Related papers: Photonic systolic array for all-optical matrix-mat…
Photonic and polaritonic systems offer a fast and efficient platform for accelerating machine learning (ML) through physics-based computing. To gain a computational advantage, however, polaritonic systems must: (1) exploit features that…
We present a deep neural network approach for encoding microphone array signals into Ambisonics that generalizes to arbitrary microphone array configurations with fixed microphone count but varying locations and frequency-dependent…
Recent success in deep neural networks has generated strong interest in hardware accelerators to improve speed and energy consumption. This paper presents a new type of photonic accelerator based on coherent detection that is scalable to…
In this paper, we propose a novel fully programmable linear photonic processor, which we call LightPro, with improved scalability, performance, and footprint. At the heart of LightPro are compact, low-loss, and programmable silicon photonic…
Hamiltonian simulation is a key workload in quantum computing, enabling the study of complex quantum systems and serving as a critical tool for classical verification of quantum devices. However, it is computationally challenging because…
An efficient numerical algorithm is presented for massively parallel simulations of dispersion-managed wavelength-division-multiplexed optical fiber systems. The algorithm is based on a weak nonlinearity approximation and independent…
Many real-world mission-critical applications require continual online learning from noisy data and real-time decision making with a defined confidence level. Probabilistic models and stochastic neural networks can explicitly handle…
A silicon photomultiplier (SiPM) array has been developed at FBK-irst having 32 channels and a dimension of 8.0 x 1.1 mm^2. Each 250 um wide channel is subdivided into 5 x 22 rectangularly arranged pixels. These sensors are developed to…
Modern large-scale deep learning workloads highlight the need for parallel execution across many devices in order to fit model data into hardware accelerator memories. In these settings, array redistribution may be required during a…
A multispectral filter array (MSFA) is one solution for capturing a multispectral image (MSI) in a single shot at low cost. We introduce our optimization method of the spectral sensitivity of the MSFAs and demosaicking, and show a new…
This paper presents a series of user parameter-free iterative Sparse Asymptotic Minimum Variance (SAMV) approaches for array processing applications based on the asymptotically minimum variance (AMV) criterion. With the assumption of…
Despite stereo matching accuracy has greatly improved by deep learning in the last few years, recovering sharp boundaries and high-resolution outputs efficiently remains challenging. In this paper, we propose Stereo Mixture Density Networks…
Adaptive representations are increasingly indispensable for reducing the in-memory and on-disk footprints of large-scale data. Usual solutions are designed broadly along two themes: reducing data precision, e.g., through compression, or…
There is an urgent need to enhance the storage density of memory devices to accommodate the exponentially increasing amount of data generated by humankind. In this work, we describe Magnonic Combinatorial Memory (MCM), where the bits of…
Colloidal plasmonic-photonic crystals represent a class of hybrid materials composed of a dielectric colloidal spheres photonic lattice and a metal plasmonic film. In this work, the optical properties of a linear array colloidal…
Here it is proposed a three-dimensional plasmonic nonvolatile memory crossbar arrays that can ensure a dual-mode operation in electrical and optical domains. This can be realized through plasmonics that serves as a bridge between photonics…
On-chip optical neural networks (ONNs) have recently emerged as an attractive hardware accelerator for deep learning applications, characterized by high computing density, low latency, and compact size. As these networks rely heavily on…
We present a new formulation for parallel matrix multiplication (MM) to out-perform the standard row-column code design. This algorithm is formulated in the MoA formalism (A Mathematics of Arrays) and combines an array view of hardware…
There has been growing interest in using photonic processors for performing neural network inference operations; however, these networks are currently trained using standard digital electronics. Here, we propose on-chip training of neural…
Transformers are at the core of modern AI nowadays. They rely heavily on matrix multiplication and require efficient acceleration due to their substantial memory and computational requirements. Quantization plays a vital role in reducing…