A5084150135- Photonic and Optical Devices
- Neural Networks and Reservoir Computing
- Advanced Memory and Neural Computing
- Optical Network Technologies
- Advanced Photonic Communication Systems
- Semiconductor Lasers and Optical Devices
- Semiconductor Quantum Structures and Devices
- Advanced Optical Sensing Technologies
- Photoreceptor and optogenetics research
- Advanced Fiber Laser Technologies
- Photonic Crystals and Applications
- Machine Learning and ELM
- Cellular Automata and Applications
- Optical Coherence Tomography Applications
- Neural Networks and Applications
- Thin-Film Transistor Technologies
- Photonic Crystal and Fiber Optics
- Neural dynamics and brain function
- Semiconductor materials and devices
- Spectroscopy and Laser Applications
- Photorefractive and Nonlinear Optics
Hewlett Packard Enterprise (United States)
2018-2025
Hewlett-Packard (United States)
2022-2024
Alibaba Group (United States)
2022
Hewlett Packard Enterprise (Ireland)
2020
University of Virginia
2017-2019
California Polytechnic State University
2014-2015
Abstract Recently, interest in programmable photonics integrated circuits has grown as a potential hardware framework for deep neural networks, quantum computing, and field arrays (FPGAs). However, these are constrained by the limited tuning speed large power consumption of phase shifters used. In this paper, we introduce memresonator, metal-oxide memristor heterogeneously with microring resonator, non-volatile silicon photonic shifter. These devices capable retention times 12 hours,...
Heterogeneous III-V-on-silicon photonic integration has proved to be an attractive and volume manufacturable solution that marries the merits of III-V compounds silicon technology for various integrated circuit (PIC) applications. The current main-stream Ethernet trends larger bandwidth are pushing higher modulation baudrate or employing advanced format datacom However, neither is likely able significantly drive overall cost energy efficiency best sweet spot, nor unfold full potential...
Programmable photonics play a crucial role in many emerging applications, from optical accelerators for machine learning to quantum information technologies. Conventionally, photonic systems are tuned by mechanisms such as the thermo-optic effect, free carrier dispersion, electro-optic or micro-mechanical movement. Although these physical effects allow either fast (>100 GHz) large contrast (>60 dB) switching, their high static power consumption is not optimal programmability,...
The convergence of deep learning and big data has spurred significant interest in developing novel hardware that can run large artificial intelligence (AI) workloads more efficiently. Over the last several years, silicon photonics emerged as a disruptive technology for next-generation accelerators machine (ML). More recently, heterogeneous integration III-V compound semiconductors opened door to integrating lasers semiconductor optical amplifiers at wafer-scale, enabling scaling size,...
<title>Abstract</title> The need for high-speed, energy-efficient computing in machine learning and real-time communication necessitates innovations beyond conventional digital analog electronics to sustain computational power advances without requiring prohibitive energy amounts. Photonics has emerged as a promising platform demonstrating significant highlights the field of linear transformations. Adopting, however, use photons within broad range applications their successful employment...
Silicon photonics provides a promising platform for energy-efficient interconnects within supercomputers and data centers. However, developing complementary metal–oxide–semiconductor compatible high-speed photodetector with low dark current has long presented challenge in the field. In this paper, we report first O-band InAs quantum dot (QD) waveguide photodiode (PD) heterogeneously integrated on silicon. Record currents as 0.01 nA, responsivities of 0.34 A/W at 1310 nm 0.9 A/W 1280 nm,...
Abstract Silicon photonics technology has drawn significant interest due to its potential for compact and high-performance photonic integrated circuits. The Ge- or III–V material-based avalanche photodiodes on silicon provide ideal high sensitivity optical receivers telecommunication wavelengths. Herein, the last advances of monolithic heterogeneous are reviewed, including different device structures semiconductor systems.
Abstract Over the past few years, extensive work on optical neural networks has been investigated in hopes of achieving orders magnitude improvement energy efficiency and compute density via all-optical matrix-vector multiplication. However, these solutions are limited by a lack high-speed power power-efficient phase tuners, on-chip non-volatile memory, proper material platform that can heterogeneously integrate all necessary components needed onto single chip. We address issues...
We will review our most recent results on non-volatile memristive III-V/Si Photonics. The wafer-bonded cell facilitates optical functionality for a variety of devices such as Mach-Zehnder Interferometers (MZIs), asymmetric MZI lattice filters, and ring resonator filters.
Abstract Non‐volatile charge‐trap flash memory (CTM) co‐located with heterogeneous III‐V/Si photonics is demonstrated. The wafer‐bonded CTM cell facilitates non‐volatile optical functionality for a variety of devices such as Mach–Zehnder Interferometers (MZIs), asymmetric MZI lattice filters, and ring resonator filters. exhibits full write/erase operation (100 cycles 500 states) wavelength shifts Δλ = 1.16 nm (Δn eff,non‐volatile ≈ 2.5 × 10 −4 ) dynamic power consumption <20 pW (limited...
