PhotoniX

姓名
邮箱
手机号码
标题
留言内容
验证码

期刊信息更多+

主办单位：
中国光学工程学会清华大学上海理工大学
名誉主编： 庄松林院士
国际主编： 顾敏院士
主编：
孙洪波教授仇旻教授
创刊：2020年3月
ISSN：2662-1991

新闻资讯更多+

最新上线

Sub-nanosecond heat-based logic, writing and reset in an antiferromagnetic magnetoresistive memory

M. Surýnek, A. Farkaš, J. Zubáč, P. Kubaščík, K. Olejník, F. Krizek, L. Nádvorník, T. Ostatnický, R. P. Campion, V. Novák, T. Jungwirth, P. Němec

doi: 10.1186/s43074-025-00207-1

Abstract(0) PDF(0)

Abstract:
Thermal logic aims to create thermal counterparts to electronic circuits. In this work, we investigate experimentally the response of an analog memory device based on a thin film of an antiferromagnetic metal CuMnAs to bursts of heat pulses generated by the absorption of femtosecond laser pulses at room ambient temperature. When a threshold temperature in the heat-based short-term memory of the device is exceeded, the output of the in-memory logic operations is transferred within the same device to a long-term memory, where it can be retrieved at macroscopic times. The long-term memory is based on magnetoresistive switching from a reference low-resistive uniform magnetic state to high-resistive metastable nanofragmented magnetic states. The in-memory heat-based logic operations and the conversion of the outputs into the electrically-readable long-term magnetoresistive memory were performed at sub-nanosecond time scales, making them compatible with the GHz frequencies of standard electronics. Finally, we demonstrate the possibility of rapidly resetting the long-term memory to the reference low-resistive state by heat pulses.

Photonic transformer chip: interference is all you need

Ye Tian, Shuiying Xiang, Xingxing Guo, Yahui Zhang, Jiashang Xu, Shangxuan Shi, Haowen Zhao, Yizhi Wang, Xinran Niu, Wenzhuo Liu, Yue Hao

doi: 10.1186/s43074-025-00182-7

Abstract(13) PDF(1)

Abstract:
As the core component of the transformer model, the attention has been proved as all you need in artificial intelligence field in recent years. However, conventional electronic processors are unable to cope with the exponentially increasing hardware costs and energy consumption of the computing-expensive attention. While the photonic neural network (NN) chips provide alternative energy-efficient solutions for accelerating the matrix multiplication (MM), existing photonic accelerators are primarily designed for weight-static NNs that involve MM between the learned weight matrix and input tensors and thus are inefficient in supporting attention mechanisms that require dynamic input operands. Here we propose an attention mechanism relying solely on the runtime-programable optical-interference. Through theoretical analyses, numerical simulations and experimental validations, we demonstrate the photonic “all-interference” attention with learning capability equivalent to classical self-attention, and implement the photonic transformer chip (PTC). Evaluation shows that the PTC is promising to exceed 200 pera-operations per second (POPS) with 1POPS/mm² computation density and 0.5 POPS/W power efficiency, much better than prior photonic accelerators, and delivers over 200 × energy reduction and 2 to 3 orders of magnitude higher computation capability compared to the electronic counterpart. The photonic transformer with “all-interference” attention proposed in this work highlights the immense potential of photonics to construct its own computing paradigm for general purpose machine learning.

Hybrid high-index composite meta-structures with atomic layer-coated nanoparticle-embedded resin

Minseok Choi, Hyunjung Kang, Dohyun Kang, Joohoon Kim, Hongyoon Kim, Junhwa Seong, Seokwoo Kim, Junsuk Rho

doi: 10.1186/s43074-025-00204-4

Abstract(12) PDF(0)

Abstract:
Metasurfaces offer great potential to replace conventional optics by enabling multi-functionalities in compact form factors. However, their mass production remains at crossroads, as most materials compatible with scalable fabrication like nanoimprint lithography (NIL) exhibit relatively low refractive indices (~ 2), which limit metasurface performance and necessitate tall, high-aspect-ratio meta-atoms prone to bending and collapsing. To address these bottlenecks, we introduce a hybrid nanoparticle-embedded resin (nano-PER) structure that reduces meta-atom height and aspect ratio. By utilizing TiO₂ nano-PER as the core material with thin TiO₂ coatings, we can implement the optical properties of high refractive index with printable material, achieving a height reduction of over 27% and an aspect ratio reduction of more than 36% compared with conventional hybrid structures using nanoimprint resin. Despite the reduced dimensions, our meta-atoms exhibit high broadband properties, with an average conversion efficiency of over 72% across blue (450 nm), green (532 nm), and red (635 nm) wavelengths. Our design provides robustness in the fabrication process, demonstrated by producing a hyperbolic metalens via NIL and experimentally verifying its optical performance, with an average focusing efficiency of 51.23%. These findings mark an important advancement in scalable, high-performance metasurfaces, paving the way for their practical integration into optical applications.

Modulation of ultrafast soliton molecules enabled by plasmonic metafibers

Chenxi Zhang, Zuxi Ouyang, Yunyu Lyu, Bo Fu, Lei Zhang, Min Qiu

doi: 10.1186/s43074-025-00205-3

Abstract(10) PDF(0)

Abstract:
Soliton molecules (SMs) are bound states of two or more fundamental solitons arising from the balance between nonlinear interactions and dispersion effects. SMs are the focus of intense research and have sparked numerous applications in optical communication, including coding, storage, and exchange. However, it remains challenging to experimentally produce SMs with the required pulse number and temporal separation in mode-locked fiber lasers owning to the gap between the theoretical prediction and the experimental results. Here, we achieve controllable output of SMs by utilizing a plasmonic metafiber and external manipulation techniques. Plasmonic metafibers with superior nonlinear performance are used as saturable absorbers to realize soliton mode-locking operation with femtosecond pulse duration. Regulation of pump power and polarization enables switching of pulse number from one to five and temporal separation from 4 ps to 10 ps, respectively. An analytical model based on the nonlinear Schrödinger equation is well established, effectively bridging the gap between experimental results and theoretical predictions. Our results shed light on the understanding of the formation mechanism, transport properties, and free regulation of SMs.

引用排行

Terahertz spectroscopy in biomedical field: a review on signal-to-noise ratio improvement

Yan Peng, Chenjun Shi, Yiming Zhu, Min Gu, Songlin Zhuang

2020, 1(1). doi: 10.1186/s43074-020-00011-z

PDF(41)

Recent advances in multi-dimensional metasurfaces holographic technologies

Ruizhe Zhao, Lingling Huang, Yongtian Wang

2020, 1(1). doi: 10.1186/s43074-020-00020-y

PDF(70)

Information Metamaterials: bridging the physical world and digital world

Qian Ma, Tie Jun Cui

2020, 1(1). doi: 10.1186/s43074-020-00006-w

PDF(68)

Vacuum-ultraviolet photodetectors

Lemin Jia, Wei Zheng, Feng Huang

2020, 1(1). doi: 10.1186/s43074-020-00022-w

PDF(39)

推荐论文更多+

下载排行更多+

留言板

Feature Article | 新型多功能超透镜助力AR显示集成化

前沿：全光纤单光子探测器

前沿：MXene传感器研究进展—从传统的生物、化学、物理传感器到智能、仿生设备的应用

前沿：基于宽视场光谱巡天望远镜（MUST）的光学系统概念设计

稳步发展！PhotoniX 2022年最新IF公布!

前沿：深度学习助力全景并行相机实现计算变焦成像

前沿：深度学习助力无序光子结构实现光场探测新范式

photonix.springeropen.com