Scalable optical hypercube-based interconnection network for massively parallel computing.

Two important parameters of a network for massively parallel computers are scalability and modularity. Scalability has two aspects: size and time (or generation). Size scalability refers to the property that the size of the network can be increased with nominal effect on the existing configuration. Also, the increase in size is expected to result in a linear increase in performance. Time scalability implies that the communication capabilities of a network should be large enough to support the evolution of processing elements through generations. A modular network enables the construction of a large network out of many smaller ones. The lack of these two important parameters has limited the use of certain types of interconnection networks in the area of massively parallel computers. We present a new modular optical interconnection network, called an optical multimesh hypercube (OMMH), which is both size and time scalable. The OMMH combines positive features of both the hypercube (small diameter, high connectivity, symmetry, simple routing, and fault tolerance) and the torus (constant node degree and size scalability) networks. Also presented is a three-dimensional optical implementation of the OMMH network. A basic building block of the OMMH network is a hypercube module that is constructed with free-space optics to provide compact and high-density localized hypercube connections. The OMMH network is then constructed by the connection of such basic building blocks with multiwavelength optical fibers to realize torus connections. The proposed implementation methodology is intended to exploit the advantages of both space-invariant free-space and multiwavelength fiber-based optical interconnect technologies. The analysis of the proposed implementation shows that such a network is optically feasible in terms of the physical size and the optical power budget.

[1]  John Cocke,et al.  Computer architecture in the 1990s , 1991, Computer.

[2]  Y Sheng,et al.  Space invariant multiple imaging for hypercube interconnections. , 1990, Applied optics.

[3]  B K Jenkins,et al.  Digital optical cellular image processor (DOCIP): experimental implementation. , 1993, Applied optics.

[4]  Frederick B. McCormick Generation Of Large Spot Arrays From A Single Laser Beam By Multiple Imaging With Binary Phase Gratings , 1989 .

[5]  John Perry Barlow Will Japan jack in? , 1992, CACM.

[6]  G. Bell A Teraflop Before its Time , 1992 .

[7]  A Louri,et al.  Design methology for three-dimensional space-invariant hypercube networks with graph bipartitioning. , 1993, Optics letters.

[8]  Karl-Heinz Brenner,et al.  Space-variance in optical computing systems , 1992 .

[9]  B D Metcalf,et al.  High-capacity wavelength demultiplexer with a large-diameter GRIN rod lens. , 1982, Applied optics.

[10]  Charles L. Seitz,et al.  The cosmic cube , 1985, CACM.

[11]  Gordon Bell,et al.  Ultracomputers: a teraflop before its time , 1992, CACM.

[12]  A Louri,et al.  Efficient implementation methodology for three-dimensional space-invariant hypercube-based optical interconnection networks. , 1993, Applied optics.

[13]  A. Hall Applied Optics. , 2022, Science.

[14]  Dharma P. Agrawal,et al.  Generalized Hypercube and Hyperbus Structures for a Computer Network , 1984, IEEE Transactions on Computers.

[15]  Anoop Gupta,et al.  The Stanford Dash multiprocessor , 1992, Computer.

[16]  H. Temkin,et al.  High-speed vertical-cavity surface emitting laser , 1993, IEEE Photonics Technology Letters.

[17]  S H Lee,et al.  Computer aided design of computer generated holograms for electron beam fabrication. , 1989, Applied optics.

[18]  L. Boivin,et al.  Multiple imaging using various types of simple phase gratings. , 1972, Applied optics.

[19]  M. J. Karol,et al.  ShuffleNet: an application of generalized perfect shuffles to multihop lightwave networks , 1988, IEEE INFOCOM '88,Seventh Annual Joint Conference of the IEEE Computer and Communcations Societies. Networks: Evolution or Revolution?.

[20]  N. Streibl,et al.  Array generation with lenslet arrays. , 1991, Applied optics.

[21]  M W Haney Self-similar grid patterns in free-space shuffle - exchange networks. , 1993, Optics letters.

[22]  Nian-Feng Tzeng,et al.  Enhanced Hypercubes , 1991, IEEE Trans. Computers.

[23]  Kai Hwang,et al.  Advanced computer architecture - parallelism, scalability, programmability , 1992 .

[24]  Ian N. Robinson,et al.  Hardware to support runtime intelligence , 1992, Computer.

[25]  R M Vasu,et al.  Multiple imaging and multichannel optical processing with split lenses. , 1987, Applied optics.

[26]  D. G. Meyer,et al.  Multiple channel architecture: a new optical interconnection strategy for massively parallel computers , 1991 .

[27]  Claus M. Habiger,et al.  Hybrid-WSI: a massively parallel computing technology? , 1993, Computer.

[28]  Hui Kang,et al.  Optical implementation of the perfect shuffle interconnection networks using CGH beamsplitter , 1993 .

[29]  Sartaj Sahni,et al.  An optimal routing algorithm for mesh-connected Parallel computers , 1980, JACM.

[30]  G. R. Hill,et al.  Wavelength domain optical network techniques , 1990, Proc. IEEE.

[31]  Thomas J. Cloonan,et al.  Optical Implementation And Performance Of One-Dimensional And Two-Dimensional Trimmed Inverse Augmented Data Manipulator Networks For Multiprocessor Computer Systems , 1989 .

[32]  S. D. Personick,et al.  Receiver design for optical fiber communication systems , 1980 .

[33]  A W Lohmann,et al.  Free-space optical mesh-connected bus networks using wavelength-division multiple access. , 1993, Applied optics.

[34]  Carlo H. Séquin,et al.  Hypertree: A Multiprocessor Interconnection Topology , 1981, IEEE Transactions on Computers.

[35]  M. Karol,et al.  Shuffle Net: an application of generalized perfect shuffles to multihop lightwave networks , 1991 .

[36]  Raymond K. Kostuk,et al.  Multiplexed substrate-mode holograms , 1990 .

[37]  David Hung-Chang Du,et al.  Distributed computing with high-speed optical networks , 1993, Computer.

[38]  P. Green Fiber Optic Networks , 1992 .

[39]  Ahmed Louri,et al.  A Design Methodology for Three-dimensional Space-invariant Hypercube Networks Using Graph Bipartitioning , 1993 .