Photon Acceleration Research Articles

Mutual advancement: How integrated circuits interact with artificial intelligence

Abstract. Integrated circuits and artificial intelligence complement each other. On one hand, the continuous optimization of electron and photon accelerators can reduce the energy consumption of artificial intelligence algorithms and adapt them to edge computing, thereby promoting the development of artificial intelligence. On the other hand, artificial intelligence not only assists in troubleshooting and optimizing of integrated circuits, but also automates the design of integrated circuits, making them more convenient and efficient. The interactive relationship between integrated circuits and artificial intelligence is presented through qualitative analysis and literature review in this paper. The mutual influence of these two technologies creates new breakthroughs and possibilities for the development of science and technology. However, while photonic artificial intelligence accelerators are suitable for high-speed neural networks, their current development is not yet mature due to the high cost, complex manufacturing process, and environmental disturbances. Additionally, artificial intelligence algorithms for integrated circuits design lack sufficient high-quality data for training due to the secrecy of integrated circuit parameters. Thus it requires a long time to train and perfect these algorithms.

Energy response of CaSO4:Dy thermoluminescence detector to medical linear accelerator photon beams.

The study aims to evaluate the energy response of the thermoluminescent dosemeter (TLD) CaSO4:Dy from high energy photon beams produced from medical linear accelerator. The test was performed on the polymethyl methacrylate phantom surface and at the depth of dose maximum with a source-to-surface distance of 100cm and the radiation field size of 12×12cm. The results were compared with the TLD response exposed to 60Co standard source. The results show that the TLD response to the linac's photon differs >30% compared to the TLD response to 60Co exposure when measured at the phantom surface, while at the depth of maximum dose the response is similar (< ±10%), indicating that the TLD response is no longer dependent on the energy. This implies that the current dose reading evaluation based on 60Co calibration curve should be corrected to obtain more accurate dose report of the radiation workers.

Photonic neuromorphic accelerators for event-based imaging flow cytometry

In this work, we present experimental results of a high-speed label-free imaging cytometry system that seamlessly merges the high-capturing rate and data sparsity of an event-based CMOS camera with lightweight photonic neuromorphic processing. This combination offers high classification accuracy and a massive reduction in the number of trainable parameters of the digital machine-learning back-end. The event-based camera is capable of capturing 1 Gevents/sec, where events correspond to pixel contrast changes, similar to the retina’s ganglion cell function. The photonic neuromorphic accelerator is based on a hardware-friendly passive optical spectrum slicing technique that is able to extract meaningful features from the generated spike-trains using a purely analogue version of the convolutional operation. The experimental scenario comprises the discrimination of artificial polymethyl methacrylate calibrated beads, having different diameters, flowing at a mean speed of 0.1 m/sec. Classification accuracy, using only lightweight digital machine-learning schemes has topped at 98.2%. On the other hand, by experimentally pre-processing the raw spike data through the proposed photonic neuromorphic spectrum slicer at a rate of 3 × 106 images per second, we achieved an accuracy of 98.6%. This performance was accompanied by a reduction in the number of trainable parameters at the classification back-end by a factor ranging from 8 to 22, depending on the configuration of the digital neural network. These results confirm that neuromorphic sensing and neuromorphic computing can be efficiently merged to a unified bio-inspired system, offering a holistic enhancement in emerging bio-imaging applications.

Photonic floating point multiplication using cascaded SSB-SC modulation.

In this paper, we present what we believe to be a novel approach to floating-point multiplication, demonstrated experimentally for the first time. This method involves encoding floating-point numbers onto RF sub-carriers, utilizing amplitude to represent the significand and sub-carrier frequency for the exponent. We employ single-sideband suppressed-carrier (SSB-SC) modulation via IQ modulators to effectively translate the floating-point numbers into the optical domain. The process involves cascaded SSB-SC modulation coupled with balanced detection, enabling the execution of scalar floating-point multiplication. In our proof-of-concept experiment, we analyzed 10 samples with subcarrier frequencies ranging between 8 GHz and 18 GHz. The results exhibit a remarkably low error in scalar multiplication-ranging from 1% to less than 10% in the significand while maintaining error-free performance in the exponent calculation. We further conducted an energy efficiency analysis comparing fixed-point and floating-point operations for matrix-vector multiplication, demonstrating that floating-point is notably more energy-efficient, particularly for large-sized matrices or vectors. These results highlight the technique's viability for high dynamic range floating-point multiplication within photonic accelerators.

