检索结果-内蒙古大学图书馆

A Moment-Based Hermite Weno Scheme with Unified Stencils for Hyperbolic Conservation Laws

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Fan, Chuan Qiu, Jianxian Zhao, Zhuang Department of Mathematical Southern University of Science and Technology Guangdong Shenzhen518055 China School of Mathematical Sciences Fujian Provincial Key Laboratory of Mathematical Modeling and High-Performance Scientific Computing Xiamen University Fujian Xiamen361005 China School of Mathematical Sciences Xiamen University Fujian Xiamen361005 China

In this paper, a fifth-order moment-based Hermite weighted essentially non-oscillatory scheme with unified stencils (termed as HWENO-U) is proposed for hyperbolic conservation laws. The main idea of the HWENO-U scheme is to modify the first-order moment by a HWENO limiter only in the time discretizations using the same information of spatial reconstructions, in which the limiter not only overcomes spurious oscillations well, but also ensures the stability of the fully-discrete scheme. For the HWENO reconstructions, a new scale-invariant nonlinear weight is designed by incorporating only the integral average values of the solution, which keeps all properties of the original one while is more robust for simulating challenging problems with sharp scale variations. Compared with previous HWENO schemes, the advantages of the HWENO-U scheme are: (1) a simpler implemented process involving only a single HWENO reconstruction applied throughout the entire procedures without any modifications for the governing equations;(2) increased efficiency by utilizing the same candidate stencils, reconstructed polynomials, and linear and nonlinear weights in both the HWENO limiter and spatial reconstructions;(3) reduced problem-specific dependencies and improved rationality, as the nonlinear weights are identical for the function $u$ and its non-zero multiple $\zeta u$. Besides, the proposed scheme retains the advantages of previous HWENO schemes, including compact reconstructed stencils and the utilization of artificial linear weights. Extensive benchmarks are carried out to validate the accuracy, efficiency, resolution, and robustness of the proposed scheme. © 2024, The Authors. All rights reserved.

关键词： Efficiency

Analyzing time-dimension communication characterizations for representative scientific applications on supercomputer systems

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2019年第6期13卷 1228-1242页

作者： Juan CHEN Wenhao ZHOU Yong DONG Zhiyuan WANG Chen CUI Feihao WU Enqiang ZHOU Yuhua TANG State Key Laboratory of High Performance Computing College of ComputerNational University of Defense TechnologyChangsha 410073China

Exascale computing is one of the major challenges of this decade,and several studies have shown that communications are becoming one of the bottlenecks for scaling parallel *** analysis on the characteristics of communications can effectively aid to improve the performance of scientific *** this paper,we focus on the statistical regularity in time-dimension communication characteristics for representative scientific applications on supercomputer systems,and then prove that the distribution of communication-event intervals has a power-law decay,which is common in scientific interests and human *** verify the distribution of communication-event intervals has really a power-law decay on the Tianhe-2 supercomputer,and also on the other six parallel systems with three different network topologies and two routing *** order to do a quantitative study on the power-law distribution,we exploit two groups of statistics:bursty *** and periodicity *** results indicate that the communication events show a“strong-bursty and weak-memory”characteristic and the communication event intervals show the periodicity and the ***,our research provides an insight into the relationship between communication optimizations and time-dimension communication characteristics.

关键词： power-law distributions supercomputer systems time-dimension communication characteristics Tianhe-2

Critical grid method: An extensible Smoothed Particle Hydrodynamics fluid general interpolation method for Fluid-Structure Interaction surface coupling based on preCICE

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Long, Sifan Guo, Xiaowei Fan, Xiaokang Yang, Canqun College of Computer Central South University Changsha410083 China Institute for Quantum Information State Key Laboratory of High Performance Computing College of Computer National University of Defense Technology Changsha410073 China College of Computer National University of Defense Technology Changsha410073 China National Supercomputing Center in Tianjin Tianjin300457 China

Solving Fluid-Structure Interaction (FSI) problems using traditional methods is a big challenge in the field of numerical simulation. As a powerful multi-physical field coupled library, preCICE has a bright application prospect for solving FSI, which supports many open/closed source software and commercial CFD solvers to solve FSI problems in the form of a black box. However, this library currently only supports mesh-based coupling schemes. This paper proposes a critical grid (mesh) as an intermediate medium for the particle method to connect a bidirectional coupling tool named preCICE. The particle and critical mesh are used to interpolate the displacement and force so that the pure Lagrangian Smoothed Particle Hydrodynamic (SPH) method can also solve the FSI problem. This method is called the particle mesh coupling (PMC) method, which theoretically solves the mesh mismatch problem based on the particle method to connect preCICE. In addition, we conduct experiments to verify the performance of the PMC method, in which the fluid and the structure is discretized by SPH and the Finite Element Method (FEM), respectively. The results show that the PMC method given in this paper is effective for solving FSI problems. Finally, our source code for the SPH fluid adapter is open-source and available on GitHuba for further developing preCICE compatibility with more meshless methods. © 2024, CC0.

