A prototype 3D optically interconnected neural network

We report the implementation of a prototype three-dimensional (3D) optoelectronic neural network that combines free-space optical interconnects with silicon-VLSI-based optoelectronic circuits. The prototype system consists of a 16-node input, 4-neuron hidden, and a single-neuron output layer, where the denser input-to-hidden-layer connections are optical. The input layer uses PLZT light modulators to generate optical outputs which are distributed over an optoelectronic neural network chip through space-invariant holographic optical interconnects. Optical interconnections provide negligible fan-out delay and allow compact, purely on-chip electronic H-tree type fan-in structure. The small prototype system achieves a measured 8-bit electronic fan-in precision and a calculated maximum speed of 640 million interconnections per second. The system was tested using synaptic weights learned off system and was shown to distinguish any vertical line from any horizontal one in an image of 4/spl times/4 pixels. New, more efficient light detector and small-area analog synapse circuits and denser optoelectronic neuron layouts are proposed to scale up the system. A high-speed, feed-forward optoelectronic synapse implementation density of up to 10/sup 4//cm/sup 2/ seems feasible using new synapse design. A scaling analysis of the system shows that the optically interconnected neural network implementation can provide higher fan-in speed and lower power consumption characteristics than a purely electronic, crossbar-based neural network implementation. >

[1]  Hendrikus J. M. Veendrick,et al.  Short-circuit dissipation of static CMOS circuitry and its impact on the design of buffer circuits , 1984 .

[2]  Ugur Cilingiroglu,et al.  A purely capacitive synaptic matrix for fixed-weight neural networks , 1991 .

[3]  J W Goodman,et al.  Fully parallel, high-speed incoherent optical method for performing discrete Fourier transforms. , 1978, Optics letters.

[4]  M. Rem,et al.  Cost and performance of VLSI computing structures , 1979, IEEE Transactions on Electron Devices.

[5]  H. B. Bakoglu,et al.  Circuits, interconnections, and packaging for VLSI , 1990 .

[6]  F.J. Leonberger,et al.  Optical interconnections for VLSI systems , 1984, Proceedings of the IEEE.

[7]  Sing H. Lee,et al.  Diffractive lenses utilizing orthogonal cylindrical Fresnel zone plates , 1990, Photonics West - Lasers and Applications in Science and Engineering.

[8]  Bing J. Sheu,et al.  Hardware annealing in analog VLSI neurocomputing , 1990 .

[9]  S H Lee,et al.  Dual-scale topology optoelectronic processor. , 1991, Optics letters.

[10]  Mark A. Franklin,et al.  Pin Limitations and Partitioning of VLSI Interconnection Networks , 1982, IEEE Transactions on Computers.

[11]  S H Lee,et al.  Comparison between optical and electrical interconnects based on power and speed considerations. , 1988, Applied optics.

[12]  Barmak Mansoorian,et al.  Design And Implementation Of Flip-chip Bonded Si/PLZT Smart Pixels , 1992, Summer Topical Meeting Digest on Broadband Analog and Digital Optoelectronics, Optical Multiple Access Networks, Integrated Optoelectronics, Smart Pixels.

[13]  Sadik C. Esener,et al.  Two-Dimensional Silicon/PLZT Spatial Light Modulators: Design Considerations And Technology , 1986 .

[14]  B K Jenkins,et al.  Architectural implications of a digital optical processor. , 1984, Applied optics.

[15]  S. Esener,et al.  A scalable optoelectronic neural system using free-space optical interconnects , 1992, IEEE Trans. Neural Networks.