New York University

Arizona Board of Regents, a Body Corporate

Frank C Hoppensteadt

Eugene Izhikevich

Andrew David Kent

Varun Narendra

Ferran Macià Bros

A nano-oscillator magnetic wave propagation system has a group of aggregated spin-torque nano-oscillators (ASTNOs), which share a magnetic propagation material. Each of the group of ASTNOs is disposed about an emanating point in the magnetic propagation material. During a non-wave propagation state of the nano-oscillator magnetic wave propagation system, the magnetic propagation material receives a polarizing magnetic field. During a wave propagation state of the nano-oscillator magnetic wave propagation system, each of the group of ASTNOs initiates spin waves through the magnetic propagation material, such that a portion of the spin waves initiated from each of the group of ASTNOs combine to produce an aggregation of spin waves emanating from the emanating point. The aggregation of spin waves may provide a sharper wave front than wave fronts of the individual spin waves initiated from each of the group of ASTNOs.

Aggregated spin-torque nano-oscillators

The present invention relates to a polychronous wave propagation system that is based on relative timing between two or more propagated waves through a wave propagation medium. The relative timing may be associated with interference patterns of energy between the propagated waves. Operational behavior of the polychronous wave propagation system is based on the relative timing of the propagated waves and distances between initiators that transmit the propagated waves and responders that receive the propagated waves. The operational behavior may include arithmetical computations, memory storage, Boolean functions, frequency-based computations, or the like. The polychronous wave propagation system relies on time delays between the propagated waves that result from propagation velocities of the propagated waves through the wave propagation medium. By incorporating the time delays into the system, operational capacity may be greatly enhanced.

Computations using a polychronous wave propagation system

The present invention mimics certain behaviors of a human brain using at least one frequency controllable oscillator block that has multiple operating frequency regions, such that some of the regions tend to lock on to a frequency, while other regions do not have any locking tendencies. A stimulation using at least one stimulation signal based on information input to the system may alter the regions and determine a stable final frequency after a brief processing interval. A control system may process the final frequency to provide system information. A controllable oscillator block may have at least one basin of attraction that corresponds with a stable region of the stimulation signal. Additionally, the controllable oscillator block may function outside of a basin of attraction that corresponds with a non-stable region of the stimulation signal.

Controllable oscillator blocks

A neural network computer () includes a weighting network () coupled to a plurality of phase-locked loop circuits (-N). The weighting network () has a plurality of weighting circuits (C, . . . , C) having output terminals connected to a plurality of adder circuits (-). A single weighting element (C) typically has a plurality of output terminals coupled to a corresponding adder circuit (). Each adder circuit () is coupled to a corresponding bandpass filter circuit () which is in turn coupled to a corresponding phase-locked loop circuit (). The weighting elements (C, . . . , C) are programmed with connection strengths, wherein the connection strengths have phase-encoded weights. The phase relationships are used to recognize an incoming pattern.

Phase-locked loop oscillatory neurocomputer

A neurocomputer () comprises n oscillating processing elements (A,B,C,D andE) that communicate through a common medium () so that there are required only n connective junctions (A,B,C,D andE). A rhythmic external forcing input () modulates the oscillatory frequency of the medium () which, in turn, is imparted to the n oscillators (A,B,C,D andE). Any two oscillators oscillating at different frequencies may communicate provided that input's power spectrum includes the frequency equal to the difference between the frequencies of the two oscillators in question. Thus, selective communication, or dynamic connectivity, between different neurocomputer oscillators occurs due to the frequency modulation of the medium () by external forcing.

Growing community of inventors

Carefree, AZ, United States of America

Frank C Hoppensteadt