摘要:
One embodiment of the present invention provides a demultiplexer implemented as a nanowire crossbar or a hybrid nanowire/microscale-signal-line crossbar with resistor-like nanowire junctions. The demultiplexer of one embodiment provides demultiplexing of signals input on k microscale address lines to 2k or fewer nanowires, employing supplemental, internal address lines to map 2k nanowire addresses to a larger, internal, n-bit address space, where n>k. A second demultiplexer embodiment of the present invention provides demultiplexing of signals input on n microscale address lines to 2k nanowires, with n>k, using 2k, well-distributed, n-bit external addresses to access the 2k nanowires. Additional embodiments of the present invention include a method for evaluating different mappings of nanowire addresses to internal address-spaces of different sizes, or to evaluate mappings of nanowires to external address-spaces of different sizes, metrics for evaluating address mapping and demultiplexer designs, and demultiplexer design methods.
摘要:
Embodiments of the present invention are employ dynamical, nanoscale devices, including memristive connections between nanowires, for constructing parallel, distributed, dynamical computational networks and systems, including perceptron networks and neural networks. In many embodiments of the present invention, neuron-like computational devices are constructed from silicon-based microscale and/or submicroscale components, and interconnected with one another by dynamical interconnections comprising nanowires and memristive connections between nanowires. In many massively parallel, distributed, dynamical computing systems, including the human brain, there may be a far greater number of interconnections than neuron-like computational nodes. Use of dynamical nanoscale devices for these connections results in enormous design, space, energy, and computational efficiencies.
摘要:
Embodiments of the present invention are employ dynamical, nanoscale devices, including memristive connections between nanowires, for constructing parallel, distributed, dynamical computational networks and systems, including perceptron networks and neural networks. In many embodiments of the present invention, neuron-like computational devices are constructed from silicon-based microscale and/or submicroscale components, and interconnected with one another by dynamical interconnections comprising nanowires and memristive connections between nanowires. In many massively parallel, distributed, dynamical computing systems, including the human brain, there may be a far greater number of interconnections than neuron-like computational nodes. Use of dynamical nanoscale devices for these connections results in enormous design, space, energy, and computational efficiencies.
摘要:
A standalone point-of-sale terminal in a unitary physical package has input/output devices for communicating transaction data between the terminal and an operator, and also has a cash drawer. These devices and a direct-access storage are controlled by a programmable processor which executes low-level routines from read-only addressable memory and serially replaceable high-level applications programs from read/write addressable memory. Both applications programs and transaction data are stored on a single non-volatile, removable storage medium.