The rapid development of artificial neural networks and applied artificial intelligence has led to many applications. However, current software implementation of neural networks is severely limited in terms of performance and energy efficiency. It is believed that further progress requires the development of neuromorphic systems, in which hardware directly mimics the neuronal network structure of a human brain. Here, we propose theoretically and realize experimentally an optical network of nodes performing binary operations. The nonlinearity required for efficient computation is provided by semiconductor microcavities in the strong quantum light-matter coupling regime, which exhibit exciton–polariton interactions. We demonstrate the system performance against a pattern recognition task, obtaining accuracy on a par with state-of-the-art hardware implementations. Our work opens the way to ultrafast and energy-efficient neuromorphic systems taking advantage of ultrastrong optical nonlinearity of polaritons.

Multicomponent Bose–Einstein condensates, quantum Hall systems, and chiral magnetic materials display twists and knots in the continuous symmetries of their order parameters known as skyrmions. Originally discovered as solutions to the nonlinear sigma model in quantum field theory, these vectorial excitations are quantified by a topological winding number dictating their interactions and global properties of the host system. Here, we report the experimental observation of a stable individual second-order meron and antimeron appearing in an electromagnetic field. We realize these complex textures by confining light into a liquid-crystal-filled cavity that, through its anisotropic refractive index, provides an adjustable artificial photonic gauge field that couples the cavity photon motion to its polarization, resulting in the formation of these fundamental vectorial vortex states of light. Our observations could help bring topologically robust room-temperature optical vector textures into the field of photonic information processing and storage.

We comparatively study donor-induced quantum dots in Si nanoscale-channel transistors for a wide range of doping concentration by analysis of single-electron tunneling transport and surface potential measured by Kelvin probe force microscopy (KPFM). By correlating KPFM observations of donor-induced potential landscapes with simulations based on Thomas-Fermi approximation, it is demonstrated that single-electron tunneling transport at lowest gate voltages (for smallest coverage of screening electrons) is governed most frequently by only one dominant quantum dot, regardless of doping concentration. Doping concentration, however, primarily affects the internal structure of the quantum dot. At low concentrations, individual donors form most of the quantum dots, i.e., “donor-atom” quantum dots. In contrast, at high concentrations above metal-insulator transition, closely placed donors instead of individual donors form more complex quantum dots, i.e., “donor-cluster” quantum dots. The potential depth of these “donor-cluster” quantum dots is significantly reduced by increasing gate voltage (increasing coverage of screening electrons), leading to the occurrence of multiple competing quantum dots.

We propose all-optical neural networks characterized by very high energy efficiency and performance density of inference. We argue that the use of microcavity exciton polaritons allows one to take advantage of the properties of both photons and electrons in a seamless manner. This results in strong optical nonlinearity without the use of optoelectronic conversion. We propose a design of a realistic neural network and estimate energy cost to be at the level of attojoules per bit, also when including the optoelectronic conversion at the input and output of the network, several orders of magnitude below state-of-the-art hardware implementations. We propose two kinds of nonlinear binarized nodes based either on optical phase shifts and interferometry or on polariton spin rotations.