An Optically Interconnected Reconfigurable Mesh

In this paper, we present an electro-optical architecture called the Optical Reconfigurable Mesh (ORM). The ORM has two layers: the deflection layer and the processing layer. The processing layer is an N×N reconfigurable mesh. The deflection layer, situated directly above the processing layer, provides unit- time free space optical interconnections for the processors. The architecture supports three types of unit-time communication mechanisms. The first is for arbitrary planar connections among sets of locally connected processors using the electrical reconfigurable mesh. The second is for arbitrary connections among N of the processors using the electrical buses on the processing layer and N2 fixed passive optical deflecting units on the deflection layer. The third is for arbitrary free space optical connections among any of the N2 processors using the N2 mechanically reconfigurable deflectors in the deflection layer. The reconfiguration time of the first two communication mechanisms is on the order of nanoseconds, while for the third mechanism it is on the order of milliseconds. Therefore, algorithms are designed to utilize all three types of mechanisms but minimize or avoide the use of the third type during the execution of algorithms. We present a set of basic data movement algorithms for ORM and, as an applicate example, outline an O(logN) time algorithm for finding the convex hulls of all figures in an N×N digitized image.

[1]  Lili Hai,et al.  Efficient parallel processing with optical interconnections , 1997 .

[2]  Ted H. Szymanski,et al.  "Hypermeshes": Optical Intercomnnection Network for Parallel Computing , 1995, J. Parallel Distributed Comput..

[3]  Dionysios I. Reisis,et al.  Parallel Computations on Reconfigurable Meshes , 1993, IEEE Trans. Computers.

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

[5]  M. M. Eshaghian Parallel algorithms for image processing on OMC , 1991 .

[6]  Gen-Huey Chen,et al.  Constant Time Algorithms for the Transitive Closure and Some Related Graph Problems on Processor Arrays with Reconfigurable Bus Systems , 1990, IEEE Trans. Parallel Distributed Syst..

[7]  Yao Li,et al.  Free-space optical interconnects using connectivity-enhanced, mesh-based networks , 1994 .

[8]  Ahmed Louri,et al.  3D optical interconnects for high-speed interchip and interboard communications , 1994, Computer.

[9]  Tut San Guan,et al.  Reconfigurable multiple-behavioural architecture using free-space optical communication , 1994, First International Workshop on Massively Parallel Processing Using Optical Interconnections.

[10]  V. K. Prasanna-Kumar,et al.  Parallel computing with optical interconnects , 1988 .

[11]  Patrick W. Dowd Wavelength Division Multiple Access Channel Hypercube Processor Interconnection , 1992, IEEE Trans. Computers.

[12]  S. Kawai,et al.  Free-space multistage optical interconnection networks using micro lens arrays , 1991 .

[13]  John A. Neff,et al.  Analysis of a three-dimensional computer optical scheme based on bidirectional free-space optical interconnects , 1995 .

[14]  Mary Mehrnoosh Eshaghian-Wilner,et al.  The systolic reconfigurable mesh , 2004, Parallel Process. Lett..

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

[16]  M. Murdocca,et al.  A model for a reconfigurable fine-grained optoelectronic processor , 1994, First International Workshop on Massively Parallel Processing Using Optical Interconnections.