Massive Interconnection Research Articles

Distributed memory systems for simulating artificial neural networks

In executing tasks involving intelligent information processing, the human brain performs better than the digital computer. The human brain derives its power from a large number [ O(10 11)] of neurons which are interconnected by a dense interconnection network [ O(10 5) connections per neuron]. Artificial neural network (ANN) paradigms adopt the structure of the brain to try to emulate the intelligent information processing methods of the brain. ANN techniques are being employed to solve problems in areas such as pattern recognition, and robotic processing. Simulation of ANNs involves implementation of large number of neurons and a massive interconnection network. In this paper, we discuss various simulation models of ANNs and their implementation on distributed memory systems. Our investigations reveal that communication-efficient networks of distributed memory systems perform better than other topologies in implementing ANNs.

Guided-wave planar optical interconnects using highly multiplexed polymer waveguide holograms

An intraplanar interconnection scheme using substrate guided modes in conjunction with a highly multiplexed waveguide volume hologram is proposed. Acoustooptically addressed 1-to-50 passive and 1-to-2-to-100 reconfigurable interconnections with a fan-out diffraction efficiency of 55% at 514-nm wavelength are demonstrated. A coordinate transformation converts the 3-D diffraction problem into a 2-D one, which significantly simplifies the theoretical calculation. Intraplane massive fan-out optical interconnection using substrate guided mode provides both collinear and coplanar fan-out capability for data and clock signals. Colinearity of signal distribution allows the 2-D input signal array to be processes. The laminated waveguide device containing a highly multiplexed dichromated gelatin (DCG) hologram has been evaluated. >

Optoelectronic implementations of neural networks

The ability of optical systems to provide the massive interconnections between processors required in most neural network models, which constitutes their chief advantage for such applications, is discussed, focusing on holography. Because of the essential nonlinearity of the holographic connections, nonlinear processing elements are needed to perform complex computations. The use of GaAs hybrid optoelectronic processing elements is examined. GaAs is an excellent material for this purpose, since it can be used to fabricate both fast electronic circuits and optical sources and detectors. It is shown how a complete hybrid neural computer can be implemented using available technology developed for conventional computing. An experimentally demonstrated network in which optics plays an even larger role is described. >

Parallel N2 weighted reconfigurable networks for optical neuron and chip-to-chip interconnects

Various N 2 weighted reconfigurable optical networks for the optical free-space either neural network (ONN) or other chip-to-chip interconnections are proposed. With these new schemes, the large number of either electronic or optical delay lines, that were used in the previous ONN implementations, are minimized, so that the massive interconnections among monolithically intergrated either optical neurons or computing chips can be implemented. Both linearly- and circularly-distributed optical neuron (chip) interconnect models are presented. A comparison between the two interconnect models is also given.

Optical implementation of the Hopfield model

Optical implementation of content addressable associative memory based on the Hopfield model for neural networks and on the addition of nonlinear iterative feedback to a vector–matrix multiplier is described. Numerical and experimental results presented show that the approach is capable of introducing accuracy and robustness to optical processing while maintaining the traditional advantages of optics, namely, parallelism and massive interconnection capability. Moreover a potentially useful link between neural processing and optics that can be of interest in pattern recognition and machine vision is established.