Euro-Par 1996: Euro-Par'96 Parallel Processing pp 287-298 | Cite as

The effect of flow control and routing adaptivity on priority-driven traffic in multiprocessor networks

  • Shobana Balakrishnan
  • Füsun Özgüner
Workshop 02 Routing and Communication in Interconnection Networks
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1123)

Abstract

We study the impact of two flow control schemes and routing adaptivity on the performance of priority-driven traffic produced by real-time applications. The first is wormhole routing (WR) with real-time extensions [1] and the second is preemptive pipelined circuit switching (PPCS-RT) [2]. Our simulations show that for a fixed number of virtual channels (VCs)/link, parallel lanes are more effective than adaptive routing alone, in reducing the number of messages that miss their deadlines. PPCS-RT performs better than WR due to the VC preemption protocol supported by it.

Keywords

Virtual Channel Link Utilization Priority Inversion Path Establishment Priority Message 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Li, J.-P., Mutka, M.W.: Real-time virtual channel flow control. J. Par. Dist. Comp. 32(1996) 49–65CrossRefGoogle Scholar
  2. 2.
    Balakrishnan, S., Özgüner, F.: A priority-based flow control mechanism to support real-traffic in pipelined direct networks. To appear Proc. Int. Conf. Par. Proc. (1996)Google Scholar
  3. 3.
    Ni, L.M., McKinley, P.K.: A survey of wormhole routing techniques in direct networks. IEEE Computer (1993) 62–76Google Scholar
  4. 4.
    Duato, J.: A new theory of deadlock-free adaptive routing in wormhole networks. IEEE Trans. Par. Dist. Sys. 4(12)(1993) 1320–1331CrossRefGoogle Scholar
  5. 5.
    Dally, W.J.: Virtual-channel flow control. IEEE Trans. Par. Dist. Sys. 3(2)(1992) 194–205CrossRefGoogle Scholar
  6. 6.
    Leung, J.Y.-T., Whitehead, J.: On the complexity of fixed-priority scheduling of periodic real-time tasks. Performance Evaluation 2(4)(1982) 237–250CrossRefGoogle Scholar
  7. 7.
    Gaughan, P.T., Yalamanchili, S.: A family of fault-tolerant routing protocols for direct multiprocessor networks. IEEE Trans. Par. Dist. Sys. 6(5)(1995) 482–497CrossRefGoogle Scholar
  8. 8.
    Dolter, J. et. al.: SPIDER: Flexible and efficient communication support for point-to-point distributed systems. Proc. Int. Conf. Dist. Comp. Sys.(1995) 574–580Google Scholar
  9. 9.
    Shukla, S.B., Agrawal, D.P.: Scheduling pipelined communication in distributed memory multiprocessors for real-time applications. Proc. Int. Symp. Comp. Arch. (1991) 222–231Google Scholar
  10. 10.
    Balakrishnan, S., Özgüner, F.: Providing message delivery guarantees in pipelined flit-buffered multiprocessor networks. To appear Proc. IEEE Real-time Technology and Applications Symp. (1996)Google Scholar
  11. 11.
    Schwetman, H.: CSIM: A C-based process-oriented simulation language. Proc.Winter Simulation Conf. (1986) 387–396Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1996

Authors and Affiliations

  • Shobana Balakrishnan
    • 1
  • Füsun Özgüner
    • 1
  1. 1.Dept. of Electrical EngineeringThe Ohio State UniversityColumbusUSA

Personalised recommendations