Search

The next generation of smart imaging and vision systems will require compact and tunable optical computing hardware to perform high-speed and low-power image processing. These requirements are driving the development of computing metasurfaces to realize efficient front-end analog optical pre-processors, especially for edge-detection capability. Yet, there is still a lack of reconfigurable or programmable schemes, which may drastically enhance the impact of these devices at the system level. Here, we propose and experimentally demonstrate a reconfigurable flat optical image processor using low-loss phase-change nonlocal metasurfaces. The metasurface is configured to realize different transfer functions in spatial frequency space, when transitioning the phase-change material between its amorphous and crystalline phases. This enables edge detection and bright-field imaging modes on the same device. The metasurface is compatible with a large numerical aperture of ~0.5, making it suitable for high resolution coherent optical imaging microscopy. The concept of phase-change reconfigurable nonlocal metasurfaces may enable emerging applications of artificial intelligence-assisted imaging and vision devices with switchable multitasking., Comment: 20 pages, 5 figures

Biological neural networks effortlessly tackle complex computational problems and excel at predicting outcomes from noisy, incomplete data, a task that poses significant challenges to traditional processors. Artificial neural networks (ANNs), inspired by these biological counterparts, have emerged as powerful tools for deciphering intricate data patterns and making predictions. However, conventional ANNs can be viewed as "point estimates" that do not capture the uncertainty of prediction, which is an inherently probabilistic process. In contrast, treating an ANN as a probabilistic model derived via Bayesian inference poses significant challenges for conventional deterministic computing architectures. Here, we use chaotic light in combination with incoherent photonic data processing to enable high-speed probabilistic computation and uncertainty quantification. Since both the chaotic light source and the photonic crossbar support multiple independent computational wavelength channels, we sample from the output distributions in parallel at a sampling rate of 70.4 GS/s, limited only by the electronic interface. We exploit the photonic probabilistic architecture to simultaneously perform image classification and uncertainty prediction via a Bayesian neural network. Our prototype demonstrates the seamless cointegration of a physical entropy source and a computational architecture that enables ultrafast probabilistic computation by parallel sampling.

Plastic self-adaptation, nonlinear recurrent dynamics and multi-scale memory are desired features in hardware implementations of neural networks, because they enable them to learn, adapt and process information similarly to the way biological brains do. In this work, we experimentally demonstrate these properties occurring in arrays of photonic neurons. Importantly, this is realised autonomously in an emergent fashion, without the need for an external controller setting weights and without explicit feedback of a global reward signal. Using a hierarchy of such arrays coupled to a backpropagation-free training algorithm based on simple logistic regression, we are able to achieve a performance of 98.2% on the MNIST task, a popular benchmark task looking at classification of written digits. The plastic nodes consist of silicon photonics microring resonators covered by a patch of phase-change material that implements nonvolatile memory. The system is compact, robust, and straightforward to scale up through the use of multiple wavelengths. Moreover, it constitutes a unique platform to test and efficiently implement biologically plausible learning schemes at a high processing speed.

Electronically reprogrammable photonic circuits based on phase-change chalcogenides present an avenue to resolve the von-Neumann bottleneck; however, implementation of such hybrid photonic-electronic processing has not achieved computational success. Here, we achieve this milestone by demonstrating an in-memory photonic-electronic dot-product engine, one that decouples electronic programming of phase-change materials (PCMs) and photonic computation. Specifically, we develop non-volatile electronically reprogrammable PCM memory cells with a record-high 4-bit weight encoding, the lowest energy consumption per unit modulation depth (1.7 nJ per dB) for Erase operation (crystallization), and a high switching contrast (158.5%) using non-resonant silicon-on-insulator waveguide microheater devices. This enables us to perform parallel multiplications for image processing with a superior contrast-to-noise ratio (greater than 87.36) that leads to an enhanced computing accuracy (standard deviation less than 0.007). An in-memory hybrid computing system is developed in hardware for convolutional processing for recognizing images from the MNIST database with inferencing accuracies of 86% and 87%.

