A5100768288- VLSI and Analog Circuit Testing
- Integrated Circuits and Semiconductor Failure Analysis
- Advanced Memory and Neural Computing
- Parallel Computing and Optimization Techniques
- Low-power high-performance VLSI design
- Physical Unclonable Functions (PUFs) and Hardware Security
- Radiation Effects in Electronics
- Interconnection Networks and Systems
- Software Testing and Debugging Techniques
- Advanced Neural Network Applications
- Ferroelectric and Negative Capacitance Devices
- Embedded Systems Design Techniques
- Advanced Data Storage Technologies
- VLSI and FPGA Design Techniques
- Semiconductor materials and devices
- Formal Methods in Verification
- Neuroscience and Neural Engineering
- Advanced Image and Video Retrieval Techniques
- Adversarial Robustness in Machine Learning
- Engineering and Test Systems
- CCD and CMOS Imaging Sensors
- Cryptography and Data Security
- Advancements in Semiconductor Devices and Circuit Design
- Graph Theory and Algorithms
- Advancements in Photolithography Techniques
Chinese Academy of Sciences
2016-2025
University of Chinese Academy of Sciences
2016-2025
Institute of Computing Technology
2016-2025
Chinese Academy of Geological Sciences
2021-2025
Czech Academy of Sciences, Institute of Geology
2025
China University of Petroleum, Beijing
2025
Peng Cheng Laboratory
2019-2024
Beijing Jiaotong University
2006-2024
Hebei University of Science and Technology
2024
North China University of Water Resources and Electric Power
2021-2024
Recent advances in Neural Networks (NN) are enabling more and innovative applications. As an energy-efficient hardware solution, machine learning accelerators for CNNs or traditional ANNs also gaining popularity the area of embedded vision, robotics cyberphysics. However, design parameters NN models vary significantly from application to application. Hence, it's hard provide one general highly-efficient solution accommodate all them, it is impractical domain-specific developers customize...
Graph neural networks (GNNs) emerge as a powerful approach to process non-euclidean data structures and have been proved in various application domains such social e-commerce. While graph maintained real-world systems can be extremely large sparse, thus employing GNNs deal with them requires substantial computational memory overhead, which induces considerable energy resource cost on CPUs GPUs. In this article, we present specialized accelerator architecture, EnGN, enable high-throughput...
The study objective was to investigate the use of peripheral blood biomarkers as predictors patient survival. aim this identify baseline associated with clinical outcome in patients early lung cancer (stage I-II) treated surgery.We included and analysed data from 376 early-stage who underwent a standard lobectomy. Univariate multivariate Cox regression analyses were performed on all assess relationships between progression-free survival (PFS) overall (OS) biomarker metrics measured before...
Applications' traffic tends to be bursty and the location of hot-spot nodes moves as time goes by. This will significantly aggregate blocking problem wormhole-routed Network-on-Chip (NoC). Most state-of-the-art balancing solutions are based on fully adaptive routing algorithms which may introduce large time/space overhead routers. Partially algorithms, other hand, efficient, but lack even or sufficient adaptiveness. Reconfigurable could provide on-demand adaptiveness for reducing blocking,...
We developed a predictive model associated with ferroptosis to provide more comprehensive view of esophageal squamous cell carcinoma (ESCC) immunotherapy. Gene expression data and corresponding clinical outcomes were obtained from the GEO The Cancer Genome Atlas (TCGA) databases, ferroptosis-related gene set was FerrDb database. identified 45 genes that differentially expressed, including enrichment in involved immune system process. established gene-based prognostic based on results...
As large language models (LLMs) like ChatGPT exhibited unprecedented machine intelligence, it also shows great performance in assisting hardware engineers to realize higher-efficiency logic design via natural interaction. To estimate the potential of process assisted by LLMs, this work attempts demonstrate an automated environment that explores LLMs generate designs from specifications. a more accessible and efficient chip development flow, we present scalable four-stage zero-code framework...
Recent advances in large language models have demonstrated their potential for automated generation of hardware description (HDL) code from high-level prompts.Researchers utilized fine-tuning to enhance the ability these (LLMs) field Chip Design.However, lack Verilog data hinders further improvement quality by LLMs.Additionally, absence a and electronic design automation (EDA) script augmentation framework significantly increases time required prepare training dataset LLM trainers.This paper...
Homogeneous manycore systems are emerging for tera-scale computation and typically utilize Network-on-Chip (NoC) as the communication scheme between embedded cores. Effective defect tolerance techniques essential to improve yield of such complex integrated circuits. We propose achieve fault by employing redundancy at core-level instead microarchitecture level. When faulty cores exist on-chip in this architecture, however, physical topologies various manufactured chips can be significantly...
Phase change at the nanoscale is heart of many biological and geological phenomena. The recent emergence global implications unconventional oil gas production from nanoporous shale further necessitate a higher understanding phase behavior these scales. Here, we directly observe condensation condensate growth light hydrocarbon (propane) in discrete sub-100 nm (∼70 nm) channels. Two different mechanisms this are distinguished, continuous discontinuous due to liquid bridging ahead meniscus,...
Prior works typically conducted the fault analysis of neural network accelerator computing arrays with simulation and focused on prediction accuracy loss models. There is still a lack systematic acceleration system that considers both degradation exceptions, such as stall running overtime. To end, we implemented representative corresponding injection modules Xilinx ARM-FPGA platform evaluated reliability under different rates when series typical models are deployed system. The entire...
Number theoretic transform (NTT) is useful for the acceleration of polynomial multiplication, which main performance bottleneck in next-generation cryptographic schemes. Different NTT-based algorithms have different security settings. The diverse application scenarios introduce cost-performance tradeoffs and hardware constraints. Motivated by emerging demand more versatile NTT accelerators, we propose a new design methodology that can generate area-efficient high-performance accelerators any...
By leveraging the power of open-source software, EDA tool offers a cost-effective and flexible solution for designers, researchers, hobbyists alike. Open-source promotes collaboration, innovation, knowledge sharing within community. It emphasizes role toolchain in accelerating development electronic systems, reducing design costs, improving quality. This paper presents an project, iEDA, aiming to build basic infrastructure technology evolution closing industrial-academic gap area. As...
Modern processors integrate carefully designed micro-architectural components within the front-end to optimize performance. These include instruction cache, micro-operation and prefetcher. Through experimentation, we observed that rate of generation in fetch unit markedly exceeds execution decode unit. However, existing frameworks fail explain this phenomenon. Consequently, empirically validate presence an optimization feature, referred as Fetch Target Queue (FTQ), Intel processor. To best...
<title>Abstract</title> The Tethyan orogenic belt underwent prolonged tectonic evolution and hosts numerous world-class porphyry copper deposits. Notably, most deposits are associated with Cenozoic continental collision, while fewer formed during Mesozoic subduction. Here we integrate detrital zircon oxybarometry geochemical data, stratigraphy, sea-level temperature fluctuations, major geological events. Our results reveal a stark redox transition–from anoxic subduction to oxidized...
With the development of intelligent manufacturing systems, data-driven fault diagnosis has become a hot research topic. Traditional methods often rely on expert-extracted features, wherein feature extraction process requires considerable effort and affects final results to great extent. However, end-to-end based deep learning can automatically learn representations from raw data. In this study, first, vibration signals various states planetary gearbox were segmented preprocessed into input...