Journal of the Institute of Electronics and Information Engineers (전자공학회논문지)

The Institute of Electronics and Information Engineers (대한전자공학회)
권호 표지
DOI QR코드

DOI QR Code

A Latency Optimization Mapping Algorithm for Hybrid Optical Network-on-Chip

하이브리드 광학 네트워크-온-칩에서 지연 시간 최적화를 위한 매핑 알고리즘

  • Lee, Jae Hun (College of Information & Communication Engineering, Sungkyunkwan University) ;
  • Li, Chang Lin (College of Information & Communication Engineering, Sungkyunkwan University) ;
  • Han, Tae Hee (College of Information & Communication Engineering, Sungkyunkwan University)
  • 이재훈 (성균관대학교 정보통신대학) ;
  • 이창림 (성균관대학교 정보통신대학) ;
  • 한태희 (성균관대학교 정보통신대학)
  • Received : 2013.01.02
  • Published : 2013.07.25
Download PDF
⟨ Previous Next ⟩

Abstract

To overcome the limitations in performance and power consumption of traditional electrical interconnection based network-on-chips (NoCs), a hybrid optical network-on-chip (HONoC) architecture using optical interconnects is emerging. However, the HONoC architecture should use circuit-switching scheme owing to the overhead by optical devices, which worsens the latency unfairness problem caused by frequent path collisions. This resultingly exert a bad influence in overall performance of the system. In this paper, we propose a new task mapping algorithm for optimizing latency by reducing path collisions. The proposed algorithm allocates a task to a certain processing element (PE) for the purpose of minimizing path collisions and worst case latencies. Compared to the random mapping technique and the bandwidth-constrained mapping technique, simulation results show the reduction in latency by 43% and 61% in average for each $4{\times}4$ and $8{\times}8$ mesh topology, respectively.

기존 전기적 상호 연결을 사용한 네트워크-온-칩(Network-on-Chip, NoC)의 전력 및 성능 한계를 보완하고자 광학적 상호연결을 이용하는 하이브리드 광학 네트워크-온-칩(HONoC)이 등장하였다. 하지만 HONoC에서는 광학적 소자 특성으로 인해 서킷 스위칭을 사용함으로써 경로 충돌이 빈번하게 발생하며 이로 인해 지연 시간 불균형의 문제가 심화되어 전체적인 시스템 성능에 악영향을 미치게 된다. 본 논문에서는 경로 충돌을 최소화 시켜 지연 시간을 최적화 할 수 있는 새로운 태스크 매핑 알고리즘을 제안하였다. HONoC 환경에서 태스크를 각 Processing Element (PE)에 할당하고 경로 충돌을 최소화하며, 부득이한 경로 충돌의 경우 워스트 케이스 (worst case) 지연 시간을 최소화 할 수 있도록 하였다. 모의실험 결과를 통해 무작위 매핑 방식, 대역폭 제한 매핑 방식과 비교하여, 제안된 알고리즘이 $4{\times}4$ 메시 토폴로지에서는 평균 43%, $8{\times}8$ 메시 토폴로지에서는 평균 61%의 지연 시간 단축 효과가 있음을 확인할 수 있었다.

Keywords

References

  1. A. Shacham and K. Bergman and L. P. Carloni, "On the design of a photonic Network-on-Chip," in Proc. of the First Int. Symp. on Networks-on-Chip, pp. 53-64, Princeton, New Jersey, USA, May 2007.
  2. S. Kwon and J. D. Cho and T. H. Han, "Switch Architecture and Routing Optimization Strategy Using Optical Interconnects for Network-on-Chip," Journal of IEEK-SD, Vol. 46, no. 9, pp. 25-32, September 2009.
  3. Y. Ye and L. Duan and J. Xu and J. Ouyang and M. K. Hung and Y. Xie, "3D optical Networks-on-Chip (NoC) for Multiprocessor Systems-on-chip (MPSoC)," in Proc. of IEEE Int. Conf. on 3D System Integration, pp. 1-6, San Francisco, California, USA, September 2009.
  4. K. H. Mo and Y. Ye and X. Wu and W. Zhang and W. Liu and J. Xu, "A hierarchical hybrid optical-electronic Network-on-Chip," in Proc. of IEEE Computer Society Annual Symposium on VLSI, pp. 327-332, Lixouri Kefalonia, Greece, July 2010.
  5. J. T. Seo and Y. J. Hwang and T. H. Han, "A Study on the Parallel Routing in Hybrid Optical Networks-on-Chip," Journal of IEEK-SD, Vol. 48, no. 8, pp. 25-32, August 2011.
  6. S. Murali and G. D. Micheli, "Bandwidth-constrained mapping of cores onto NoC architectures," in Proc. of Conf. on Design, automation and test in Europe, Vol. 2, pp. 896-901, Paris, France, February 2004.
  7. J. Hu and R. Marculescu, "Energy- and performance-aware mapping for regular NoC architectures," IEEE Trans. Computer-Aided Design of Integrated Circuits and Systems, Vol. 24, no. 4, pp. 551-562, April 2005. https://doi.org/10.1109/TCAD.2005.844106
  8. J. Hu and R. Marculescu, "Energy-aware communication and task scheduling for network-on-chip architectures under real-time constraints," in Proc. of Conf. Design, automation and test in Europe, Vol. 1, pp. 234-239, Paris, France, February 2004.
  9. J. Hu and C. Buckl and A. Raabe and A. Knoll, "Energy-aware task allocation for network-on-chip based heterogeneous multiprocessor systems," in Proc. of 19th Euromicro International Conf. Parallel, Distributed and Network-Based Processing, pp. 447-454, Ayia Napa, Cyprus, February 2011.
  10. H. Gu and K. H. Mo and J. Xu and W. Zhang, "A low-power low-cost optical router for optical networks-on-chip in multiprocessor systems-on-chip," in Proc. of IEEE Computer Society Annual Symp. VLSI, pp. 19-24, Tampa, Florida, USA, May 2009.
  11. J. A. Lott and N. N. Ledentsov and V. A. Shchukin and A. Mutig and S. A. Blokhin and A. M. Nadtochiy and G. Fiol and D. Bimberg, "850nm VCSELs for up to 40 Gbit/s Short Reach Data Links," in Proc. of 2010 Conference on Lasers and Electro-Optics (CLEO) and Quantum Electronics and Laser Science Conference (QELS), pp. 1-2, San Jose, California, USA, May 2010.
  12. R. B. Dick and D. L. Rhodes and W. Wolf, "TGFF: task graphs for free," in Proc. of the 6th international workshop on Hardware/software codesign , pp. 97-101, Seattle, WA, USA, March 1998.

Cited by

  1. Topology Design for Energy/Latency Optimized Application-specific Hybrid Optical Network-on-Chip (HONoC) vol.51, pp.11, 2014, https://doi.org/10.5573/ieie.2014.51.11.083