Plasmonic metasurfaces based on the extraordinary optical transmission effect (EOT) can be designed to efficiently transmit specific spectral bands from the visible to the far-infrared regimes, offering numerous applications in im-portant technological fields such as compact multispectral imaging, biological and chemical sensing, or color displays. However, due to their subwavelength nature, EOT metasurfaces are nowadays fabricated with nano- and micro-lithographic techniques, requiring many processing steps and carried out in expensive cleanroom environments. In this work, we propose and experimentally demonstrate a novel, single-step process for the rapid fabrication of high performance mid- and long-wave infrared EOT metasurfaces employing ultrafast direct laser writing (DLW). Micro-hole arrays composing extraordinary transmission metasurfaces were fabricated over areas of 4 mm2 in timescales of units of minutes, employing single pulse ablation of 40 nm thick Au films on dielectric substrates mounted on a high-precision motorized stage. We show how by carefully characterizing the influence of only three key experimental pa-rameters on the processed micro-morphologies (namely laser pulse energy, scan velocity and beam shaping slit), we can have on-demand control of the optical characteristics of the extraordinary transmission effect in terms of transmission wavelength, quality factor and polarization sensitivity of the resonances. To illustrate this concept, a set of EOT metasurfaces having different performances and operating in different spectral regimes has been successfully designed, fabricated and tested. Comparison between transmittance measurements and numerical simulations have revealed that all the fabricated devices behave as expected, thus demonstrating the high performance, flexibility and reliability of the proposed fabrication method.

Neural processing on devices and circuits is fast becoming a popular approach to emulate biological neural networks. Elaborate CMOS and memristive technologies have been employed to achieve this, including chalcogenide-based in-memory computing concepts. Here we show that nano-scaled films of chalcogenide semiconductors can serve as building-blocks for novel types of neural computations where their tunable electronic and optical properties are jointly exploited. We demonstrate that ultrathin photoactive cavities of Ge-doped Selenide can emulate the computationally powerful non-linear operations of three-factor neo-Hebbian plasticity and the shunting inhibition. We apply this property to solve a maze game through reinforcement learning, as well as a single-neuron solution to the XOR, which is a linearly inseparable problem with point-neurons. Our results point to a new breed of memristors with broad implications for neuromorphic computing.

Over a century ago, Ivan P. Pavlov, in a classic experiment, demonstrated how dogs can learn to associate a ringing bell with food, thereby causing a ring to result in salivation. Today, it is rare to find the use of Pavlovian type associative learning for artificial intelligence (AI) applications even though other learning concepts, in particular backpropagation on artificial neural networks (ANNs) have flourished. However, training using the backpropagation method on 'conventional' ANNs, especially in the form of modern deep neural networks (DNNs), is computationally and energy intensive. Here we experimentally demonstrate a form of backpropagation-free learning using a single (or monadic) associative hardware element. We realize this on an integrated photonic platform using phase-change materials combined with on-chip cascaded directional couplers. We then develop a scaled-up circuit network using our monadic Pavlovian photonic hardware that delivers a distinct machine-learning framework based on single-element associations and, importantly, using backpropagation-free architectures to address general learning tasks. Our approach reduces the computational burden imposed by learning in conventional neural network approaches, thereby increasing speed, whilst also offering higher bandwidth inherent to our photonic implementation., Comment: 24 pages, 5 figures

Research in photonic computing has flourished due to the proliferation of optoelectronic components on photonic integration platforms. Photonic integrated circuits have enabled ultrafast artificial neural networks, providing a framework for a new class of information processing machines. Algorithms running on such hardware have the potential to address the growing demand for machine learning and artificial intelligence, in areas such as medical diagnosis, telecommunications, and high-performance and scientific computing. In parallel, the development of neuromorphic electronics has highlighted challenges in that domain, in particular, related to processor latency. Neuromorphic photonics offers sub-nanosecond latencies, providing a complementary opportunity to extend the domain of artificial intelligence. Here, we review recent advances in integrated photonic neuromorphic systems, discuss current and future challenges, and outline the advances in science and technology needed to meet those challenges., Comment: 21 pages, 6 figures

We present the first demonstration of an integrated photonic phase-change memory using GeSbTe-225 on silicon-on-insulator and demonstrate reliable multilevel operation with a single programming pulse. We also compare our results on silicon with previous demonstrations on silicon nitride. Crucially, achieving this on silicon enables tighter integration of traditional electronics with photonic memories in future, making phase-change photonic memory a viable and integrable technology.

