We present the design of a planar-integrated optoelectronic vector-matrix multiplier. The inherent parallel-processing potential is fully exploited by optical implementation of multiplications and summations. Planar integration makes the free-space optical system compatible with electronic VLSI technologies. It is composed of phase-only diffractive optical elements, which implement lens and multiple-beam-splitter functions. A demonstrator version of the optical system for a matrix of size 10 x 10 was fabricated on quartz glass by means of multimask lithography and reactive ion etching. It shows low cross talk and good uniformity of the signals.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1364/ao.39.005367 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Memristive Field-Programmable Analog Arrays for Analog Computing.
Adv Mater
September 2023
Department of Electrical and Computer Engineering, University of Massachusetts, Amherst, MA, 01003, USA.
The increasing interests in analog computing nowadays call for multipurpose analog computing platforms with reconfigurability. The advancement of analog computing, enabled by novel electronic elements like memristors, has shown its potential to sustain the exponential growth of computing demand in the new era of analog data deluge. Here, a platform of a memristive field-programmable analog array (memFPAA) is experimentally demonstrated with memristive devices serving as a variety of core analog elements and CMOS components as peripheral circuits.
View Article and Find Full Text PDFNonideality-Aware Training for Accurate and Robust Low-Power Memristive Neural Networks.
Adv Sci (Weinh)
June 2022
Department of Electronic and Electrical Engineering, University College London, London, WC1E 7JE, UK.
Recent years have seen a rapid rise of artificial neural networks being employed in a number of cognitive tasks. The ever-increasing computing requirements of these structures have contributed to a desire for novel technologies and paradigms, including memristor-based hardware accelerators. Solutions based on memristive crossbars and analog data processing promise to improve the overall energy efficiency.
View Article and Find Full Text PDFOptics is a promising platform in which to help realize the next generation of fast, parallel, and energy-efficient computation. We demonstrate a reconfigurable free-space optical multiplier that is capable of over 3000 computations in parallel, using spatial light modulators with a pixel resolution of only 340×340. This enables vector-matrix multiplication and parallel vector-vector multiplication with vector size of up to 56.
View Article and Find Full Text PDFSilicon-on-insulator Mach-Zehnder interferometer structures that utilize a photonic crystal nanobeam waveguide in each of two connecting arms are proposed here as efficient 2 × 2 resonant, wavelength-selective electro-optical routing switches that are readily cascaded into on-chip N × N switching networks. A localized lateral PN junction of length ~2 μm within each of two identical nanobeams is proposed as a means of shifting the transmission resonance by 400 pm within the 1550 nm band. Using a bias swing ΔV = 2.
View Article and Find Full Text PDFA low-power, time-division-multiplexed vector matrix-multiplier for a vestibular prosthesis.
Annu Int Conf IEEE Eng Med Biol Soc
December 2015
A custom analog vector matrix multiplier (VMM) for a vestibular prosthesis is reported. The VMM functions to reduce misalignment between implanted angular rate sensors and associated peripheral sense organs and precompensate for spurious electrical stimulation of vestibular neurons. Operating in the CMOS subthreshold region, the VMM performs a 3-by-3 vector matrix multiplication of rate sensor outputs, magnitude <; ±250 mV, and bandwidth <; 1.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!