Thomas Ferreira de Lima

Recent progress in semiconductor excitable lasers for photonic spike processing

Recently, there has been tremendous interest in excitable optoelectronic devices and in particular excitable semiconductor lasers that could potentially enable unconventional processing approaches beyond conventional binary-logic-based approaches. In parallel, there has been renewed investigation of non-von Neumann architectures driven in part by incipient limitations in aspects of Moore’s law. These neuromorphic architectures attempt to decentralize processing by interweaving interconnection with computing while simultaneously incorporating time-resolved dynamics, loosely classified as spiking (a.k.a. excitability). The rapid and efficient advances in CMOS-compatible photonic interconnect technologies have led to opportunities in optics and photonics for unconventional circuits and systems. Effort in the budding research field of photonic spike processing aims to synergistically integrate the underlying physics of photonics with bio-inspired processing. Lasers operating in the excitable regime are dynamically analogous with the spiking dynamics observed in neuron biophysics but roughly 8 orders of magnitude faster. The field is reaching a critical juncture at which there is a shift from studying single devices to studying an interconnected network of lasers. In this paper, we review the recent research in the information processing abilities of such lasers, dubbed “photonic neurons,” “laser neurons,” or “optical neurons.” An integrated network of such lasers on a chip could potentially grant the capacity for complex, ultrafast categorization and decision making to provide a range of computing and signal processing applications, such as sensing and manipulating the radio frequency spectrum and for hypersonic aircraft control.

Multi-channel control for microring weight banks.

We demonstrate 4-channel, 2GHz weighted addition in a silicon microring filter bank. Accurate analog weight control becomes more difficult with increasing number of channels, N, as feedback approaches become impractical and brute force feedforward approaches take O(2N) calibration measurements in the presence of inter-channel dependence. We introduce model-based calibration techniques for thermal cross-talk and cross-gain saturation, which result in a scalable O(N) calibration routine and 3.8 bit feedforward weight accuracy on every channel. Practical calibration routines are indispensible for controlling large-scale microring systems. The effect of thermal model complexity on accuracy is discussed. Weighted addition based on silicon microrings can apply the strengths of photonic manufacturing, wideband information processing, and multiwavelength networks towards new paradigms of ultrafast analog distributed processing.

Neuromorphic photonic networks using silicon photonic weight banks

Photonic systems for high-performance information processing have attracted renewed interest. Neuromorphic silicon photonics has the potential to integrate processing functions that vastly exceed the capabilities of electronics. We report first observations of a recurrent silicon photonic neural network, in which connections are configured by microring weight banks. A mathematical isomorphism between the silicon photonic circuit and a continuous neural network model is demonstrated through dynamical bifurcation analysis. Exploiting this isomorphism, a simulated 24-node silicon photonic neural network is programmed using “neural compiler” to solve a differential system emulation task. A 294-fold acceleration against a conventional benchmark is predicted. We also propose and derive power consumption analysis for modulator-class neurons that, as opposed to laser-class neurons, are compatible with silicon photonic platforms. At increased scale, Neuromorphic silicon photonics could access new regimes of ultrafast information processing for radio, control, and scientific computing.Photonic systems for high-performance information processing have attracted renewed interest. Neuromorphic silicon photonics has the potential to integrate processing functions that vastly exceed the capabilities of electronics. We report first observations of a recurrent silicon photonic neural network, in which connections are configured by microring weight banks. A mathematical isomorphism between the silicon photonic circuit and a continuous neural network model is demonstrated through dynamical bifurcation analysis. Exploiting this isomorphism, a simulated 24-node silicon photonic neural network is programmed using “neural compiler” to solve a differential system emulation task. A 294-fold acceleration against a conventional benchmark is predicted. We also propose and derive power consumption analysis for modulator-class neurons that, as opposed to laser-class neurons, are compatible with silicon photonic platforms. At increased scale, Neuromorphic silicon photonics could access new regimes of ultrafast information processing for radio, control, and scientific computing.

