A compact, electrically driven light source integrated on silicon is a key component for large-scale integration of electronic and photonic integrated circuits. Here we demonstrate electrically injected continuous-wave lasing in InP-based microdisk lasers coupled to a sub-micron silicon wire waveguide, fabricated through heterogeneous integration of InP on silicon-on-insulator (SOI). The InP-based microdisk has a diameter of 7.5 mum and a thickness of 1 mum. A tunnel junction was incorporated to efficiently contact the p-side of the pn-junction. The laser emits at 1.6 mum, with a threshold current as low as 0.5 mA under continuous-wave operation at room temperature, and a threshold voltage of 1.65 V. The SOI-coupled laser slope efficiency was estimated to be 30 muW/mA, with a maximum unidirectional output power of 10 muW.

http://pcphotonics.intec.ugent.be/download/pub_2116.pdf

Electrically pumped InP-based microdisk lasers integrated with a nanophotonic silicon-on-insulator waveguide circuit.

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.

/pdf/neuromorphic-photonic-networks-using-silicon-photonic-weight-3o4fn6wx4m.pdf

Neuromorphic photonic networks using silicon photonic weight banks.

This paper reviews advanced optical burst switching (OBS) and optical packet switching (OPS) technologies and discusses their roles in the future photonic Internet. Discussions include optoelectronic and optical systems technologies as well as systems integration into viable network elements (OBS and OPS routers). Optical label switching (OLS) offers a unified multiple-service platform with effective and agile utilization of the available optical bandwidth in support of voice, data, and multimedia services on the Internet Protocol. In particular, OLS routers with wavelength routing switching fabrics and parallel optical labeling allow forwarding of asynchronously arriving variable-length packets, bursts, and circuits. By exploiting contention resolution in wavelength, time, and space domains, the OLS routers can achieve high throughput without resorting to a store-and-forward method associated with large buffer requirements. Testbed demonstrations employing OLS edge routers show high-performance networking in support of multimedia and data communications applications over the photonic Internet with optical packets and bursts switched directly at the optical layer

Optical Packet and Burst Switching Technologies for the Future Photonic Internet

This paper presents an analysis of optical buffers based on slow-light optical delay lines. The focus of this paper is on slow-light delay lines in which the group velocity is reduced using linear processes, including electromagnetically induced transparency (EIT), population oscillations (POs), and microresonator-based photonic-crystal (PC) filters. We also consider slow-light delay lines in which the group velocity is reduced by an adiabatic process of bandwidth compression. A framework is developed for comparing these techniques and identifying fundamental physical limitations of linear slow-light technologies. It is shown that slow-light delay lines have limited capacity and delay-bandwidth product. In principle, the group velocity in slow-light delay lines can be made to approach zero. But very slow group velocity always comes at the cost of very low bandwidth or throughput. In many applications, miniaturization of the delay line is an important consideration. For all delay-line buffers, the minimum physical size of the buffer for a given number of buffered data bits is ultimately limited by the physical size of each stored bit. We show that in slow-light optical buffers, the minimum achievable size of 1 b is approximately equal to the wavelength of light in the buffer. We also compare the capabilities and limitations of a range of delay-line buffers, investigate the impact of waveguide losses on the buffer capacity, and look at the applicability of slow-light delay lines in a number of applications.

https://web.eecs.umich.edu/~peicheng/people/publications/2005%20-%20Slow-Light%20Optical%20Buffers-%20Capabilities%20and%20Fundamental%20Limitations.pdf

Slow-light optical buffers: capabilities and fundamental limitations

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, particularly 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. Photonics offers an attractive platform for implementing neuromorphic computing due to its low latency, multiplexing capabilities and integrated on-chip technology.