Integrated photonic fractional convolution accelerator

An integrated photonic circuit architecture to perform a modified-convolution operation based on the discrete fractional Fourier transform (DFrFT) is introduced. This is accomplished by utilizing two nonuniformly coupled waveguide lattices with equally spaced eigenmode spectra, the lengths of which are chosen so that the DFrFT and its inverse operations are achieved. A programmable modulator array is interlaced so that the required fractional convolution operation is performed. Numerical simulations demonstrate that the proposed architecture can effectively perform smoothing and edge detection tasks even for noisy input signals, which is further verified by electromagnetic wave simulations. Notably, mild lattice defects do not jeopardize the architecture performance, showing its resilience to manufacturing errors.

Partial coherence enhances parallelized photonic computing

Advancements in optical coherence control1–5 have unlocked many cutting-edge applications, including long-haul communication, light detection and ranging (LiDAR) and optical coherence tomography6–8. Prevailing wisdom suggests that using more coherent light sources leads to enhanced system performance and device functionalities9–11. Our study introduces a photonic convolutional processing system that takes advantage of partially coherent light to boost computing parallelism without substantially sacrificing accuracy, potentially enabling larger-size photonic tensor cores. The reduction of the degree of coherence optimizes bandwidth use in the photonic convolutional processing system. This breakthrough challenges the traditional belief that coherence is essential or even advantageous in integrated photonic accelerators, thereby enabling the use of light sources with less rigorous feedback control and thermal-management requirements for high-throughput photonic computing. Here we demonstrate such a system in two photonic platforms for computing applications: a photonic tensor core using phase-change-material photonic memories that delivers parallel convolution operations to classify the gaits of ten patients with Parkinson’s disease with 92.2% accuracy (92.7% theoretically) and a silicon photonic tensor core with embedded electro-absorption modulators (EAMs) to facilitate 0.108 tera operations per second (TOPS) convolutional processing for classifying the Modified National Institute of Standards and Technology (MNIST) handwritten digits dataset with 92.4% accuracy (95.0% theoretically).

Symmetric silicon microring resonator optical crossbar array for accelerated inference and training in deep learning

Photonic integrated circuits are emerging as a promising platform for accelerating matrix multiplications in deep learning, leveraging the inherent parallel nature of light. Although various schemes have been proposed and demonstrated to realize such photonic matrix accelerators, the in situ training of artificial neural networks using photonic accelerators remains challenging due to the difficulty of direct on-chip backpropagation on a photonic chip. In this work, we propose a silicon microring resonator (MRR) optical crossbar array with a symmetric structure that allows for simple on-chip backpropagation, potentially enabling the acceleration of both the inference and training phases of deep learning. We demonstrate a 4×4 circuit on a Si-on-insulator platform and use it to perform inference tasks of a simple neural network for classifying iris flowers, achieving a classification accuracy of 93.3%. Subsequently, we train the neural network using simulated on-chip backpropagation and achieve an accuracy of 91.1% in the same inference task after training. Furthermore, we simulate a convolutional neural network for handwritten digit recognition, using a 9×9 MRR crossbar array to perform the convolution operations. This work contributes to the realization of compact and energy-efficient photonic accelerators for deep learning.

Photonics for Neuromorphic Computing: Fundamentals, Devices, and Opportunities.

In the dynamic landscape of Artificial Intelligence (AI), two notable phenomena are becoming predominant: the exponential growth of large AI model sizes and the explosion of massive amount of data. Meanwhile, scientific research such as quantum computing and protein synthesis increasingly demand higher computing capacities. As the Moore's Law approaches its terminus, there is an urgent need for alternative computing paradigms that satisfy this growing computing demand and break through the barrier of the von Neumann model. Neuromorphic computing, inspired by the mechanism and functionality of human brains, uses physical artificial neurons to do computations and is drawing widespread attention. This review studies the expansion of optoelectronic devices on photonic integration platforms that has led to significant growth in photonic computing, where photonic integrated circuits (PICs) have enabled ultrafast artificial neural networks (ANN) with sub-nanosecond latencies, low heat dissipation, and high parallelism. In particular, various technologies and devices employed in neuromorphic photonic AI accelerators, spanning from traditional optics to PCSEL lasers are examined. Lastly, it is recognized that existing neuromorphic technologies encounter obstacles in meeting the peta-level computing speed and energy efficiency threshold, and potential approaches in new devices, fabrication, materials, and integration to drive innovation are also explored. As the current challenges and barriers in cost, scalability, footprint, and computing capacity are resolved one-by-one, photonic neuromorphic systems are bound to co-exist with, if not replace, conventional electronic computers and transform the landscape of AI and scientific computing in the foreseeablefuture.

