Abstract:
A filter to transmit incident radiation at a predetermined incidence angle includes a plurality of photonic crystal structures disposed substantially along a surface normal direction of the filter. The photonic crystal structure includes a multilayer cell that comprises a first layer having a first dielectric permittivity, and a second layer having a second dielectric permittivity different from the first dielectric permittivity. The first layer and the second layer define a Brewster angle substantially equal to the predetermined incidence angle based on the first dielectric permittivity and the second permittivity. Each photonic crystal structure in the plurality of photonic crystal structures defines a respective bandgap, and the respective bandgaps of the plurality of photonic crystal structures, taken together, cover a continuous spectral region of about 50 nm to about 100 mm.
Abstract:
An optical neural network is constructed based on photonic integrated circuits to perform neuromorphic computing. In the optical neural network, matrix multiplication is implemented using one or more optical interference units, which can apply an arbitrary weighting matrix multiplication to an array of input optical signals. Nonlinear activation is realized by an optical nonlinearity unit, which can be based on nonlinear optical effects, such as saturable absorption. These calculations are implemented optically, thereby resulting in high calculation speeds and low power consumption in the optical neural network.
Abstract:
An apparatus includes at least one conductive layer, an electromagnetic (EM) wave source, and an electron source. The conductive layer has a thickness less than 5 nm. The electromagnetic (EM) wave source is in electromagnetic communication with the at least one conductive layer and transmits a first EM wave at a first wavelength in the at least one conductive layer so as to generate a surface plasmon polariton (SPP) field near a surface of the at least one conductive layer. The electron source propagates an electron beam at least partially in the SPP field so as to generate a second EM wave at a second wavelength less than the first wavelength.
Abstract:
A photonic parallel network can be used to sample combinatorially hard distributions of Ising problems. The photonic parallel network, also called a photonic processor, finds the ground state of a general Ising problem and can probe critical behaviors of universality classes and their critical exponents. In addition to the attractive features of photonic networks—passivity, parallelization, high-speed and low-power—the photonic processor exploits dynamic noise that occurs during the detection process to find ground states more efficiently.