Photonics for artificial intelligence and neuromorphic computing

We present a 1 /spl times/ 2 all-optical packet switch. All the processing of the header information is carried out in the optical domain. The optical headers are recognized by employing the two-pulse correlation principle in a semiconductor laser amplifier in loop optical mirror (SLALOM) configuration. The processed header information is stored in an optical flip-flop memory that is based on a symmetric configuration of two coupled lasers. The optical flip-flop memory drives a wavelength routing switch that is based on cross-gain modulation in a semiconductor optical amplifier. We also present an alternative optical packet routing concept that can be used for all-optical buffering of data packets. In this case, an optical threshold function that is based on a asymmetric configuration of two coupled lasers is used to drive a wavelength routing switch. Experimental results are presented for both the 1 /spl times/ 2 optical packet switch and the optical buffer switch.

Optical packet switching and buffering by using all-optical signal processing methods

We present a model for polarization-dependent gain saturation in strained bulk semiconductor optical amplifiers. We assume that the polarized optical field can be decomposed into transverse electric and transverse magnetic components that have indirect interaction with each other via the gain saturation. The gain anisotropy due to tensile strain in the amplifier is accounted for by a population imbalance factor. The model is applied to a nonlinear polarization switch, for which results are obtained, that are in excellent agreement with experimental data. Finally, we describe an all-optical flip-flop memory that is based on two coupled nonlinear polarization switches.

/pdf/nonlinear-polarization-rotation-in-semiconductor-optical-1357igqstq.pdf

Nonlinear polarization rotation in semiconductor optical amplifiers: theory and application to all-optical flip-flop memories

An all-optical set-reset flip-flop is presented that is based on two coupled lasers with separate cavities and lasing at different wavelengths. The lasers are coupled so that lasing in one of the lasers quenches lasing in the other laser. The flip-flop state is determined by the laser that is currently lasing. A rate-equation based model for the flip-flop is developed and used to obtain steady-state characteristics. Important properties of the system, such as the minimum coupling between lasers and the optical power required for switching, are derived from the model. These properties are primarily dependent on the laser mirror reflectivity, the inter-laser coupling, and the power emitted from one of the component lasers, affording the designer great control over the flip-flop properties. The flip-flop is experimentally demonstrated with two lasers constructed from identical semiconductor optical amplifiers (SOAs) and fiber Bragg gratings of different wavelengths. Good agreement between the theory and experiment is obtained. Furthermore, switching over a wide range of input wavelengths is shown; however, increased switching power is required for wavelengths far from the SOA gain peak.

/pdf/all-optical-flip-flop-based-on-coupled-laser-diodes-13a5q7pywz.pdf

All-optical flip-flop based on coupled laser diodes

A 1/spl times/2 all-optical packet switch is presented The header processing is implemented by using a SLALOM structure and an optical flip-flop memory is used to store the processed header bits The packets are switched in wavelength by using cross-gain modulation Experimental results are presented

/pdf/1x2-optical-packet-switch-using-all-optical-header-59w4j1jyu1.pdf

1x2 optical packet switch using all-optical header processing

We demonstrate novel optical signal processing functions based on self-induced nonlinear polarization rotation in a semiconductor optical amplifier (SOA). Numerical and experimental results are presented, which demonstrate that a nonlinear polarization switch can be employed to achieve all-optical logic. We demonstrate an all-optical header processing system, an all-optical seed pulse generator for packet synchronization, and an all-optical arbiter that can be employed for optical buffering at a bit rate of 10 Gb/s. Experimental results indicate that optical signal processing functions based on self-polarization rotation have a higher extinction ratio and a lower power operation compared with similar functions based on self-phase modulation.

/pdf/optical-signal-processing-based-on-self-induced-polarization-37pvnpzdu6.pdf

H.J.S. Dorren

Papers

Optical packet switching and buffering by using all-optical signal processing methods

Nonlinear polarization rotation in semiconductor optical amplifiers: theory and application to all-optical flip-flop memories

All-optical flip-flop based on coupled laser diodes

1x2 optical packet switch using all-optical header processing

Optical signal processing based on self-induced polarization rotation in a semiconductor optical amplifier