traffic

  • 김영규 (경북대학교 전자전기컴퓨터학부) ;
  • 문병인 (경북대학교 IT대학 전자공학부)
  • 투고 : 2012.04.10
  • 심사 : 2012.11.06
  • 발행 : 2012.11.30
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

공유메모리 멀티프로세서 시스템에서, lock을 사용하는 전통적인 동기화 방식의 문제점들을 극복하기 위하여 트랜잭셔널 메모리(transactional memory)가 제안되었고, 고성능 트랜잭셔널 메모리를 실용화하기 위한 다양한 구현 방법들이 계속해서 연구되고 있다. 하지만 이러한 연구들은 트랜잭셔널 메모리의 실용화 및 수행 속도 개선에 주력하고 있으며, 충돌 관리 정책(conflict management policy)에 따른 트랜잭셔널 메모리의 시스템 오버헤드를 분석하는 연구는 부족한 실정이다. 이에 본 논문은 트랜잭셔널 메모리의 한 종류인 하드웨어 트랜잭셔널 메모리를 충돌 관리 정책에 따라 네 가지로 분류하고, 모델링과 시뮬레이션을 통해 이 네 가지의 성능과 시스템 버스 트래픽을 비교 분석한다. 그리고 이러한 비교 분석 결과를 바탕으로 시스템 성능에 가장 크게 기여 할 수 있는 효율적인 충돌 관리 정책을 제시한다.

The transactional memory was proposed to solve the problems of the conventional lock-based synchronization methods in the shared memory multiprocessor system. Various implementation methods for putting the high performance transactional memory to practical use have been continuously studied. However, these studies focus only on the commercialization and performance enhancement of the transactional memory. Besides, there have been few studies to analyze the system overhead of the transactional memory according to the conflict management policy. Thus this paper classifies hardware transactional memory, which is one kind of transactional memories, into four types according to the conflict management policy, and then compares and analyzes their performance and system bus traffic through their modeling and simulation. In addition, the most effective conflict management policy for the hardware transactional memory is presented through these comparison and analysis.

키워드

참고문헌

  1. K. Asanovic, R. Bodik, B. Catanzaro, J. Gebis, P. Husbands, K. Keutzer, D. Patterson, W. Plishker, J. Shalf, S. Williams, K. Uelick, The Landscape of Parallel Computing Research: A View form Berkeley, Technical Report No. UCB/EECS-2006-183, Electrical Engineering and Computer Sciences University of California at Berkeley, 2006.
  2. J. Larus, R. Rajwar, Transactional Memory, Morgan & Claypool, 2010.
  3. J. May, Parallel I/O for High Performance computing, Morgan Kaufmann, 2001.
  4. T. Anderson, "The Performance of Spin Lock Alternatives for Shared-Memory Multiprocessors", IEEE Transactions on Parallel and Distributed Systems, vol. 1, no. 1, pp. 6-16, Jan. 1990. https://doi.org/10.1109/71.80120
  5. M. Herligy, J. Moss, "Transactional Memory: Architectural Support for Lock-Free Data Structures", Proceedings of ISCA 1993, pp. 289-300, San Diego, CA, USA, May. 1993.
  6. J. Bobba, K. Moore, H. Volos, L. Yen, M. Hill, M. Swift, D. Wood, "Performance Pathologies in Hardware Transactional Memory", Proceedings of ISCA 2007, pp. 81-91, San Diego, CA, USA, Jun. 2007.
  7. J. Mellor-Crummey, M. Scott, "Algorithms for Scalable Synchronization on Shared-Memory Multiprocessors", ACM Transactions on Computer Systems, vol. 9, no. 1, pp. 21-65, Feb. 1991. https://doi.org/10.1145/103727.103729
  8. J. Larus, C. Kozyrakis, "Transactional Memory", Communications of the ACM, vol. 51, no. 7, pp. 80-88, July. 2008. https://doi.org/10.1145/1364782.1364800
  9. C. Cascaval, C. Blundell, M. Michael, H. Cain, P. Wu, S. Chiras, S. Chatterjee "Software Transactional Memory: Why is it Only a Research Toy?", Communications of the ACM, vol. 51, no. 11, pp. 40-46, Nov. 2008.
  10. P. Damron, A. Fedorova, Y. Lev, "Hybrid Transactional Memory", Proceedings of ASPLOS-XII, pp. 336-346, San Jose, CA, USA, Oct. 2006.
  11. K. Moore, J. Bobba, M. Moravan, M. Hill, D. Wood, "LogTM: Log-based Transactional Memory", Proceedings of HPCA-12, pp. 254-265, Austin, Texas, USA, Feb. 2006.
  12. S. Tomic, C. Perfumo, C. Kulkarni, A. Armejach, A. Cristal, O. Unsal, T. Harris, M. Valero, "EazyHTM: Eager-Lazy Hardware Transactional Memory," Proceedings of MICRO-42, pp. 145-155, New York, USA, Dec. 2009.
  13. D. Culler, J. Singh, A. Gupta, Parallel Computer Architecture: A Hardware/Software Approach, Morgan Kaufmann, 1999.
  14. R. Yoo, H. Lee, "Adaptive Transaction Scheduling for Transactional Memory System", Proceedings of the SPAA 2008, pp. 169-178, Munich, Germany, June. 2008.
  15. D. Burger, J. Goodman, A. Kagi, "Limited Bandwidth to Affect Processor Design", Micro, IEEE, vol. 17, no. 6, pp. 55-62, Nov/Dec. 1997. https://doi.org/10.1109/40.641597
  16. N. Hardavellas, M. Ferdman, B. Falsafi, A. Ailamaki, "Reactive NUCA: Near-Optimal Block Placement and Replication in Distributed Caches", Proceedings of ISCA 2009, Austin, pp. 184-195, TX, USA, June. 2009.
  17. Windows Desktop Development Center, http://msdn.microsoft.com/en-us/library/windows/desktop/ms644904(v=vs.85).aspx.