21 January 2025 in Journal of Lightwave Technology
The prowess of Artificial Intelligence (AI) and Large Language Models (LLMs) to compute the exponentially growing data, coupled with the digital electronic AI computing systems reaching their physical plateaus, have stimulated extensive research into next-gen AI accelerators. Optical Neural Networks (ONNs) have been among the most favored candidates, since the unparalleled speed of light can provide massive compute parallelism by the physical dimensions of time, wavelength, and space, offering higher computational speeds and lower power consumption. In this paper, we experimentally demonstrate novel time-space-wavelength multiplexed arrayed waveguided gratings router (AWGR)-based architectures exploited as photonic matrix- and tensor-multiplier engines achieving a total computational power of up to 163.8 TOPS. This marks a substantial ∼15x improvement over state-of-the-art waveguide-based optical accelerators. The two demonstrators were tested for three different NN applications, under two experimental testbeds. The matrix-multiplier was evaluated as a fully connected NN (FCNN), classifying images from the FMNIST dataset. An 8×8 AWGR was employed for providing the passive routing, with modulators driven at 10Gbaud, exhibiting a hardware-based inference accuracy of 87.1% with respect to the software. The tensor-multiplier, using a 16x16 AWGR, was evaluated as a FCNN and a convolutional NN (CNN), detecting DDoS attacks and classifying handwritten digits at 20 Gbaud. The experimental results exhibited a Cohen's kappa score of 0.799 for the DDoS detection and a classification accuracy of 93.35%, respectively.
12 September 2022 in Journal of Lightwave Technology
During the past years, the limited access-times and memory bandwidth in electronic random-access memory (RAM) has triggered a concerted research effort towards deploying optical memory circuitry for exploiting the inherent high-bandwidth characteristics of light-based technologies. Yet the deployment of optical memory elements in functional optical RAM demonstrations has still been limited to single-bit capacity levels, with the few multi-bit demonstrations still offering only storage functionality without any true random-access capabilities. In this article, we demonstrate an experimental 16-bit all-optical RAM bank prototype that follows a 4 × 4 matrix organization and supports addressable random access storage of four 4-bit WDM-formatted optical Data Words at a 20 Gb/s memory-throughput. The proposed architecture constitutes of completely passive Row/Column decoding (RD/CD) subsystems, four Semiconductor Optical Amplifier Mach-Zehnder Interferometer (SOA-MZI) Access Gates (AGs) for enabling the access towards the 4 available Word Lines (WLs) and 16 all-optical Flip-Flops (AOFF) arranged in a 4 × 4 RAM bank. Each FF is incorporated in an indicative photonic integrated circuit (PIC). The experimental validation of the complete 16-bit RAM bank reveals error-free operation in Write mode with a peak power penalty between 8.3–10.2 dB at 10 −9 bit-error-rate (BER) measurements, indicating a roadmap for two-dimensional addressable (2D) RAM banks that can penetrate the optical computing application domain.
16 September 2022 in Journal of Lightwave Technology
We report for the first time, to the best of our knowledge, the experimental demonstration of a complete all-optical 2-bit Address Look-Up (AL) table that performs address matching and forwarding and operates at 10Gb/s. The circuit incorporates an optical Content Addressable Memory (CAM) and an optical Random Access Memory (RAM) table that are interconnected via an optical Encoding/Decoding circuit. The AL table has been experimentally validated for different sets of CAM and RAM word contents as well as for different CAM and RAM row interconnect settings at 10Gb/s, revealing a power penalty of around 9 dB at a bit-error-rate (BER) of 10 -9 . Furthermore, we demonstrate experimental results of a fully integrated InP-based optical CAM cell, which includes an integrated XGM Flip-Flop (FF) memory cell and a Mach-Zehnder Interferometer (MZI) XNOR Access Gate (AG) and supports content addressing operation at 5 Gb/s. This fully-integrated optical CAM cell indicates a possible integration roadmap for the AL constituent building blocks towards the deployment of integrated optical ALs that can be utilized in optical routing or, more general, in optical content matching and forwarding operations.