Continuous Calibration of Microring Weights for Analog Optical Networks

Continuously configurable weighted addition is a key function for multivariate analog signal processing. The wavelength-division multiplexed version of weighted addition based on microrings can apply the strengths of photonic integration, wideband photonic information processing, and multiwavelength networks toward new approaches to ultrafast analog distributed processing. In this letter, we experimentally demonstrate single-channel functionality using a 5-GHz signal, showing a continuous range of complementary (+/-) weighting in a bank of silicon microrings using a feedforward controller and balanced detection technique. Weight tuning accuracy with a dynamic range of 9.2 dB (i.e., 3.1 bits) is shown, and steps to improve performance are discussed. Systems of more microring weights and weight banks could address hitherto unsolvable problems in real-time radio frequency processing and high-performance control.

An integrated analog O/E/O link for multi-channel laser neurons

We demonstrate an analog O/E/O electronic link to allow integrated laser neurons to accept many distinguishable, high bandwidth input signals simultaneously. This device utilizes wavelength division multiplexing to achieve multi-channel fan-in, a photodetector to sum signals together, and a laser cavity to perform a nonlinear operation. Its speed outpaces accelerated-time neuromorphic electronics, and it represents a viable direction towards scalable networking approaches.

Microring Weight Banks

Microring weight banks could enable novel signal processing approaches in silicon photonics. We analyze factors limiting channel count in microring weight banks, which are central to analog wavelength-division multiplexed processing networks in silicon. We find that microring weight banks require a fundamentally different analysis compared to other wavelength-division multiplexing circuits (e.g., demultiplexers). By introducing a quantitative description of independent weighting, we establish performance tradeoffs between channel count and power penalty. This performance is significantly affected by coherent multiresonator interactions through bus waveguides. We experimentally demonstrate these effects in a fabricated device. Analysis relies on the development of a novel simulation technique combining parametric programming with generalized transmission theory. Experimental measurement fitting of an 8-channel weight bank is presented as an example of another application of the simulator.

Progress in neuromorphic photonics

Abstract As society’s appetite for information continues to grow, so does our need to process this information with increasing speed and versatility. Many believe that the one-size-fits-all solution of digital electronics is becoming a limiting factor in certain areas such as data links, cognitive radio, and ultrafast control. Analog photonic devices have found relatively simple signal processing niches where electronics can no longer provide sufficient speed and reconfigurability. Recently, the landscape for commercially manufacturable photonic chips has been changing rapidly and now promises to achieve economies of scale previously enjoyed solely by microelectronics. By bridging the mathematical prowess of artificial neural networks to the underlying physics of optoelectronic devices, neuromorphic photonics could breach new domains of information processing demanding significant complexity, low cost, and unmatched speed. In this article, we review the progress in neuromorphic photonics, focusing on photonic integrated devices. The challenges and design rules for optoelectronic instantiation of artificial neurons are presented. The proposed photonic architecture revolves around the processing network node composed of two parts: a nonlinear element and a network interface. We then survey excitable lasers in the recent literature as candidates for the nonlinear node and microring-resonator weight banks as the network interface. Finally, we compare metrics between neuromorphic electronics and neuromorphic photonics and discuss potential applications.

Photonic Implementation of Spike-Timing-Dependent Plasticity and Learning Algorithms of Biological Neural Systems

The neurobiological learning algorithm, spike-timing-dependent plasticity (STDP), is demonstrated in a simple photonic system using the cooperative nonlinear effects of cross gain modulation and nonlinear polarization rotation, and supervised and unsupervised learning using photonic neuron principles are examined. An STDP-based supervised learning scheme is presented which is capable of mimicking a desirable spike pattern through learning and adaptation. Furthermore, unsupervised learning is illustrated by a principal component analysis system operating under similar learning rules. Finally, a photonic-distributed processing network capable of STDP-based unsupervised learning is theoretically explored.

Principles of Neuromorphic Photonics.

