Journal of the Korea Society for Simulation (한국시뮬레이션학회논문지)

The Korea Society for Simulation (한국시뮬레이션학회)
권호 표지

Placement and Performance Analysis of I/O Resources for Torus Multicomputer

토러스 다중컴퓨터를 위한 입출력 자원의 배치와 성능 분석

  • 안중석 (한컴기술연구소 개발실)
  • Published : 1997.12.01
Download PDF
⟨ Previous Next ⟩

Abstract

Performance bottleneck of parallel computer systems has mostly been I/O devices because of disparity between processor speed and I/O speed. Therefore I/O node placement strategy is required such that it can minimize the number of I/O nodes, I/O access time and I/O traffic in an interconnection network. In this paper, we propose an optimal distance-k embedding algorithm, and analyze its effect on system performance when this algorithm is applied to n x n torus architecture. We prove this algorithm is an efficient I/O node placement using software simulation. I/O node placement using the proposed algorithm shows the highest performance among other I/O node placements in all cases. It is because locations of I/O nodes are uniformly distributed in the whole network, resulting in reduced traffic in the intE'rconnection network.

Keywords

References

  1. Scalable Shared-Memory Multiprocessors The MIT alewife machine: a large-scale distributed-memory multiprocessor A. Agarwal(et al.)
  2. Algebraic coding theory E. R. Berlekamp
  3. IEEE Transactions on Computers v.44 no.8 Lee distance and topological properties of k-ary n-cubes B. Bose(et al.)
  4. PROTEUS User Documentation, 545 Technology Square(0.5 edition) E. A. Brewer;C. N. Dellarocas
  5. Proceedings. of Supercomputing '89, St. Petersberg Concurrent file system - Making highly parallel mass storage transparent S. Cannon
  6. IEEE Transations on Computers v.C-36 no.5 Deadlock-free message routing in multiprocessor interconnection networks W. J. Dally;C. L. Seitz
  7. Proceedings of the Fifth International Parallel Processing Symposium Parallel I/O subsystems for hypercube multicomputers J. Ghosh;B. Agarwal
  8. Journal of Parallel and Distributed Computing v.17 Performance evaluation of a paralle I/O subsystem for hypercube multicomputers J. Ghosh(et al.)
  9. Advanced Computer Architecture K. Hwang
  10. Ph.D Dissertation, University of Illinois Disk I/O in high-performance computing systems D. Jensen
  11. IEEE Transactions on Parallel and Distributed Systems v.6 no.8 Optimal hot spot allocation on meshes for large-scale data-parallel algorithms S. Y . Lee;C. M. Chen
  12. '89 International Conference on Parallel Processing Resource placement in k-ary n-cubes P. Ramanathan;S. Chalasani
  13. Proc. 1988 Int. Conference on Parallel Processing I/O embeddings in hypercubes A. L. N. Reddy;P. Banerjee;S. G. Abraham
  14. IEEE Transactions on Parallel and Distributed Systems v.1 no.2 Design, analysis, and simulation of I/O architectures for hypercube multiprocessors A. L. N. Reddy;P. Banerjee
  15. Multicomputer Networks D. Reed.;R. Fujimoto