Performance analysis of a class of hierarchical hypercube multicomputer networks

doi:10.1016/0166-5316(91)90053-6

Performance Evaluation

Volume 13, Issue 3, 1 November 1991, Pages 159-179

https://doi.org/10.1016/0166-5316(91)90053-6 Get rights and content

Abstract

Hierarchical interconnection networks (HINs) have been proposed to interconnect large numbers of processors in a multicomputer system. It has been shown that HINs provide better cost-benefit ratios than the corresponding nonhierarchical interconnection networks. This article discusses performance of two schemes that improve fault-tolerance of binary hypercube-based HINs. Both these schemes use hardware redundancy. In one, a standby-spare node is provided to reduce the impact of key node failures on the network reliability; in the other, a part of the network is duplicated. Both these schemes improve the network reliability substantially. The analysis presented here shows that, from a performance point of view. neither of the two schemes dominates the other for all parameter values and system characteristics. If the system supports applications that have varying degrees of communication locality and/or different computation-communication ratios, the duplication scheme is to be recommended. On the other hand, when applications exhibit high degrees of communication locality and high computation-communication ratios, the standby-spare node scheme provides better performance.

The impact of three routing algorithms is also considered. We derive bounds on message delay and saturation message generation rate and compare the performance of these routing algorithms in achieving these bounds. It is shown by means of analytical and simulation models that performance of the replication scheme is less sensitive to the routing algorithm used.

References (37)

D.P. Agrawal et al.
Evaluating the performance of multicomputer configurations
IEEE Comput.
(1986)
B.W. Arden et al.
Analysis of chordal ring network
IEEE Trans. Comput.
(1981)
L.N. Bhuyan et al.
Design and performance of generalized interconnection networks
IEEE Trans. Comput.
(1983)
L.N. Bhuyan et al.
Generalized hypercube and hyperbus structures for a computer network
IEEE Trans. Comput.
(1984)
L.N. Bhuyan et al.
Dependability evaluation of multicomputer networks
D. Carlson
The mesh with a global mesh: a flexible, high-speed organization for parallel computation
S.P. Dandamudi
A performance comparison of routing algorithms of hierarchical hypercube multicomputer networks
S.P. Dandamudi
A comparison of link-oriented multicomputer interconnection networks
(1990)
S.P. Dandamudi
Hierarchical Hypercube Multicomputer Interconnection Networks
(1991)
S.P. Dandamudi et al.
Hierarchical interconnection networks for multicomputer systems
IEEE Trans. Comput.
(1990)

S.P. Dandamudi et al.

On hypercube-based hierarchical interconnection network design

J. Parallel and Distributed Comput.

(1991)

C.R. Das et al.

Dependability modeling for multiprocessors

IEEE Comput.

(1990)

N. Deo

Graph Theory with Applications to Engineering and Computer Science

(1974)

T.-Y. Feng

A survey of interconnection networks

IEEE Comput.

(1981)

K. Ghose et al.

The HCN: a versatile interconnection network based on cubes

K. Ghose et al.

The design and evaluation of the hierarchical cubic network

J.R. Goodman et al.

Hypertree: a multiprocessor interconnection topology

IEEE Trans. Comput.

(1981)

G. Gopal et al.

Delay analysis of broadcast routing in packet-switching networks

IEEE Trans. Comput.

(1981)

Cited by (3)

Fast and simple decomposition techniques for the reliability analysis of interconnection networks
1999, Journal of Systems and Software
To study and analyze the reliability of complex systems such as multistage interconnection networks (MINs) and hierarchical interconnection networks (HINs), traditional techniques such as simulation, the Markov chain modeling, and probabilistic techniques have been widely used. Unfortunately, these traditional approaches become intractable when dealing with real-life applications. For the purpose of reliability analysis, the Petri net (PN) theory can also be used, but the major drawback of PN models is the state space explosion with increasing model complexity. Therefore, in this paper, we propose two analytical decomposition techniques for computing (approximately) the transient state space solution of large stochastic PN (SPN) models of MINs and HINs. In the first technique, a large scale SPN model is partitioned into smaller submodels. These submodels are compressed and combined to calculate the entire net. Throughout our reliability analysis of a simple example, we will understand how to decompose an entire net of a modeled system into subnets and what kinds of quantities are required to pass from one subnet to another. The modeling power of this technique is illustrated through the reliability analysis of an 8 × 8 Omega network as a practical example of MINs. In the second approach, we use the topological properties of HINs to build an SPN generic modeling methodology to aid the reliability evaluation of different HIN configurations. Finally, we illustrate and discuss the numerical reliability results obtained from the proposed techniques. It has been shown that the suggested techniques give results quite close to those obtained by the exact method with an enormous saving in computation time and memory usage.
DBG MANs and their routing performance
2000, IEE Proceedings: Communications
Simulation Study of the Performance of a Cluster-Based Hypercube Multicomputer
1997, Simulation

View full text

Performance Evaluation

Performance analysis of a class of hierarchical hypercube multicomputer networks

Abstract

Evaluating the performance of multicomputer configurations

IEEE Comput.

Analysis of chordal ring network

IEEE Trans. Comput.

Design and performance of generalized interconnection networks

IEEE Trans. Comput.

Generalized hypercube and hyperbus structures for a computer network

IEEE Trans. Comput.

Dependability evaluation of multicomputer networks

The mesh with a global mesh: a flexible, high-speed organization for parallel computation

A performance comparison of routing algorithms of hierarchical hypercube multicomputer networks

A comparison of link-oriented multicomputer interconnection networks

Hierarchical Hypercube Multicomputer Interconnection Networks

Hierarchical interconnection networks for multicomputer systems

IEEE Trans. Comput.

On hypercube-based hierarchical interconnection network design

J. Parallel and Distributed Comput.

Dependability modeling for multiprocessors

IEEE Comput.

Graph Theory with Applications to Engineering and Computer Science

A survey of interconnection networks

IEEE Comput.

The HCN: a versatile interconnection network based on cubes

The design and evaluation of the hierarchical cubic network

Hypertree: a multiprocessor interconnection topology

IEEE Trans. Comput.

Delay analysis of broadcast routing in packet-switching networks

IEEE Trans. Comput.

Fast and simple decomposition techniques for the reliability analysis of interconnection networks

DBG MANs and their routing performance

Simulation Study of the Performance of a Cluster-Based Hypercube Multicomputer