The Journal of Supercomputing

, Volume 53, Issue 2, pp 329–351 | Cite as

Metacube—a versatile family of interconnection networks for extremely large-scale supercomputers



In the next decade, the high-performance supercomputers will consist of several millions of CPUs. The interconnection networks in such supercomputers play an important role for achieving high performance. In this paper, we introduce the Metacube (MC), a versatile family of interconnection network that can connect an extremely large number of nodes with a small number of links per node and keep the diameter rather low. An MC network has a 2-level hypercube structure. An MC(k,m) network can connect \(2^{2^{k}m+k}\) nodes with m+k links per node, where k is the dimension of the high-level hypercubes (classes) and m is the dimension of the low-level hypercubes (clusters). An MC is a symmetric network with short diameter, easy and efficient routing and broadcasting similar to that of the hypercube. However, the MC network can connect millions of nodes with up to 6 links per node. An MC(2,3) with 5 links per node has 16,384 nodes and an MC(3,3) with 6 links per node has 134,217,728 nodes. We describe the MC network’s structure, topological properties, routing and broadcasting algorithms, and the Hamiltonian cycle embedding in the Metacube networks.


Interconnection network Hypercube Routing and broadcasting Hamiltonian cycle Algorithm 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Adiga NR, Blumrich MA, Chen D, Coteus P, Gara A, Giampapa ME, Heidelberger P, Singh S, Steinmacher-Burow BD, Takken T, Tsao M, Vranas P (2005) Blue Gene/L torus interconnection network. IBM J Res Dev 49(2/3):265–276. CrossRefGoogle Scholar
  2. 2.
    Amawy AE, Latifi S (1991) Properties and performance of folded hypercubes. IEEE Trans Parallel Distrib Syst 2(1):31–42 CrossRefGoogle Scholar
  3. 3.
    Arabnia HR (1993) A transputer-based reconfigurable parallel system. Transput Res Appl (NATUG) 6:153–169 Google Scholar
  4. 4.
    Bhandarkar SM, Arabnia HR (1995) The Hough transform on a reconfigurable multi-ring network. J Parallel Distrib Comput 24(1):107–114 CrossRefGoogle Scholar
  5. 5.
    Bhandarkar SM, Arabnia HR (1995) The refined multiprocessor theoretical properties and algorithms. Parallel Comput 21(11):1783–1806 CrossRefGoogle Scholar
  6. 6.
    Chang C, Lin C, Tan J, Huang H, Hsu L (2009) The super spanning connectivity and super spanning laceability of the enhanced hypercubes. J Supercomput 48(1):66–87 CrossRefGoogle Scholar
  7. 7.
    Efe K (1992) The crossed cube architecture for parallel computation. IEEE Trans Parallel Distrib Syst 3(5):513–524 CrossRefGoogle Scholar
  8. 8.
    Fu JS (2004) Hamiltonian-connectedness of the WK-recursive network. In: Proceedings of the 7th international symposium on parallel architectures, algorithms and networks (ISPAN’04), 2004, pp 569–574 Google Scholar
  9. 9.
    Ghose K, Desai KR (1995) Hierarchical cubic networks. IEEE Trans Parallel Distrib Syst 6(4):427–435 CrossRefGoogle Scholar
  10. 10.
    Hayes JP, Mudge TN (1989) Hypercube supercomputers. Proc IEEE 17(12):1829–1841 CrossRefGoogle Scholar
  11. 11.
    Leighton FT (1992) Introduction to parallel algorithms and architectures: arrays, trees and hypercubes. Morgan Kaufmann, San Mateo MATHGoogle Scholar
  12. 12.
    Li Y, Peng S, Chu W (2002) Hamiltonian cycle embedding for fault tolerance in dual-cube. In: Proceedings of the IASTED international conference on networks, parallel and distributed processing, and applications (NPDPA 2002), Tsukuba, Japan, October 2002, pp 1–6 Google Scholar
  13. 13.
    Li Y, Peng S, Chu W (2004) Efficient collective communications in dual-cube. J Supercomput 28(1):71–90 MATHCrossRefGoogle Scholar
  14. 14.
    Li Y, Peng S, Chu W (2008) Prefix computation and sorting in dual-cube. In: Proceedings of the 37th international conference on parallel processing (ICPP’08), Portland, Oregon, USA, Sept. 8–12, 2008, pp 389–396 Google Scholar
  15. 15.
    Li Y, Peng S, Chu W (2008) An effective structure for algorithmic design and a parallel prefix algorithm on Metacube. In: Proceedings of the international conference on parallel and distributed computing, applications and technologies (PDCAT’08), Dunedin, New Zealand, Dec. 1–4, 2008, pp 54–61 Google Scholar
  16. 16.
    Preparata FP, Vuillemin J (1981) The cube-connected cycles: a versatile network for parallel computation. Commun ACM 24:300–309 CrossRefMathSciNetGoogle Scholar
  17. 17.
    SGI (1997) Origin2000 Rackmount owner’s guide, 007-3456-003. 1997
  18. 18.
    TOP500 (2008) Supercomputer sites., Jun. 2008
  19. 19.
    Tucker LW, Robertson GG (1988) Architecture and applications of the connection machine. IEEE Comput 21:26–38 Google Scholar
  20. 20.
    Vanvoorst B, Seidel S, Barscz E (1994) Workload of an ipsc/860. In: Proceedings of the scalable high-performance computing conference, 1994, pp 221–228 Google Scholar
  21. 21.
    Wang D (2008) On embedding Hamiltonian cycles in crossed cubes. IEEE Trans Parallel Distrib Syst 19(3):334–346 CrossRefGoogle Scholar
  22. 22.
    Ziavras SG (1994) RH: a versatile family of reduced hypercube interconnection networks. IEEE Trans Parallel Distrib Syst 5(11):1210–1220 CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2009

Authors and Affiliations

  1. 1.Department of Computer ScienceHosei UniversityTokyoJapan
  2. 2.Department of Computer HardwareUniversity of AizuAizu-WakamatsuJapan

Personalised recommendations