A review of emerging trends in photonic deep learning accelerators

Deep learning has revolutionized many sectors of industry and daily life, but as application scale increases, performing training and inference with large models on massive datasets is increasingly unsustainable on existing hardware. Highly parallelized hardware like Graphics Processing Units (GPUs) are now widely used to improve speed over conventional Central Processing Units (CPUs). However, Complementary Metal-oxide Semiconductor (CMOS) devices suffer from fundamental limitations relying on metallic interconnects which impose inherent constraints on bandwidth, latency, and energy efficiency. Indeed, by 2026, the projected global electricity consumption of data centers fueled by CMOS chips is expected to increase by an amount equivalent to the annual usage of an additional European country. Silicon Photonics (SiPh) devices are emerging as a promising energy-efficient CMOS-compatible alternative to electronic deep learning accelerators, using light to compute as well as communicate. In this review, we examine the prospects of photonic computing as an emerging solution for acceleration in deep learning applications. We present an overview of the photonic computing landscape, then focus in detail on SiPh integrated circuit (PIC) accelerators designed for different neural network models and applications deep learning. We categorize different devices based on their use cases and operating principles to assess relative strengths, present open challenges, and identify new directions for further research.

Magnetization reversal by multiple optical pulses for a photonic spiking neuron with the leaky integrate and fire model

Photonic accelerators are anticipated to be the next generation of hardware processors, replacing traditional electronic accelerators. In current photonic accelerators based on artificial neural networks, photonic integrated circuits are incorporated with electronic integrated circuits to leverage their strengths: photonic circuits are used to perform linear calculations, while electronic circuits are used to perform nonlinear calculations. However, this architecture requires optoelectric conversion at each layer and is unable to leverage the superiority of light. We propose a novel photonic spiking neuron with a magneto-optical synapse and an all-optical spiking neural network. This study experimentally demonstrates that the magnetization reversal of CoFeB, which occurs during thermal accumulation owing to multiple optical pulses, is similar to the behavior of the leaky integrated and fire model of spiking neurons.

Lithium tantalate photonic integrated circuits for volume manufacturing

Electro-optical photonic integrated circuits (PICs) based on lithium niobate (LiNbO3) have demonstrated the vast capabilities of materials with a high Pockels coefficient1,2. They enable linear and high-speed modulators operating at complementary metal–oxide–semiconductor voltage levels3 to be used in applications including data-centre communications4, high-performance computing and photonic accelerators for AI5. However, industrial use of this technology is hindered by the high cost per wafer and the limited wafer size. The high cost results from the lack of existing high-volume applications in other domains of the sort that accelerated the adoption of silicon-on-insulator (SOI) photonics, which was driven by vast investment in microelectronics. Here we report low-loss PICs made of lithium tantalate (LiTaO3), a material that has already been adopted commercially for 5G radiofrequency filters6 and therefore enables scalable manufacturing at low cost, and it has equal, and in some cases superior, properties to LiNbO3. We show that LiTaO3 can be etched to create low-loss (5.6 dB m−1) PICs using a deep ultraviolet (DUV) stepper-based manufacturing process7. We demonstrate a LiTaO3 Mach–Zehnder modulator (MZM) with a half-wave voltage–length product of 1.9 V cm and an electro-optic bandwidth of up to 40 GHz. In comparison with LiNbO3, LiTaO3 exhibits a much lower birefringence, enabling high-density circuits and broadband operation over all telecommunication bands. Moreover, the platform supports the generation of soliton microcombs. Our work paves the way for the scalable manufacture of low-cost and large-volume next-generation electro-optical PICs.

Photonic convolution accelerator based on a hybrid integrated multi-wavelength laser array by photonic wire bonding for real-time image classification.

