Explore the vast range of titles available.

Software implementations of brain-inspired computing underlie many important computational tasks, from image processing to speech recognition, artificial intelligence and deep learning applications. Yet, unlike real neural tissue, traditional computing architectures physically separate the core computing functions of memory and processing, making fast, efficient and low-energy computing difficult to achieve. To overcome such limitations, an attractive alternative is to design hardware that mimics neurons and synapses. Such hardware, when connected in networks or neuromorphic systems, processes information in a way more analogous to brains. Here we present an all-optical version of such a neurosynaptic system, capable of supervised and unsupervised learning. We exploit wavelength division multiplexing techniques to implement a scalable circuit architecture for photonic neural networks, successfully demonstrating pattern recognition directly in the optical domain. Such photonic neurosynaptic networks promise access to the high speed and high bandwidth inherent to optical systems, thus enabling the direct processing of optical telecommunication and visual data. An optical version of a brain-inspired neurosynaptic system, using wavelength division multiplexing techniques, is presented that is capable of supervised and unsupervised learning.

Frequency combs produced by solitons in silicon-based optical microresonators are used to transmit data streams of more than 50 terabits per second in telecommunication wavelength bands. Scaling up telecommunications Frequency combs—light sources that emit a wide spectrum of sharp lines with equally spaced frequencies—have recently become of interest for use in high-capacity optical data transmission. The possibility of producing frequency combs using compact, chip-integrated microresonators promises scalability and practical applicability. Christian Koos et al . make use of a recently developed technique whereby frequency combs are produced by continuously circulating optical solitons—waveforms that preserve their shape during propagation—in silicon-based microresonators. They use two interleaved, chip-based frequency combs to demonstrate transmission of a data stream of more than 50 terabits per second on 179 individual optical carriers in telecommunication wavelength bands. The technology could be used to develop efficient, highly scalable communication systems that could help to address the challenge of a continually growing demand for data capacity. Solitons are waveforms that preserve their shape while propagating, as a result of a balance of dispersion and nonlinearity 1 , 2 . Soliton-based data transmission schemes were investigated in the 1980s and showed promise as a way of overcoming the limitations imposed by dispersion of optical fibres. However, these approaches were later abandoned in favour of wavelength-division multiplexing schemes, which are easier to implement and offer improved scalability to higher data rates. Here we show that solitons could make a comeback in optical communications, not as a competitor but as a key element of massively parallel wavelength-division multiplexing. Instead of encoding data on the soliton pulse train itself, we use continuous-wave tones of the associated frequency comb as carriers for communication. Dissipative Kerr solitons (DKSs) 3 , 4 (solitons that rely on a double balance of parametric gain and cavity loss, as well as dispersion and nonlinearity) are generated as continuously circulating pulses in an integrated silicon nitride microresonator 5 via four-photon interactions mediated by the Kerr nonlinearity, leading to low-noise, spectrally smooth, broadband optical frequency combs 6 . We use two interleaved DKS frequency combs to transmit a data stream of more than 50 terabits per second on 179 individual optical carriers that span the entire telecommunication C and L bands (centred around infrared telecommunication wavelengths of 1.55 micrometres). We also demonstrate coherent detection of a wavelength-division multiplexing data stream by using a pair of DKS frequency combs—one as a multi-wavelength light source at the transmitter and the other as the corresponding local oscillator at the receiver. This approach exploits the scalability of microresonator-based DKS frequency comb sources for massively parallel optical communications at both the transmitter and the receiver. Our results demonstrate the potential of these sources to replace the arrays of continuous-wave lasers that are currently used in high-speed communications. In combination with advanced spatial multiplexing schemes 7 , 8 and highly integrated silicon photonic circuits 9 , DKS frequency combs could bring chip-scale petabit-per-second transceivers into reach.

Microcombs have sparked a surge of applications over the past decade, ranging from optical communications to metrology 1 – 4 . Despite their diverse deployment, most microcomb-based systems rely on a large amount of bulky elements and equipment to fulfil their desired functions, which is complicated, expensive and power consuming. By contrast, foundry-based silicon photonics (SiPh) has had remarkable success in providing versatile functionality in a scalable and low-cost manner 5 – 7 , but its available chip-based light sources lack the capacity for parallelization, which limits the scope of SiPh applications. Here we combine these two technologies by using a power-efficient and operationally simple aluminium-gallium-arsenide-on-insulator microcomb source to drive complementary metal–oxide–semiconductor SiPh engines. We present two important chip-scale photonic systems for optical data transmission and microwave photonics, respectively. A microcomb-based integrated photonic data link is demonstrated, based on a pulse-amplitude four-level modulation scheme with a two-terabit-per-second aggregate rate, and a highly reconfigurable microwave photonic filter with a high level of integration is constructed using a time-stretch approach. Such synergy of a microcomb and SiPh integrated components is an essential step towards the next generation of fully integrated photonic systems. A simple and power-efficient microcomb source is used to drive complementary metal–oxide–semiconductor silicon photonic engines, a step towards the next generation of fully integrated photonic systems.

