Related papers: Energy-efficient neural network inference with mic…
Photonic computing is a computing paradigm which have great potential to overcome the energy bottlenecks of electronic von Neumann architecture. Throughput and power consumption are fundamental limitations of…
The explosive growth of data and its related energy consumption is pushing the need to develop energy-efficient brain-inspired schemes and materials for data processing and storage. Here, we demonstrate experimentally that Co/Pt films can…
Exciton polaritons based on atomically thin semiconductors are essential building blocks of quantum optoelectronic devices. Their properties are governed by an ultrafast and oscillatory energy transfer between their excitonic and photonic…
Sparse neural networks can greatly facilitate the deployment of neural networks on resource-constrained platforms as they offer compact model sizes while retaining inference accuracy. Because of the sparsity in parameter matrices, sparse…
We present an approach to adaptively utilize deep neural networks in order to reduce the evaluation time on new examples without loss of accuracy. Rather than attempting to redesign or approximate existing networks, we propose two schemes…
External driving of spinor degrees of freedom by magnetic or optical fields in quantum systems underpin many applications ranging from nuclear magnetic resonance to coherent state control in quantum computing. Although spinor polariton…
Recent works on Binary Neural Networks (BNNs) have made promising progress in narrowing the accuracy gap of BNNs to their 32-bit counterparts. However, the accuracy gains are often based on specialized model designs using additional 32-bit…
The field of artificial intelligence faces significant challenges in achieving both biological plausibility and computational efficiency, particularly in visual learning tasks. Current artificial neural networks, such as convolutional…
Exciton-polaritons provide a great platform for developing ultrafast all-optical logic gates for quantum and optical chips. However, progress toward practical polariton logic remains limited due to incomplete logical functionality on a…
We investigate the possibility of creating X-waves, or localized wave packets, in resonantly excited exciton-polariton superfluids. We demonstrate the existence of X-wave traveling solutions in the coupled exciton-photon system past the…
Deep convolution Neural Network (DCNN) has been widely used in computer vision tasks. However, for edge devices even inference has too large computational complexity and data access amount. The inference latency of state-of-the-art models…
Convolutional neural networks have recently achieved significant breakthroughs in various image classification tasks. However, they are computationally expensive,which can make their feasible mplementation on embedded and low-power devices…
The realization of exciton-polaritons -- hybrid excitations of semiconductor quantum well excitons and cavity photons -- has been of great technological and scientific significance. In particular, the short-range collisional interaction…
Neural networks have proven effective for solving many difficult computational problems. Implementing complex neural networks in software is very computationally expensive. To explore the limits of information processing, it will be…
By exploiting the polarization multistability of polaritons, we show that polarized signals can be conducted in the plane of a semiconductor microcavity along controlled channels or "neurons". Furthermore due to the interaction of…
Microcavity polaritons are light-matter quasiparticles that arise from the strong coupling between excitons and photons confined in a semiconductor microcavity. They typically operate at visible or near visible wavelengths. They combine the…
We propose an all-optical nonlinear router based on a double barrier gate connected to periodically modulated guides. A semiconductor microcavity is driven nonresonantly in-between the barriers to form an exciton-polariton condensate on a…
We propose theoretically an optical diode based on exciton-polaritons in semiconductor microcavities. A flow of polaritons in the bistable regime is used to send signals through an asymmetric fixed potential that favours the tunneling of…
Deep neural networks (DNNs) have been demonstrated as effective prognostic models across various domains, e.g. natural language processing, computer vision, and genomics. However, modern-day DNNs demand high compute and memory storage for…
For successful deployment of deep neural networks on highly--resource-constrained devices (hearing aids, earbuds, wearables), we must simplify the types of operations and the memory/power resources used during inference. Completely avoiding…