A5101619573- Photonic and Optical Devices
- Optical Network Technologies
- Neural Networks and Reservoir Computing
- Photoacoustic and Ultrasonic Imaging
- Phase-change materials and chalcogenides
- Advanced Fiber Optic Sensors
- Optical Coherence Tomography Applications
- Advanced Photonic Communication Systems
- Photonic Crystals and Applications
- Spectroscopy Techniques in Biomedical and Chemical Research
- Mechanical and Optical Resonators
- Advanced Fiber Laser Technologies
- Thermography and Photoacoustic Techniques
- Expert finding and Q&A systems
- Topic Modeling
- Graph theory and applications
- Advanced Neural Network Applications
- Colorectal Cancer Treatments and Studies
- Graph Labeling and Dimension Problems
- Spectroscopy and Laser Applications
- Plasma Diagnostics and Applications
- Adversarial Robustness in Machine Learning
- Photochromic and Fluorescence Chemistry
- Advanced Memory and Neural Computing
- Particle accelerators and beam dynamics
George Washington University
2019-2024
Peking University
2021-2023
Shanxi University
2022-2023
Wuhan University
2023
Xinjiang University
2023
Collaborative Innovation Center of Quantum Matter
2021
Photonic Random-Access Memories (P-RAM) are an essential component for the on-chip non-von Neumann photonic computing by eliminating optoelectronic conversion losses in data links. Emerging Phase-Change Materials (PCMs) have been showed multilevel memory capability, but demonstrations still yield relatively high optical loss and require cumbersome WRITE-ERASE approaches increasing power consumption system package challenges. Here we demonstrate a multistate electrically programmed low-loss...
Solving mathematical equations faster and more efficiently has been a Holy Grail for centuries scientists engineers across all disciplines. While electronic digital circuits have revolutionized equation solving in recent decades, it become apparent that performance gains from brute-force approaches of compute-solvers are quickly saturating over time. Instead, paradigms leverage the universes' natural tendency to minimize system's free energy, such as annealers or Ising Machines, being sought...
Photonic neural networks benefit from both the high channel capacity-and wave nature of light acting as an effective weighting mechanism through linear optics.The neuron's activation function, however, requires nonlinearity which can be achieved either nonlinear optics or electro-optics.Nonlinear optics, while potentially faster, is challenging at low optical power.With electro-optics, a photodiode integrating weighted products photonic perceptron paired directly to modulator, creates...
Liquid-crystal microcavity lasers have attracted considerable attention because of their extraordinary tunability and sensitive response to external stimuli, they operate generally within a specific phase. Here, we demonstrate liquid-crystal laser operated in the phase transition which reorientation molecules occurs from aligned disordered states. A significant wavelength shift microlaser is observed, resulting dramatic changes refractive index microdroplets during transition. This...
Abstract The retention-of-state functionality provided by memories is fundamental to any Turing machine and neural network, hence critical for information system today. While emerging optical learning accelerators photonic neuromorphic computing paradigms provide promising signal processing performance, the lack of a photon-photon force in universe makes storing challenging. Fortunately, phase change materials such missing memristive nonvolatile function via their reconfigurable crystalline...
We propose and demonstrate experimentally the strong dissipative acousto-optic interaction between a suspended vibrating microfiber whispering-gallery microcavity. On one hand, response driven by an external stimulus of acoustic waves is found to be stronger than dispersive 2 orders magnitude. other dead points emerge with zero at certain parameters, promising potentials in physical sensing such as precise measurements magnetic field temperature. The then explored for ultrasensitive...
Key to recent successes in the field of artificial intelligence (AI) has been ability train a growing number parameters which form fixed connectivity matrices between layers nonlinear nodes. This "deep learning" approach AI historically required an exponential growth processing power far exceeds computational throughput digital hardware as well trends efficiency. New computing paradigms are therefore enable efficient information while drastically improving throughput. Emerging strategies for...
Here we present a multi-level discrete-state nonvolatile photonic memory based on an ultra-compact ( ) hybrid phase change material GSST-silicon Mach Zehnder modulator, with low insertion losses (3dB), to serve as node in neural network. Emulating opportunely trained 100×100 fully connected multilayered perceptron network this weighting functionality embedded memory, shows up 92% inference accuracy and robustness towards noise when performing predictions of unseen data.
Digital-to-analog converters (DAC) are indispensable functional units in signal processing instrumentation and wide-band telecommunication links for both civil military applications. Since photonic systems capable of high data throughput low latency, an increasingly found system limitation stems from the required domain-crossing such as digital-to-analog, electronic-to-optical. A DAC implementation, contrast, enables a seamless conversion with respect to energy efficiency short delay, often...
Optical ultrasonic probes, exemplified by Fabry–Perot cavities on optical fibers, have small sizes, high sensitivity, and pure characteristics, making them highly attractive in high-resolution ultrasonic/photoacoustic imaging, especially near-field or endoscopic scenarios. Taking a different approach, we demonstrate an ultrasensitive broadband ultrasound microprobe formed whispering-gallery-mode polymer microcavity coupled to U-shaped microfiber. With the high-quality ( Q ) factors (>10 6...
Here we demonstrate a photonic tensor core based on silicon photonics dot-product engine. Utilizing compact electronic phase-change-material memory and WDM show the highest throughput density to date of 3.8 MAC/s/mm 2 .
Solving mathematical equations faster and more efficiently has been a Holy Grail for centuries scientists engineers across all disciplines. While electronic digital circuits have revolutionized equation solving in recent decades, it become apparent that performance gains from brute-force approaches of compute-solvers are quickly saturating over time. Instead, paradigms leverage the universes natural tendency to minimize systems free energy, such as annealers or Ising Machines, being sought...
Here we present a multi-level discrete-state nonvolatile photonic memory based on an ultra-compact (<4μm) hybrid phase change material GSST-silicon Mach Zehnder modulator, with low insertion losses (3dB), to serve as node in neural network. Emulating opportunely trained 100 × fully connected multilayered perceptron network this weighting functionality embedded memory, shows up 92% inference accuracy and robustness towards noise when performing predictions of unseen data.
Here we demonstrate an on-chip programmable multi-level non-volatile photonic memory based on ultra-compact (<4pm) hybrid GSST-silicon Mach Zehnder modulator, with low insertion losses (3dB), used as node in a neural network that effortlessly perform inference.
Here we demonstrate an on-chip programmable multi-level non-volatile photonic memory used as node in a neural network that effortlessly perform inference at the edge of passive and reprogrammable filter.
Digital-To-Analog Converters In article number 2000033, Volker J. Sorger and co-workers demonstrate a novel photonic digital-toanalog converter (DAC) that bypasses the usually deployed optical-to-electrical-to-optical conversions. Furthermore, this parallel binary-weighted DAC offers high sampling efficiencies compact form-factor is exceptionally suited to be at network-edges peripheral sensors. (Image designed by Dr. Mario Miscuglio)
Here we demonstrate an ultra-low-loss multi-state photonic memory with phase change materials (GeSbSe), which can be efficiently reprogrammed on-chip. 4-bit and over 100,000 cycle tests are shown for this material.
Photonic Random-Access Memories (P-RAM) are an essential component for the on-chip non-von Neumann photonic computing by eliminating optoelectronic conversion losses in data links. Emerging Phase Change Materials (PCMs) have been showed multilevel memory capability, but demonstrations still yield relatively high optical loss and require cumbersome WRITE-ERASE approaches increasing power consumption system package challenges. Here we demonstrate a multi-state electrically-programmed low-loss...