关键词： Fluid structure interaction

Exploring Winograd Convolution for Cost-effective Neural Network Fault Tolerance

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Xue, Xinghua Liu, Cheng Liu, Bo Huang, Haitong Wang, Ying Luo, Tao Zhang, Lei Li, Huawei Li, Xiaowei State Key Lab of Processors Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China University of Chinese Academy of Sciences Beijing100190 China Beijing Institute of Control Engineering China Institute of High Performance Computing Astar Singapore

Winograd is generally utilized to optimize convolution performance and computational efficiency because of the reduced multiplication operations, but the reliability issues brought by winograd are usually overlooked. In this work, we observe the great potential of winograd convolution in improving neural network (NN) fault tolerance. Based on the observation, we evaluate winograd convolution fault tolerance comprehensively from different granularities ranging from models, layers, and operation types for the first time. Then, we explore the use of inherent fault tolerance of winograd convolution for cost-effective NN protection against soft errors. Specifically, we mainly investigate how winograd convolution can be effectively incorporated with classical fault-tolerant design approaches including triple modular redundancy (TMR), fault-aware retraining, and constrained activation functions. According to our experiments, winograd convolution can reduce the fault-tolerant design overhead by 55.77% on average without any accuracy loss compared to standard convolution, and further reduce the computing overhead by 17.24% when the inherent fault tolerance of winograd convolution is considered. When it is applied on fault-tolerant neural networks enhanced with fault-aware retraining and constrained activation functions, the resulting model accuracy generally shows significant improvement in presence of various faults. Copyright © 2023, The Authors. All rights reserved.

关键词： Convolution

Soft Error Reliability Analysis of Vision Transformers

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Xue, Xinghua Liu, Cheng Wang, Ying Yang, Bing Luo, Tao Zhang, Lei Li, Huawei Li, Xiaowei State Key Lab of Processors Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China University of Chinese Academy of Sciences Beijing100190 China Department of Computer Science and Technology Harbin University of Science of Technology Harbin150006 China Institute of High Performance Computing A*STAR Singapore

Vision Transformers (ViTs) that leverage self-attention mechanism have shown superior performance on many classical vision tasks compared to convolutional neural networks (CNNs) and gain increasing popularity recently. Existing ViTs works mainly optimize performance and accuracy, but ViTs reliability issues induced by soft errors in large-scale VLSI designs have generally been overlooked. In this work, we mainly study the reliability of ViTs and investigate the vulnerability from different architecture granularities ranging from models, layers, modules, and patches for the first time. The investigation reveals that ViTs with the self-attention mechanism are generally more resilient on linear computing including general matrix-matrix multiplication (GEMM) and full connection (FC) and show a relatively even vulnerability distribution across the patches. ViTs involve more fragile non-linear computing such as softmax and GELU compared to typical CNNs. With the above observations, we propose a lightweight block-wise algorithm-based fault tolerance (LB-ABFT) approach to protect the linear computing implemented with distinct sizes of GEMM and apply a range-based protection scheme to mitigate soft errors in non-linear computing. According to our experiments, the proposed fault-tolerant approaches enhance ViTs accuracy significantly with minor computing overhead in presence of various soft errors. Copyright © 2023, The Authors. All rights reserved.

关键词： Reliability analysis

A moment-based Hermite WENO scheme with unified stencils for hyperbolic conservation laws1

学校读者我要写书评

暂无评论

arXiv 2024年

In this paper, a fifth-order moment-based Hermite weighted essentially non-oscillatory scheme with unified stencils (termed as HWENO-U) is proposed for hyperbolic conservation laws. The main idea of the HWENO-U scheme is to modify the first-order moment by a HWENO limiter only in the time discretizations using the same information of spatial reconstructions, in which the limiter not only overcomes spurious oscillations well, but also ensures the stability of the fully-discrete scheme. For the HWENO reconstructions, a new scale-invariant nonlinear weight is designed by incorporating only the integral average values of the solution, which keeps all properties of the original one while is more robust for simulating challenging problems with sharp scale variations. Compared with previous HWENO schemes, the advantages of the HWENO-U scheme are: (1) a simpler implemented process involving only a single HWENO reconstruction applied throughout the entire procedures without any modifications for the governing equations;(2) increased efficiency by utilizing the same candidate stencils, reconstructed polynomials, and linear and nonlinear weights in both the HWENO limiter and spatial reconstructions;(3) reduced problem-specific dependencies and improved rationality, as the nonlinear weights are identical for the function u and its non-zero multiple ζu. Besides, the proposed scheme retains the advantages of previous HWENO schemes, including compact reconstructed stencils and the utilization of artificial linear weights. Extensive benchmarks are carried out to validate the accuracy, efficiency, resolution, and robustness of the proposed scheme. Copyright © 2024, The Authors. All rights reserved.

关键词： Efficiency

Low-Complexity and Interpretable Machine Learning Method for Multiple Parameters Estimation

学校读者我要写书评

暂无评论

Low-Complexity and Interpretable Machine Learning Method for...