With the proliferation of ultrahigh-speed mobile networks and internet-connected devices, along with the rise of artificial intelligence (AI) 1 , the world is generating exponentially increasing amounts of data that need to be processed in a fast and efficient way. Highly parallelized, fast and scalable hardware is therefore becoming progressively more important 2 . Here we demonstrate a computationally specific integrated photonic hardware accelerator (tensor core) that is capable of operating at speeds of trillions of multiply-accumulate operations per second (10 12 MAC operations per second or tera-MACs per second). The tensor core can be considered as the optical analogue of an application-specific integrated circuit (ASIC). It achieves parallelized photonic in-memory computing using phase-change-material memory arrays and photonic chip-based optical frequency combs (soliton microcombs 3 ). The computation is reduced to measuring the optical transmission of reconfigurable and non-resonant passive components and can operate at a bandwidth exceeding 14 gigahertz, limited only by the speed of the modulators and photodetectors. Given recent advances in hybrid integration of soliton microcombs at microwave line rates 3 – 5 , ultralow-loss silicon nitride waveguides 6 , 7 , and high-speed on-chip detectors and modulators, our approach provides a path towards full complementary metal–oxide–semiconductor (CMOS) wafer-scale integration of the photonic tensor core. Although we focus on convolutional processing, more generally our results indicate the potential of integrated photonics for parallel, fast, and efficient computational hardware in data-heavy AI applications such as autonomous driving, live video processing, and next-generation cloud computing services. An integrated photonic processor, based on phase-change-material memory arrays and chip-based optical frequency combs, which can operate at speeds of trillions of multiply-accumulate (MAC) operations per second, is demonstrated.

Due to its unique intensity distribution, self-acceleration, and beam self-healing properties, Airy beam holds great potential for optical wireless communications in challenging channels, such as underwater environments. As a vital part of 6G wireless network, the Internet of Underwater Things requires high-stability, low-latency, and high-capacity underwater wireless optical communication (UWOC). Currently, the primary challenge of UWOC lies in the prevalent time-varying and complex channel characteristics. Conventional blue Gaussian beam-based systems face difficulties in underwater randomly perturbed links. In this work, we report a full-color circular auto-focusing Airy beams metasurface transmitter for reliable, large-capacity and long-distance UWOC links. The metasurface is designed to exhibits high polarization conversion efficiency over a wide band (440-640 nm), enabling an increased data transmission rate of 91% and reliable 4 K video transmission in wavelength division multiplexing (WDM) based UWOC data link. The successful application of this metasurface in challenging UWOC links establishes a foundation for underwater interconnection scenarios in 6G communication. Authors present an adaptive underwater optical communication (UWOC) technology based on multi-wavelength lasers and a full-color metasurface for converting visible-band Gaussian to circular autofocusing Airy beams. The potential of Airy beams to mitigate optical power degradation is demonstrated, enabling stable data rate transmission via 4 K video transmission for these systems.

Fiber-optical networks are a crucial telecommunication infrastructure in society. Wavelength division multiplexing allows for transmitting parallel data streams over the fiber bandwidth, and coherent detection enables the use of sophisticated modulation formats and electronic compensation of signal impairments. Optical frequency combs can replace the multiple lasers used for the different wavelength channels. Beyond multiplexing, it has been suggested that the broadband phase coherence of frequency combs could simplify the receiver scheme by performing joint reception and processing of several wavelength channels, but an experimental validation in a fiber transmission experiment remains elusive. Here we demonstrate and quantify joint reception and processing of several wavelength channels in a full transmission system. We demonstrate two joint processing schemes; one that reduces the phase-tracking complexity and one that increases the transmission performance. Frequency combs have the potential to be used as multi-wavelength sources in future optical communications through fiber. Here the authors demonstrate joint phase processing of multi-wavelength comb transmission, and show two schemes to improve performance and reduce complexity.