We propose and experimentally demonstrate a compact and efficient photonic convolution accelerator based on a hybrid integrated multi-wavelength DFB laser array by photonic wire bonding. The photonic convolution accelerator operates at 60.12 GOPS for one 3 × 3 kernel with a convolution window vertical sliding stride of 1 and generates 500 images of real-time image classification. Furthermore, real-time image classification on the MNIST database of handwritten digits with a prediction accuracy of 93.86% is achieved. This work provides a novel, to the best of our knowledge, compact hybrid integration platform to realize the optical convolutional neural networks.

Silicon photonic neuromorphic accelerator using integrated coherent transmit-receive optical sub-assemblies

Neural networks, having achieved breakthroughs in many applications, require extensive convolutions and matrix-vector multiplication operations. To accelerate these operations, benefiting from power efficiency, low latency, large bandwidth, massive parallelism, and CMOS compatibility, silicon photonic neural networks have been proposed as a promising solution. In this study, we propose a scalable architecture based on a silicon photonic integrated circuit and optical frequency combs to offer high computing speed and power efficiency. A proof-of-concept silicon photonics neuromorphic accelerator based on integrated coherent transmit–receive optical sub-assemblies, operating over 1TOPS with only one computing cell, is experimentally demonstrated. We apply it to process fully connected and convolutional neural networks, achieving a competitive inference accuracy of up to 96.67% in handwritten digit recognition compared to its electronic counterpart. By leveraging optical frequency combs, the approach’s computing speed is possibly scalable with the square of the cell number to realize over 1 Peta-Op/s. This scalability opens possibilities for applications such as autonomous vehicles, real-time video processing, and other high-performance computing tasks.

Asymmetric leader-laggard cluster synchronization for collective decision-making with laser network

Photonic accelerators have recently attracted soaring interest, harnessing the ultimate nature of light for information processing. Collective decision-making with a laser network, employing the chaotic and synchronous dynamics of optically interconnected lasers to address the competitive multi-armed bandit (CMAB) problem, is a highly compelling approach due to its scalability and experimental feasibility. We investigated essential network structures for collective decision-making through quantitative stability analysis. Moreover, we demonstrated the asymmetric preferences of players in the CMAB problem, extending its functionality to more practical applications. Our study highlights the capability and significance of machine learning built upon chaotic lasers and photonic devices.

The Descartes Code (Spin Orbital Rotation of Photons)–IV. The Harress-Sagnac Color Excess in the Rotation Curves of Galaxies

In 1933, Zwicky formulated the significant bifurcation point in the history of astrophysics. Zwicky postulated an unknown dark matter in order to explain the color excess in the observed Doppler effect in the rotation curves of galaxies. However, until now we have not detected these dark matter particles. In this contribution, we propose to come back to this bifurcation point and to interpret the observed color excess as the Harress-Sagnac effect known in the rotating interferometers. We have described photon properties in those rotating interferometers based on the old color theory of Descartes—rotating light globules (spin-orbital rotation of photons). Therefore, the color excess observed in the rotation curves of galaxies can be interpreted as the additional color excess of photons emitting from distant rotating stars. Formulae for these color excesses are given for different locations of the rotating galaxy inspired by Zwicky’s core-shell model of galaxies. A part of Zwicky’s enigma is an unknown influence of the Solar gravitational field at 1 AU on the distant photon properties. From the forgotten Gerber’s retarded potential formula we have derived an expression for the acceleration of distant photons towards the Sun at 1 AU as ax = 1.171 × 10⁻¹⁰ ms⁻². This is a very surprising coincidence with the Milgrom empirical value a0 = (1.2 ± 0.1) × 10⁻¹⁰ ms⁻². Therefore, we propose to test this new formula at the surface of Mercury and Mars in order to reveal if our gravitational models are universal or “geocentric”, valid for the surroundings of the Earth only.

Solving multi-armed bandit problems using a chaotic microresonator comb

The Multi-Armed Bandit (MAB) problem, foundational to reinforcement learning-based decision-making, addresses the challenge of maximizing rewards amid multiple uncertain choices. While algorithmic solutions are effective, their computational efficiency diminishes with increasing problem complexity. Photonic accelerators, leveraging temporal and spatial-temporal chaos, have emerged as promising alternatives. However, despite these advancements, current approaches either compromise computation speed or amplify system complexity. In this paper, we introduce a chaotic microresonator frequency comb (chaotic comb) to tackle the MAB problem, where each comb mode is assigned to a slot machine. Through a proof-of-concept experiment, we employ 44 comb modes to address an MAB with 44 slot machines, demonstrating performance competitive with both conventional software algorithms and other photonic methods. Furthermore, the scalability of decision making is explored with up to 512 slot machines using experimentally obtained temporal chaos in different time slots. Power-law scalability is achieved with an exponent of 0.96, outperforming conventional software-based algorithms. Moreover, we find that a numerically calculated chaotic comb accurately reproduces experimental results, paving the way for discussions on strategies to increase the number of slot machines.