OptoElectronics and Communications Conference, OECC

作者： Zhiquan Wan Zhenming Yu Kun Xu Kun Yin Hui Yu Chuliang Hu Zhejiang Lab Research Center of High Efficiency Computing System Hangzhou China State Key Laboratory of Information Photonics and Optical Communications Beijing University of Posts and Telecommunications Beijing China Chinese Academy of Sciences Institute of Innovative Computing Technology Hangzhou China

We experimentally demonstrate a random forest- based method for simultaneous modulation format identification and OSNR monitoring. Compared with neural network-based method, this method is interpretable and reduces co... 详细信息

ISBN: (数字)9798350308396

ISBN: (纸本)9798350308402

关键词： Radio frequency Accuracy Parameter estimation Estimation Forestry Optical noise Optical modulation Monitoring Random forests Signal to noise ratio

Neuron Circuit Design and Signal Integrity Analysis of a Large-scale Memristor-based Crossbar Array

学校读者我要写书评

暂无评论

IEEE Transactions on Components, Packaging and Manufacturing technology 2025年

作者： Li, Yan Fang, Lidan Li, Da Zhang, Ling Zhou, Haomiao Wang, Jiawei Liu, En-Xiao Li, Er-Ping China Jiliang University Key Laboratory of Electro-magnetic Wave Information Technology and Metrology of Zhejiang Province College of Information Engineering Hangzhou310018 China China Jiliang University Modern Science and Technology College Yiwu322000 China Zhejiang University Key Laboratory of Advanced Micro/Nano Electronic Devices and Smart System Hangzhou310027 China Urbana–Champaign Institute Zhejiang University Zhejiang University–the University of Illinois Haining314400 China A*STAR Institute of High Performance Computing 1 Fusionpolis Way Singapore Singapore

This paper addresses the limitations of standard chips in processing large-scale datasets by leveraging neuromorphic architectures, particularly spiking neural networks (SNNs), to simulate the pulsed signals of biological brains for enhanced power efficiency and performance. Initially, we developed a CMOS circuit based on the Integrate-and-Fire (IF) neuron model, which generates spike signals with the spatiotemporal dynamics typical of neurons. This approach provides a novel perspective for neuromorphic chip design. Next, we introduce the Partial Element Equivalent Circuit (PEEC) method to establish a unit circuit model for the memristor crossbar array, accounting for all resistance-inductance-capacitance (RLC) parasitic couplings. This model enables precise signal integrity (SI) analysis, offering deeper insights into the signal transmission mechanisms within memristor crossbar arrays and serving as a foundation for optimizing neuromorphic chip performance. Building upon this model, we designed a 24×24 crossbar array hardware prototype, experimentally validating the reliability and feasibility of the circuit. Finally, we perform SI analysis to demonstrate that the performance of the memristor crossbar array can be enhanced by optimizing the rise time, adjusting the memristor’s resistance state, and adding inductance at the output. These findings provide both a theoretical foundation and a technical pathway for improving the performance of neuromorphic chips in practical applications. © 2025 IEEE. All rights reserved.

关键词： Memristors

Sumanene monolayer of pure carbon: a two-dimensional Kagome-analogy lattice with desirable band gap, ultrahigh carrier mobility and strong exciton binding energy

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Shi, Xiaoran Gao, Weiwei Liu, Hongsheng Fu, Zhen-Guo Zhang, Gang Zhang, Yong-Wei Gao, Junfeng Zhao, Jijun Key Laboratory of Materials Modification by Laser Ion and Electron Beams Dalian University of Technology Ministry of Education Dalian116024 China Institute of Applied Physics and Computational Mathematics Beijing100088 China Institute of High Performance Computing A*STAR Singapore138632 Singapore

Design and synthesis of novel two-dimensional (2D) materials that possess robust structural stability and unusual physical properties may open up enormous opportunities for device and engineering applications. Herein we propose a 2D sumanene lattice that be regarded as a derivative of the conventional Kagome lattice. Our tight-binding analysis demonstrates sumanene lattice contains two sets of Dirac cones and two sets of flat bands near the Fermi surface, distinctively different from the Kagome lattice. Using first-principles calculations, we theoretically suggest two possible routines for realization of stable 2D sumanene monolayers (named as α phase and β phase), and α-sumanene monolayer can be experimentally synthesized with chemical vapor deposition using C21H12 as a precursor. Small binding energies on Au(111) surface (e.g. -37.86 eV/Å2 for α phase) signify the possibility of their peel-off after grown on the noble metal substrate. Importantly, our GW plus Bethe-Salpeter equation calculations demonstrate both monolayers have moderate band gaps (1.94 eV for α) and ultrahigh carrier mobilities (3.4×104 cm2V−1s−1 for α). In particular, αsumanene monolayer possesses a strong exciton binding energy of 0.73 eV, suggesting potential applications in optics. Copyright © 2023, The Authors. All rights reserved.

关键词： Monolayers