Neural networks have enabled applications in artificial intelligence through machine learning, and neuromorphic computing. Software implementations of neural networks on conventional computers that have separate memory and processor (and that operate
sequentially) are limited in speed and energy efficiency. Neuromorphic engineering aims to build processors in which hardware mimics neurons and synapses in the brain for distributed and parallel processing. Neuromorphic engineering enabled by photonics (optical physics) can offer sub-nanosecond latencies and high bandwidth with low energies to extend the domain of artificial intelligence and neuromorphic computing applications to machine learning acceleration, nonlinear programming, intelligent signal processing, etc. Photonic neural networks have been demonstrated on integrated platforms and free-space optics depending on the class of applications being targeted. Here, we discuss the prospects and demonstrated applications of these photonic neural networks.
In the 1960s, computer engineers had to address the tyranny of numbers problem in which improvements in computing and its applications required integrating an increasing number of electronic components. From the first computers powered by vacuum tube
s to the billions of transistors fabricated on a single microprocessor chip today, transformational advances in integration have led to remarkable processing performance and new unforeseen applications in computing. Today, quantum scientists and engineers are facing similar integration challenges. Research labs packed with benchtop components, such as tunable lasers, tables filled with optics, and racks of control hardware, are needed to prepare, manipulate, and read out quantum states from a modest number of qubits. Analogous to electronic circuit design and fabrication nearly five decades ago, scaling quantum systems (i.e. to thousands or millions of components and quantum elements) with the required functionality, high performance, and stability will only be realized through novel design architectures and fabrication techniques that enable the chip-scale integration of electronic and quantum photonic integrated circuits (QPIC). In the next decade, with sustained research, development, and investment in the quantum photonic ecosystem (i.e. PIC-based platforms, devices and circuits, fabrication and integration processes, packaging, and testing and benchmarking), we will witness the transition from single- and few-function prototypes to the large-scale integration of multi-functional and reconfigurable QPICs that will define how information is processed, stored, transmitted, and utilized for quantum computing, communications, metrology, and sensing. This roadmap highlights the current progress in the field of integrated quantum photonics, future challenges, and advances in science and technology needed to meet these challenges.
Low latency, high throughput inference on Convolution Neural Networks (CNNs) remains a challenge, especially for applications requiring large input or large kernel sizes. 4F optics provides a solution to accelerate CNNs by converting convolutions int
o Fourier-domain point-wise multiplications that are computationally free in optical domain. However, existing 4F CNN systems suffer from the all-positive sensor readout issue which makes the implementation of a multi-channel, multi-layer CNN not scalable or even impractical. In this paper we propose a simple channel tiling scheme for 4F CNN systems that utilizes the high resolution of 4F system to perform channel summation inherently in optical domain before sensor detection, so the outputs of different channels can be correctly accumulated. Compared to state of the art, channel tiling gives similar accuracy, significantly better robustness to sensing quantization (33% improvement in required sensing precision) error and noise (10dB reduction in tolerable sensing noise), 0.5X total filters required, 10-50X+ throughput improvement and as much as 3X reduction in required output camera resolution/bandwidth. Not requiring any additional optical hardware, the proposed channel tiling approach addresses an important throughput and precision bottleneck of high-speed, massively-parallel optical 4F computing systems.
Photodetectors are key optoelectronic building blocks performing the essential optical-to-electrical signal conversion, and unlike solar cells, operate at a specific wavelength and at high signal or sensory speeds. Towards achieving high detector per
formance, device physics, however, places a fundamental limit of the achievable detector sensitivity, such as responsivity and gain, when simultaneously aimed to increasing the detectors temporal response, speed, known as the gain-bandwidth product (GBP). While detectors GBP has been increasing in recent years, the average GBP is still relatively modest (~10^6-10^7 Hz-A/W). Here we discuss photodetector performance limits and opportunities based on arguments from scaling length theory relating photocarrier channel length, mobility, electrical resistance with optical waveguide mode constrains. We show that short-channel detectors are synergistic with slot-waveguide approaches, and when combined, offer a high-degree of detector design synergy especially for the class of nanometer-thin materials. Indeed, we find that two dimensional material-based detectors are not limited by their low mobility and can, in principle, allow for 100 GHz fast response rates. However, contact resistance is still a challenge for such thin materials, a research topic that is still not addressed yet. An interim solution is to utilize heterojunction approaches for functionality separation. Nonetheless, atomistically- and nanometer-thin materials used in such next-generation scaling length theory based detectors also demand high material quality and monolithic integration strategies into photonic circuits including foundry-near processes. As it stands, this letter aims to guide the community if achieving the next generation photodetectors aiming for a performance target of GBP = 10^12 Hz-A/W.