Benchmark A standardized task that can be performed by disparate computing approaches, used to assess their relative processing merit in specific cases. Bifurcation A qualitative change in behavior of a dynamical system in response to parameter variation. Examples include cusp (from monostable to bistable), Hopf (from stable to oscillating), and transcritical (exchange of stability between two steady states). Brain-inspired computing (a.k.a. neuroinspired computing) A biologically inspired approach to build processors, devices, and computing models for applications including adaptive control, machine learning, and cognitive radio. Similarities with biological signal processing include architectural, such as distributed; representational, such as analog or spiking; or algorithmic, such as adaptation. Broadcast and Weight A multiwavelength analog networking protocol in which multiple all photonic neuron outputs are multiplexed and distributed to all neuron inputs. Weights are reconfigured by tunable spectral filters. Excitability A far-from-equilibrium nonlinear dynamical mechanism underlying all-or-none responses to small perturbations. Fan-in The number of inputs to a neuron. Layered network A network topology consisting of a series of sets (i.e., layers) of neurons. The neurons in each set project their outputs only to neurons in the subsequent layer. Most commonly used type of network used for machine learning. Metric A quantity assessing performance of a device in reference to a specific computing approach. Microring weight bank A silicon photonic implementation of a reconfigurable spectral filter capable of independently setting transmission at multiple carrier wavelengths. Modulation The act of representing an abstract variable in a physical quantity, such as photon rate (i.e., optical power), free carrier density (i.e., optical gain), and carrier drift (i.e., current). Electro-optic modulators are devices that convert from an electrical signal to the power envelope of an optical signal. Moore’s law An observation that the number of transistors in an integrated circuit doubles every 18 to 24 months, doubling its performance. Multiply-accumulate (MAC) A common operation that represents a single multiplication followed by an addition: a a + (b c). Neural networks A wide class of computing models consisting of a distributed set of nodes, called neurons, interconnected with configurable or adaptable strengths, called weights. Overall neural network behavior canIn an age overrun with information, the ability to process reams of data has become crucial. The demand for data will continue to grow as smart gadgets multiply and become increasingly integrated into our daily lives. Next-generation industries in artificial intelligence services and high-performance computing are so far supported by microelectronic platforms. These data-intensive enterprises rely on continual improvements in hardware. Their prospects are running up against a stark reality: conventional one-size-fits-all solutions offered by digital electronics can no longer satisfy this need, as Moores law (exponential hardware scaling), interconnection density, and the von Neumann architecture reach their limits. With its superior speed and reconfigurability, analog photonics can provide some relief to these problems; however, complex applications of analog photonics have remained largely unexplored due to the absence of a robust photonic integration industry. Recently, the landscape for commercially-manufacturable photonic chips has been changing rapidly and now promises to achieve economies of scale previously enjoyed solely by microelectronics. The scientific community has set out to build bridges between the domains of photonic device physics and neural networks, giving rise to the field of \emph{neuromorphic photonics}. This article reviews the recent progress in integrated neuromorphic photonics. We provide an overview of neuromorphic computing, discuss the associated technology (microelectronic and photonic) platforms and compare their metric performance. We discuss photonic neural network approaches and challenges for integrated neuromorphic photonic processors while providing an in-depth description of photonic neurons and a candidate interconnection architecture. We conclude with a future outlook of neuro-inspired photonic processing.

All-optical digital-to-spike conversion using a graphene excitable laser

Neuromorphic (brain-inspired) photonic systems process information encoded in the pulses of light, i.e., “spikes” that are analog in time but digital in amplitude. Applying these systems to process commonly used digital data requires a simple and effective interfacing solution to converting binary digits into spike sequence in the optical domain. Laser systems offer a variety of useful nonlinear functionalities, including excitable dynamics that can be found in the time-resolved “spiking” of neurons. We propose and demonstrate, both numerically and experimentally, an all-optical digital-to-spike (DTS) conversion scheme using a single graphene excitable laser (GEL) without clock signal synchronization. We first study the DTS conversion mechanism based on the simulation platform of an integrated GEL, which achieve a DTS conversion rate up to 10 Gbps. Our DTS conversion scheme can be operated under flexible input power conditions and exhibits a strong logic-level restoration capability. We then verify the feasibility of our approach via a proof-of-principle experiment where a fiber-based GEL obtains a DTS conversion rate of 40 Kbps, and a bit error rate (BER) of 10−9 with an input power of −24 dBm. This technology can be potentially applied in future neuromorphic photonic systems for information processing and computing.