All-dielectric metasurfaces consisting of arrays of nanostructured high-refractive-index materials, typically Si, are re-writing what is achievable in terms of the manipulation of light. Such devices support very strong magnetic, as well as electric, resonances, and are free of ohmic losses that severely limit the performance of their plasmonic counterparts. However, the functionality of dielectric-based metasurfaces is fixed-by-design, i.e. the optical response is fixed by the size, arrangement and index of the nanoresonators. A far wider range of applications could be addressed if active/reconfigurable control were possible. We demonstrate this here, via a new hybrid metasurface concept in which active control is achieved by embedding deeply sub-wavelength inclusions of a tuneable chalcogenide phase-change material within the body of high-index Si nanocylinders. Moreover, by strategic placement of the phase-change layer, and switching of its phase-state, we show selective and active control of metasuface resonances. This yields novel functionality, which we showcase via a dual- to mono-band meta-switch operating simultaneously in the O and C telecommunication bands.

Collocated data processing and storage are the norm in biological systems. Indeed, the von Neumann computing architecture, that physically and temporally separates processing and memory, was born more of pragmatism based on available technology. As our ability to create better hardware improves, new computational paradigms are being explored. Integrated photonic circuits are regarded as an attractive solution for on-chip computing using only light, leveraging the increased speed and bandwidth potential of working in the optical domain, and importantly, removing the need for time and energy sapping electro-optical conversions. Here we show that we can combine the emerging area of integrated optics with collocated data storage and processing to enable all-photonic in-memory computations. By employing non-volatile photonic elements based on the phase-change material, Ge2Sb2Te5, we are able to achieve direct scalar multiplication on single devices. Featuring a novel single-shot Write/Erase and a drift-free process, such elements can multiply two scalar numbers by mapping their values to the energy of an input pulse and to the transmittance of the device, codified in the crystallographic state of the element. The output pulse, carrying the information of the light-matter interaction, is the result of the computation. Our all-optical approach is novel, easy to fabricate and operate, and sets the stage for development of entirely photonic computers.

Atomically thin materials such as graphene are uniquely responsive to charge transfer from adjacent materials, making them ideal charge transport layers in phototransistor devices. Effective implementation of organic semiconductors as a photoactive layer would open up a multitude of applications in biomimetic circuitry and ultra-broadband imaging but polycrystalline and amorphous thin films have shown inferior performance compared to inorganic semiconductors. Here, we utilize the long-range order in rubrene single crystals to engineer organic semiconductor-graphene phototransistors surpassing previously reported photo-gating efficiencies by one order of magnitude. Phototransistors based upon these interfaces are spectrally selective to visible wavelengths and, through photoconductive gain mechanisms, achieve responsivity as large as 10^7 A/W and a detectivity of 1.5 10^9 Jones at room temperature. These findings point towards implementing low-cost, flexible materials for amplified imaging at ultra-low light levels., Comment: Published in Advanced Materials. Supplementary material available at http://onlinelibrary.wiley.com/doi/10.1002/adma.201702993/full

Transient currents in atomically thin MoTe$_2$ field-effect transistor are measured during cycles of pulses through the gate electrode. The transients are analyzed in light of a newly proposed model for charge trapping dynamics that renders a time-dependent change in threshold voltage the dominant effect on the channel hysteretic behavior over emission currents from the charge traps. The proposed model is expected to be instrumental in understanding the fundamental physics that governs the performance of atomically thin FETs and is applicable to the entire class of atomically thin-based devices. Hence, the model is vital to the intelligent design of fast and highly efficient opto-electronic devices., Comment: 25 pages, 7 figures, including the supporting information

We present a stochastic simulator for polycrystalline phase-change materials capable of spatio-temporal modelling of complex anneals. This is based on consideration of bulk and surface energies to generate rates of growth and decay of crystallites built up of `monomers' that themselves may be quite complex molecules. We perform a number of simulations of this model using a Gillespie algorithm. The simulations are performed at molecular scale and using an approximation of local free energy changes that depend only on immediate neighbours. The sites are on a lattice that neither correspond to the crystal lattice nor to individual monomers, but instead gives information about a two-state local phase $r$ (where $r=0$ corresponds to amorphous and 1 corresponds to crystalline) and a continuous crystal orientation $\phi$ at each site. As an example we use this to model crystallisation in chalcogenide GST ($GeSbTe$) alloys used for example in phase-change memory devices, where reversible changes between amorphous and crystalline regimes are used to store and process information. We use our model to simulate anneals of GST including ones with non-trivial spatial and temporal variation of temperature; this gives good agreement to experimental incubation times at low temperatures while modelling non-trivial crystal size distributions and melting dynamics at higher temperatures.