Kolmogorov-Arnold Networks (KAN) models were recently proposed and claimed to provide improved parameter scaling interpretability compared conventional multilayer perceptron (MLP) models. Inspired by the KAN architecture, we propose Photonic -- an integrated all-optical neuromorphic platform leveraging highly parametric optical nonlinear transfer functions along edges. In this work, implement such nonlinearities in form of cascaded ring-assisted Mach-Zehnder Interferometer (MZI) devices....
We demonstrate a heterogeneously integrated optical III-V/Si MZI memristor for the purposes of non-volatile memory. The results indicate possibility implementing large scale photonics and neural networks with inherent memory functions.
Waveguide-integrated photodiodes with InGaAs/ GaAsSb type-II quantum well absorption regions designed to absorb light at 2 μm are presented. A novel dual-integrated waveguide-depletion layer was used maximize efficiency in thin absorbers for high-speed optical response. Low dark currents (1 nA -1 V) and an internal responsivity of 0.84 A/W along a bandwidth above 10 GHz open eye diagram Gb/s have been demonstrated μm. The carrier dynamics within InGaAs/GaAsSb wells explored the first time...
We report a heterogeneous GaAs-based quantum dot (QD) avalanche photodiode (APD) on silicon with an ultralow dark current of 10 pA at <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"> <mml:mrow class="MJX-TeXAtom-ORD"> <mml:mo>−</mml:mo> </mml:mrow> <mml:mn>1</mml:mn> <mml:mspace width="thickmathspace"/> <mml:mi mathvariant="normal">V</mml:mi> </mml:math> , 3 dB bandwidth 20 GHz and record gain-bandwidth product (GBP) 585 GHz. Furthermore, open eye diagrams up to 32...
Abstract Silicon photonics has evolved from lab research to commercial products in the past decade as it plays an increasingly crucial role data communication for next‐generation centers and high‐performance computing. Recently, programmable silicon also found new applications quantum classical information processing. A key component of photonic integrated circuits (PICs) is phase shifter, traditionally realized via thermo‐optic or free‐carrier effects that are weak, volatile, power hungry....
This paper introduces our recent efforts on scalable, energy-efficient, and low-latency tensorized optical neural networks, including design considerations, options for wavelength-parallel photonic tensor cores, memory non-volatile tuning.
Herein, we demonstrate the first silicon photonic cellular automaton. Results were measured using p-n junction microring resonators on a platform. Demonstrated are rules 90, 250 and 110 automata, signifying that this computing system is Turing complete can compute any program.
We present and discuss the performance characteristics of InP-based p-i-n photodiodes (PDs) with InGaAs/GaAsSb type-II multiple quantum wells absorption regions designed to absorb light at mid-infrared wavelengths. Top-illuminated waveguide-integrated PDs are fabricated dark currents as low 100 nA -2 V, an external responsivity high 0.27 A/W 2 μm 0.3 1.55 μm, a 3-dB bandwidth 3.5 GHz μm.
Here we demonstrate the first-ever semiconductor laser with non-volatile tuning functionality. A GaAs-on Silicon quantum dot is integrated an Al <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> O xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> memristor, which results in lasing mode being manipulated by memristor states. The controls microring output wavelength a manner.
We demonstrate a heterogeneous III-V/Si (de-)interleaver based on an integrated memristor for non-volatile tuning. Passband tuning via phase shifts are demonstrated with full set/reset states. Non-volatile retention times reported in 24 hour period. The results indicate the possibility of implementing large scale photonics and optical neural networks inherent memory functions.
We present memristors heterogeneously integrated on a silicon photonics plat- form. Memristor microring modulators and lasers have been demon- strated can be together to create photonic neuromorphic circuits opto-electronic memory modules.
We present a novel technology wherein memristors are heterogeneously integrated with optoelectronic devices on silicon photonic platform. results memristor microring modulators and lasers non-volatile memory. Furthermore, multiple combined optical waveguides to create circuits neuromorphic computing. By pairing memory directly photonics, we can integrate memory, computing, high-speed interconnects all together the same chip.
This paper proposes a wavelength-parallel photonic tensor core by exploiting multiple free spectral ranges (multi-FSRs) of the microring resonator crossbar array architecture. As an example, four-FSR 4×4 is designed and simulated.
We demonstrated a non-volatile III-V-on-silicon photonic phase shifter based on HfO 2 memristor with ~400 fJ switching energy, 100 ns speed, and an excellent endurance of over 800 cycles.