Photonic Neural Networks (PNNs) implemented on silicon photonic (SiPho) platforms stand out as a promising candidate to endow neural network hardware, offering the potential for energy efficient and ultra-fast computations through exploiting the unique primitives of light i.e., THz bandwidth, low-power and low-latency. In this paper, we review the state-of-the-art photonic linear processors discuss their challenges and propose solutions for future photonic-assisted machine learning engines. Additionally, we will present experimental results on the recently introduced SiPho 4x4 coherent crossbar (Xbar) architecture, that migrates from existing Singular Value Decomposition (SVD)-based schemes while offering single time-step programming complexity. The Xbar architecture utilizes silicon germanium (SiGe) Electro-Absorption Modulators (EAMs) as its computing cells and Thermo-Optic (TO) Phase Shifters (PS) for providing the sign information at every weight matrix node. Towards experimentally evaluating our Xbar architecture, we performed 10,024 arbitrary linear transformations over the SiPho processor, with the respective fidelity values converging to 100%. Followingly, we focus on the execution of the non-linear part of the NN by demonstrating a programmable analog optoelectronic circuit that can be configured to provide a plethora of non-linear activation functions, including tanh, sigmoid, ReLU and inverted ReLU at 2 GHz update rate. Finally, we provide a holistic overview on optics-informed neural networks towards improving the classification accuracy and performance of optics-specific Deep Learning (DL) computational tasks by leveraging the synergy of optical physics and DL.
In this work, we present the design process and experimental evaluation of a 1×2 asymmetric power splitters based on the self-imaging principle that is applied on an ultra-low-loss 800nm thick Si3N4 platform. The asymmetry in the multimode interference region is induced by removing a rectangular piece from the edge of the coupler, prompting a disruption at the interference pattern and adjusting accordingly the splitting power ratio. The design of the MMIs operating in the 1500- 1600nm wavelength region was realized through 3D-FDTD calculation method and the experimental results agree with theory providing an error of 5% in splitting ratio and less than -0.6dB insertion losses.
Integrated tunable lasers based on the co-integration of InP-based SOAs with low-loss Si3N4 dielectric waveguides have emerged as promising solutions in applications where the control of light phase is fundamental. Μicrowave photonics, coherent communications and LIDARs are only some of the applications where sub-KHz linewidths have already been achieved. Nevertheless, the majority of these demonstrations are based on Si3N4 platforms featuring thicknesses lower than 300nm and providing modes with effective indices below 1.6 imposing a major restriction on the achievable FSR values and devices’ footprint. In this work, we present the design of Vernier ring-inspired reflectors based on an 800nm- thick Si3N4 platform providing a TE fundamental mode with an effective index close to 1.71 for a width of 800nm, a group index close to 2.08 at λ=1550nm wavelength, and propagation losses as low as 0.2dB/cm. The proposed thick- Si3N4 designs are based on a simple dual ring Vernier configuration achieving an experimental FSR near 38nm and a 15dB side-mode suppression. These results are in close agreement with the ones obtained theoretically through a detailed Transfer Matrix Formulation verifying the accuracy of the presented semi-analytical model. This simulation model is then employed for the prediction of the performance of more advanced structures such as triple cascaded and high-order Vernier Ring designs, towards extending the achievable FSR and SMSR metrics.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.