Tensor analytics lays the mathematical basis for the prosperous promotion of multiway signal processing. To increase computing throughput, mainstream processors transform tensor convolutions into matrix multiplications to enhance the parallelism of computing. However, such order-reducing transformation produces data duplicates and consumes additional memory. Here, we propose an integrated photonic tensor flow processor (PTFP) without digitally duplicating the input data. It outputs the convolved tensor as the input tensor ‘flows’ through the processor. The hybrid manipulation of optical wavelengths, space dimensions, and time delay steps, enables the direct representation and processing of high-order tensors in the optical domain. In the proof-of-concept experiment, an integrated processor manipulating wavelengths and delay steps is implemented for demonstrating the key functionalities of PTFP. The multi-channel images and videos are processed at the modulation rate of 20 Gbaud. A convolutional neural network for video action recognition is demonstrated on the processor, which achieves an accuracy of 97.9%. Convolutional operation is a very efficient way to handle tensor analytics, but it consumes a large quantity of additional memory. Here, the authors demonstrate an integrated photonic tensor processor which directly handles high-order tensors without tensor-matrix transformation.

The past two decades have witnessed the stagnation of the clock speed of microprocessors followed by the recent faltering of Moore’s law as nanofabrication technology approaches its unavoidable physical limit. Vigorous efforts from various research areas have been made to develop power-efficient and ultrafast computing machines in this post-Moore’s law era. With its unique capacity to integrate complex electro-optic circuits on a single chip, integrated photonics has revolutionized the interconnects and has shown its striking potential in optical computing. Here, we propose an electronic-photonic computing architecture for a wavelength division multiplexing-based electronic-photonic arithmetic logic unit, which disentangles the exponential relationship between power and clock rate, leading to an enhancement in computation speed and power efficiency as compared to the state-of-the-art transistors-based circuits. We experimentally demonstrate its practicality by implementing a 4-bit arithmetic logic unit consisting of 8 high-speed microdisk modulators and operating at 20 GHz. This approach paves the way to future power-saving and high-speed electronic-photonic computing circuits. Integrated photonics allows integration of complex optical circuits on a single chip. Here, the authors propose a wavelength division multiplexing based electronic-photonic arithmetic logic unit for computing at high speeds and with improved power consumption to help with the physical limits of Moore’s law.

Countering the optical network ‘capacity crunch’ calls for a radical development in optical fibres that could simultaneously minimize nonlinearity penalties, chromatic dispersion and maximize signal launch power. Hollow-core fibres (HCF) can break the nonlinear Shannon limit of solid-core fibre and fulfil all above requirements, but its optical performance need to be significantly upgraded before they can be considered for high-capacity telecommunication systems. Here, we report a new HCF with conjoined-tubes in the cladding and a negative-curvature core shape. It exhibits a minimum transmission loss of 2 dB km −1 at 1512 nm and a <16 dB km −1 bandwidth spanning across the O, E, S, C, L telecom bands (1302–1637 nm). The debut of this conjoined-tube HCF, with combined merits of ultralow loss, broad bandwidth, low bending loss, high mode quality and simple structure heralds a new opportunity to fully unleash the potential of HCF in telecommunication applications. Countering the optical network ‘capacity crunch’ requires developments in optical fibres. Here, the authors report a hollow-core fibre with conjoined tubes in the cladding and a negative-curvature core shape. It exhibits a transmission loss of 2 dB/km at 1512 nm and less than 16 dB/km bandwidth in the 1302–1637 nm range.

We demonstrate experimentally the feasibility of continuous variable quantum key distribution (CV-QKD) in dense-wavelength-division multiplexing networks (DWDM), where QKD will typically have to coexist with several co-propagating (forward or backward) C-band classical channels whose launch power is around 0 dBm. We have conducted experimental tests of the coexistence of CV-QKD multiplexed with an intense classical channel, for different input powers and different DWDM wavelengths. Over a 25 km fiber, a CV-QKD operated over the 1530.12 nm channel can tolerate the noise arising from up to 11.5 dBm classical channel at 1550.12 nm in the forward direction (9.7 dBm in backward). A positive key rate (0.49 kbits s−1) can be obtained at 75 km with classical channel power of respectively −3 and −9 dBm in forward and backward. Based on these measurements, we have also simulated the excess noise and optimized channel allocation for the integration of CV-QKD in some access networks. We have, for example, shown that CV-QKD could coexist with five pairs of channels (with nominal input powers: 2 dBm forward and 1 dBm backward) over a 25 km WDM-PON network. The obtained results demonstrate the outstanding capacity of CV-QKD to coexist with classical signals of realistic intensity in optical networks.