Scalable Optical Interconnection Networks for Large-Scale Parallel Computers

Abstract

A new scalable interconnection topology called the Spanning Multichannel Linked Hypercube SMLH is proposed along with an optical implementation methodology that combines both the advantages of free space optics with those of wavelength division multiplexing techniques. The SMLH uses the hypercube topology as a basic building block and connects such building blocks using two-dimensional multichannel links similar to spanning buses. In doing so, the SMLH combines positive features of both the hypercube small diameter, high connectivity, symmetry, simple routing, and fault tolerance and the spanning bus hypercube SBH constant node degree, scalability, and ease of physical implementation, while at the same time circumventing their disadvantages. The SMLH topology supports many communication patterns found in di erent classes of computation, such as bus-based, mesh-based, and treebased problems as we l l a s h ypercube-based problems. A v ery attractive feature of the SMLH network is its ability to support a large number of processors with the possibility of maintaining a constant degree and a constant diameter. Other positive features include symmetry, incremental scalability, and faulttolerance. It is shown that the SMLH network provides better average message distance, average tra c density, and queueing delay than many similar networks, including the binary hypercube, the SBH, etc. Additionally, the SMLH has comparable performance to other high-performance hypercubic networks, including the Generalized Hypercube and the Hypermesh.

Very Large Scale Optical Interconnect Systems For Different Types of Optical Interconnection Networks

“…It is evident that optical crossbar connected cluster network is higher node degree, link complexity, bisection bandwidth, number of links, number of processors, transmission bit rate capacity for audio and video signals, and higher additional network costs than optical hypercube connected cluster network. We have summarized the complete comparison of large scale of our optical interconnection systems with their medium optical interconnection systems [15]. It is clear from the comparison the best performance of our large scale optical interconnection systems in large number of processors and links and transmission bit rate capacity of audio and video signals in both types of optical interconnection networks over medium scale optical interconnection systems as mentioned in the previous studies.…”

“…viii) Figs. (12)(13)(14)(15)(16)(17)(18)(19)(20)(21) have assured that as number of processors per cluster, number of tunable VCSELs, number of detectors, number of waveguides, number of demultiplexers, and number of clusters increase this result in increasing in network cost in the case of both optical crossbar connected cluster and optical hypercube connected cluster networks. We have indicated that optical crossbar connected cluster network is higher network cost than optical hypercube connected cluster network.…”

Torus Embedded Hypercube Interconnection Network: A Comparative Study

“…Scalability of a network [4], [5] is defined as the property by which the size of the system can be expanded with nominal changes in the existing configuration provided that system expansion results with improvement in performance. The torus embedded hypercube network is highly scalable network.…”

“…The advantages of torus network can be imposed on to the mesh embedded hypercube network [4], [5] to give rise to an embedded architecture [6]- [8] called torus embedded hypercube scalable interconnection network and hence an architectural enhancement of mesh embedded hypercube network.…”

Prediction and preemptive control of network congestion in distributed real-time environment

“…In this research, hypercube network topology is chosen because of its superior topological characteristics which includes small diameter ( communication delays are less when network's diameter is small), high connectivity, simple routing, and fault tolerance [62]. …”