Conflict-free joint decision by lag and zero-lag synchronization in laser network

With the end of Moore’s Law and the increasing demand for computing, photonic accelerators are garnering considerable attention. This is due to the physical characteristics of light, such as high bandwidth and multiplicity, and the various synchronization phenomena that emerge in the realm of laser physics. These factors come into play as computer performance approaches its limits. In this study, we explore the application of a laser network, acting as a photonic accelerator, to the competitive multi-armed bandit problem. In this context, conflict avoidance is key to maximizing environmental rewards. We experimentally demonstrate cooperative decision-making using zero-lag and lag synchronization within a network of four semiconductor lasers. Lag synchronization of chaos realizes effective decision-making and zero-lag synchronization is responsible for the realization of the collision avoidance function. We experimentally verified a low collision rate and high reward in a fundamental 2-player, 2-slot scenario, and showed the scalability of this system. This system architecture opens up new possibilities for intelligent functionalities in laser dynamics.

Dephasingless plasma wakefield photon acceleration.

Sandberg and Thomas [Phys. Rev. Lett. 130, 085001 (2023)0031-900710.1103/PhysRevLett.130.085001] proposed a scheme to generate ultrashort, high-energy pulses of XUV photons though dephasingless photon acceleration in a beam-driven plasma wakefield. An ultrashort laser pulse is placed in the plasma wake behind a relativistic electron bunch such that it experiences a comoving negative density gradient and therefore shifts up in frequency. Using a tapered density profile provides phase-matching between driver and witness pulses. In this paper, we give the details of the wakefield solutions and phase-matching conditions used to generate the phase-matching density profile. The short, high-density, and weak driver limits are considered. We show, explicitly, the numerical algorithm used to calculate the density profiles.

Comparison of the standards for absorbed dose to water of the IAEA and the KRISS, Korea in accelerator photon beams

A bilateral comparison was conducted between the International Atomic Energy Agency (IAEA) and the Korea Research Institute of Standards and Science (KRISS) to measure the absorbed dose to water in accelerator photon beams. KRISS served as a linking laboratory to compare the IAEA standard with the key comparison reference value (KCRV) of the BIPM.RI(I)–K6 program, in which KRISS participated in 2017.Two ionization chambers from the IAEA were used as transfer instruments for the comparison. Both laboratories measured the calibration coefficients of these instruments and calculated the ratios. The ratio of the KRISS standard to the KCRV was applied to obtain the degree of equivalence of the IAEA, along with its uncertainty. The largest deviation of the IAEA measurement from the KCRV was 3.4 mGy/Gy, significantly smaller than the expanded uncertainty of 10.7 mGy/Gy (k = 2, 95% level of confidence).This study demonstrates the equivalence of IAEA's measurement standard for accelerator photon beams to other primary standard dosimetry laboratories. It provides evidence for the satisfactory operation of IAEA's quality management system and enhances the international credibility of the IAEA SSDL network, particularly in high-energy accelerator photon beams from linear accelerators.

Photon Acceleration by Superluminal Ionization Fronts

This paper explores the use of superluminal ionization fronts to accelerate and amplify electromagnetic radiation. These fronts are defined as optical boundaries between two regions of a gas, the neutral region and the plasma region, characterized by two different values of the refractive index. For that reason, the front velocity is not necessarily related to the motion of material particles, such as neutral atoms, ions and electrons, which can stay at rest. The fronts can therefore become superluminal without violating causality. In recent years, different experimental configurations, such as the flying focus, showed that it is possible to create superluminal fronts in the laboratory. These fronts can easily be described theoretically in a special reference frame, called the time frame, which is used here. In this frame, superluminal fronts reduce to time refraction, a process that is symmetrical to the well-known optical refraction. It is shown that propagation through such fronts can lead to considerable frequency shifts and energy amplification of probe laser beams. This could eventually be used to develop new sources of tunable radiation.