Title :
Three-dimensional integral optic threshold elements
Author_Institution :
Sci. Res. Inst., Moscow
Abstract :
A novel 3D optical threshold element and an optical neural network computer (ONNC) with a layer-module architecture are proposed. It is demonstrated that the ONNC may be implemented on a single chip with high volume density by placing the neural network on different planes and by using a third dimension for forming the synaptic electrooptic connection system. Such elements can be programmed by mask control to produce a standard set of Boolean functions in terms of a threshold logic basis and appear well-matched to the parallelism of optics
Keywords :
Boolean functions; integrated optoelectronics; masks; neural chips; optical elements; optical logic; optical neural nets; 3D optical threshold element; Boolean functions; integrated optoelectronics; layer-module architecture; mask control; optical neural network computer; parallelism; synaptic electrooptic connection system; threshold logic; High speed optical techniques; Integrated optics; Lighting control; Neural networks; Optical computing; Optical control; Optical devices; Optical films; Optical modulation; Photodetectors;
Conference_Titel :
Neuroinformatics and Neurocomputers, 1992., RNNS/IEEE Symposium on
Conference_Location :
Rostov-on-Don
Print_ISBN :
0-7803-0809-3
DOI :
10.1109/RNNS.1992.268598