Femtojoule electro-optic modulation using a silicon– organic hybrid device

Energy-efficient electro-optic modulators are at the heart of short-reach optical interconnects,and silicon photonics is considered the leading technology for realizing such devices.However,the performance of all-silicon devices is limited by intrinsic material properties.In particular,the absence of linear electro-optic effects in silicon renders the integration of energy-efficient photonic–electronic interfaces challenging.Silicon–organic hybrid(SOH)integration can overcome these limitations by combining nanophotonic silicon waveguides with organic cladding materials,thereby offering the prospect of designing optical properties by molecular engineering.In this paper,we demonstrate an SOH Mach–Zehnder modulator with unprecedented efficiency:the 1-mm-long device consumes only 0.7 fJ bit^(-1) to generate a 12.5 Gbit s^(-1) data stream with a bit-error ratio below the threshold for hard-decision forward-error correction.This power consumption represents the lowest value demonstrated for a non-resonant Mach–Zehnder modulator in any material system.It is enabled by a novel class of organic electro-optic materials that are designed for high chromophore density and enhanced molecular orientation.The device features an electro-optic coefficient of r33<180 pm V^(-1) and can be operated at data rates of up to 40 Gbit s^(-1).

From computer vision to short text understanding: Applying similar approaches into different disciplines

With the development of IoT and 5G technologies,more and more online resources are presented in trendy multimodal data forms over the Internet.Hence,effectively processing multimodal information is significant to the development of various online applications,including e-learning and digital health,to just name a few.However,most AI-driven systems or models can only handle limited forms of information.In this study,we investigate the correlation between natural language processing(NLP)and pattern recognition,trying to apply the mainstream approaches and models used in the computer vision(CV)to the task of NLP.Based on two different Twitter datasets,we propose a convolutional neural network based model to interpret the content of short text with different goals and application backgrounds.The experiments have demonstrated that our proposed model shows fairly competitive performance compared to the mainstream recurrent neural network based NLP models such as bidirectional long short-term memory(Bi-LSTM)and bidirectional gate recurrent unit(Bi-GRU).Moreover,the experimental results also demonstrate that the proposed model can precisely locate the key information in the given text.