16 September 2021 in Journal of Lightwave Technology
During the past years, optical memories have emerged as the main way to enable fast access times and increased memory bandwidth in synergy with optical interconnections. The migration, however, to fully functional and practical optical RAMs and cache memories will require an additional yet essential step: Optical memories have to get organized into an optical memory subsystem for delivering seamless cooperation between storage and peripheral decoding units, exploiting also the wavelength dimension for efficient RAM architectural layouts. In this paper we demonstrate the first experimentally verified Wavelength Division Multiplexing (WDM) -enabled optical RAM Row subsystem comprising both the peripheral decoding circuitry and a 4-bit optical RAM Row for storing 20Gb/s 4-bit WDM-formatted optical data words. The proposed scheme comprises a single shared multi-λ Semiconductor Optical Amplifier Mach-Zehnder Interferometer (SOA-MZI) Access Gate (AG) for granting access-control to the complete RAM Row, a passive Column Decoder (CD) that directs the incoming WDM-formatted data words to the respective RAM cells that are in turn based on four broadband Indium Phosphide (InP) photonic integrated Flip-Flops (FFs). Our experimental verification firstly investigates the broadband capability of the elementary InP monolithic RAM cell at 5 Gb/s, revealing error-free Write functionality with <4.5 dB power penalty across the whole C-band confirming in this way the potential of efficient scaling to multi-cell WDM-enabled RAMs. Then, experimental validation of the complete 4-bit WDM-enabled optical RAM Row with 20 Gb/s WDM-formatted data words is performed, featuring simultaneous Access gating, decoding and storage operation for all four RAM cells, having a peak power penalty within the range of [4.6-5.6] dB for Write and [2.2-2.9] dB for Read functionalities, respectively, at a Bit Error Rate of 10 −9 .
1 November 2025 in APL Photonics
The ever-increasing volume of data demarcating from the exponential scale of Artificial Intelligence (AI) and Deep Learning (DL) models motivated research into specialized AI accelerators in order to complement digital processors. Photonic Neural Networks (PNNs), with their unique ability to capitalize on the interplay of multiple physical dimensions, including time, wavelength, and space, have been brought forward with a credible promise for boosting computational power and energy efficiency in AI processors. In this article, we experimentally demonstrate a novel multidimensional arrayed waveguide grating router (AWGR)-based photonic AI accelerator that can offload bandwidth-bounded linear algebra while leaving memory hierarchy, control, and nonlinearities to electronics and can execute tensor multiplications at a record-high total computational power of 262 TOPS, offering a ∼24× improvement over the existing waveguide-based optical accelerators. It consists of a 16 × 16 AWGR that exploits the time-, wavelength-, and space-division multiplexing (T-W-SDM) for weight and input encoding, together with an integrated Si3N4-based frequency comb for multi-wavelength generation. The photonic AI accelerator has been experimentally validated in both Fully Connected (FC) and Convolutional NN (CNN) models, with the FC and CNN being trained for DDoS attack identification and MNIST classification, respectively. The experimental inference at 32 Gbaud achieved a Cohen’s kappa score of 0.8652 for DDoS detection and an accuracy of 92.14% for MNIST classification, respectively, closely matching the software performance.
22 February 2023 in IEEE Journal of Selected Topics in Quantum Electronics
The need of faster access-times and increased memory bandwidths has triggered a concerted research effort towards deploying optical memory circuitry, targeting at the apex of the memory hierarchy pyramid that is traditionally reserved for the fastest access time memory configurations. The advances in integrated photonics allowed for remarkable progress in the field of optical memory implementations, but the majority of optical random-access memory (RAM) demonstrations still expands along single cell developments. A complete optical cache memory demonstration is currently completely missing, with the limited amount of multi-bit optical memory layouts reported so far failing to incorporate the rich organizational and functional framework offered by a cache memory. In this article, we demonstrate experimentally the first 16-bit photonic cache memory hardware prototype, organized in four addressable cache lines with a 4-bit wavelength-division multiplexing (WDM)-optical Data Word stored per line, capable of performing at a 20 Gb/s memory-throughput. The proposed optical cache layout combines a WDM-enabled optical RAM bank and a complete set of cache peripherals, implementing for the first time all cache functionalities directly in the optical domain.