Data processing: artificial intelligence – Neural network – Structure
Patent
1997-07-09
1999-01-26
Downs, Robert W.
Data processing: artificial intelligence
Neural network
Structure
G06F 1518, H01L 218232
Patent
active
058648356
ABSTRACT:
An apparatus for simulating the behavior of a biological neuron. In accordance with one aspect of the invention, a charge-carrying component is fabricated from semiconductor material and is disposed above a semiconductor substrate. The charge-carrying component is separated from the substrate, and substantially surrounded, by an isolating material, such as an oxide (e.g., silicon dioxide). Advantageously, the substantial isolation of the component helps ensure the preservation of electrical charge on the charge-carrying component. Indeed, in accordance with a preferred embodiment, virtually no charge leakage occurs, and absent purposeful charge injection or drainage (described below), the charge deposited on the charge carrying component will be preserved for a very long time. In order to control the quantity of charge carried on the charge-carrying component, a charge injector is disposed adjacent to the charge-carrying component and is configured to controllably transfer charge to the charge-carrying component. Similarly, a drain disposed adjacent to the charge-carrying component and configured to controllably transfer charge away from the charge-carrying component. The charge injector and drain of the preferred embodiment of the present invention are structurally similar to charge injectors and drains conventionally used in EEPROM technology. To generate an output, a charge sensor disposed in proximity with a charge sensing location on the charge-carrying component, and an output is provided and disposed in communication with the charge sensor. As a result, the output delivers an electric signal that is indicative of the charge sensed by the charge sensor.
REFERENCES:
patent: 5258657 (1993-11-01), Shibata et al.
patent: 5263122 (1993-11-01), Nunally
patent: 5343555 (1994-08-01), Yayla et al.
patent: 5411540 (1995-05-01), Edell et al.
patent: 5480820 (1996-01-01), Roth et al.
patent: 5509105 (1996-04-01), Roenker et al.
patent: 5537512 (1996-07-01), Hsia et al.
patent: 5592589 (1997-01-01), Poon
patent: 5615305 (1997-03-01), Nunally
patent: 5621563 (1997-04-01), Boulet et al.
patent: 5627392 (1997-05-01), Diorio et al.
patent: 5636327 (1997-06-01), Nakahira et al.
patent: 5640494 (1997-06-01), Jabri et al.
P. Hasler, et al., "Single Transistor Learning Synapse with Long Term Storage," 1995 IEEE Symp. on Circuits and Systems, vol. 3, pp. 1660-1663, Apr. 1995.
P. Hasler, et al., "Single Transistor Learning Synapses," Advances in Neural Information Processing Systems 7, MIT Press, pp. 817-824, Dec. 1995 .
LandOfFree
Apparatus for simulating a biological neuron does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus for simulating a biological neuron, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus for simulating a biological neuron will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-1458530