检索结果-内蒙古大学图书馆

heterogeneous parallel computing accelerated iterative subpixel digital image correlation

Science China(Technological Sciences) 2018年第1期61卷 74-85页

作者： HUANG JianWen ZHANG LingQi JIANG ZhenYu DONG ShouBin CHEN Wei LIU YiPing LIU ZeJia ZHOU LiCheng TANG LiQun State Key Laboratory of Subtropical Building Science School of Civil Engineering and Transportation South China University of Technology School of Computer Science South China University of Technology The State Key Laboratory of Nonlinear Mechanics Institute of Mechanics Chinese Academy of Sciences

parallel computing techniques have been introduced into digital image correlation(DIC) in recent years and leads to a surge in computation speed. The graphics processing unit(GPU)-based parallel computing demonstrated a surprising effect on accelerating the iterative subpixel DIC, compared with CPU-based parallel computing. In this paper, the performances of the two kinds of parallel computing techniques are compared for the previously proposed path-independent DIC method, in which the initial guess for the inverse compositional Gauss-Newton(IC-GN) algorithm at each point of interest(POI) is estimated through the fast Fourier transform-based cross-correlation(FFT-CC) algorithm. Based on the performance evaluation, a heterogeneous parallel computing(HPC) model is proposed with hybrid mode of parallelisms in order to combine the computing power of GPU and multicore CPU. A scheme of trial computation test is developed to optimize the configuration of the HPC model on a specific computer. The proposed HPC model shows excellent performance on a middle-end desktop computer for real-time subpixel DIC with high resolution of more than 10000 POIs per frame.

关键词： digital image correlation(DIC) inverse compositional Gauss-Newton(IC-GN) algorithm heterogeneous parallel computing graphics processing unit(GPU) multicore CPU real-time DIC

来源：评论

学校读者我要写书评

暂无评论

Fast hydrological model calibration based on the heterogeneous parallel computing accelerated shuffled complex evolution method

引用

ENGINEERING OPTIMIZATION 2018年第1期50卷 106-119页

作者： Kan, Guangyuan He, Xiaoyan Ding, Liuqian Li, Jiren Hong, Yang Zuo, Depeng Ren, Minglei Lei, Tianjie Liang, Ke Minist Water Resources China Inst Water Resources & Hydropower Res Res Ctr Flood & Drought Disaster Reduct State Key Lab Simulat & Regulat Water Cycle River Beijing Peoples R China Tsinghua Univ Dept Hydraul Engn State Key Lab Hydrosci & Engn Beijing Peoples R China Univ Oklahoma Dept Civil Engn & Environm Sci Norman OK 73019 USA Beijing Normal Univ Coll Water Sci Beijing Peoples R China Hohai Univ Coll Hydrol & Water Resources Nanjing Jiangsu Peoples R China

Hydrological model calibration has been a hot issue for decades. The shuffled complex evolution method developed at the University of Arizona (SCE-UA) has been proved to be an effective and robust optimization approach. However, its computational efficiency deteriorates significantly when the amount of hydrometeorological data increases. In recent years, the rise of heterogeneous parallel computing has brought hope for the acceleration of hydrological model calibration. This study proposed a parallel SCE-UA method and applied it to the calibration of a watershed rainfall-runoff model, the Xinanjiang model. The parallel method was implemented on heterogeneous computing systems using OpenMP and CUDA. Performance testing and sensitivity analysis were carried out to verify its correctness and efficiency. Comparison results indicated that heterogeneous parallel computing-accelerated SCE-UA converged much more quickly than the original serial version and possessed satisfactory accuracy and stability for the task of fast hydrological model calibration.

关键词： Model calibration Xinanjiang model heterogeneous parallel computing OpenMP CUDA

来源：评论

学校读者我要写书评

暂无评论

heterogeneous parallel computing based real-time chaotic video encryption and its application to drone-oriented secure communication

引用

CHAOS SOLITONS & FRACTALS 2024年 181卷

作者： Shi, Fan-feng Li, Tao Hu, Hao-yu Li, Yi-fei Shan, Dan Jiang, Dong Yangzhou Polytech Inst Sch Informat Engn Yangzhou 225007 Peoples R China Anhui Univ Sch Internet Hefei 230039 Peoples R China Nanjing Univ Dept Math Nanjing 210093 Peoples R China Anhui Univ Natl Engn Res Ctr Agroecol Big Data Anal & Applica Hefei 230601 Peoples R China Hefei Financial China Informat & Technol Co Ltd Fin China Anhui Univ Joint Lab Financial Big Data Hefei 230022 Peoples R China

This paper proposes a real -time video encryption strategy based on multi-round confusion-diffusion architecture and heterogeneous parallel computing. It leverages the powerful computing capacity of the Central Processing Unit (CPU) and the high parallel capability of the Graphics Processing Unit (GPU) to perform byte generation, confusion and diffusion operations concurrently, thereby enhancing computational efficiency. Statistical and security analysis demonstrate that the proposed method exhibits exceptional statistical properties and provides resistance against different types of attacks. Encryption speed evaluation shows that it can realize latency-free 768x768 30FPS video encryption using Intel Xeon Gold 6226R and NVIDIA GeForce RTX 3090, with an average encryption time of 25.12 ms, despite performing seven rounds of confusion and six rounds of diffusion operations on each frame. Additionally, the proposed strategy is adopted to implement a droneoriented secure video communication system, achieving latency-free 256x256 29FPS video encryption with NVIDIA Jetson Xavier NX (NVIDIA Camel ARM CPU and Volta GPU).

关键词： Real-time video encryption heterogeneous parallel computing Chaos Confusion and diffusion

来源：评论

学校读者我要写书评

暂无评论

Utilization of a Web Browser for Complex heterogeneous parallel computing Using Multi-core CPU/GPU Systems 16th

Utilization of a Web Browser for Complex Heterogeneous Paral...

引用

16th International Conference on Computer Aided Systems Theory (EUROCAST)

作者： Woda, Marek Hajduga, Adam Wroclaw Univ Technol Dept Comp Engn 11-17 Janiszewskiego PL-50372 Wroclaw Poland

ISBN: (纸本)9783319747187;9783319747170

Since the invention of the first microprocessor has passed many years. Technological developments in CPU construction is primarily based on increasing the performance of devices, their miniaturisation and the reduction of manufacturing costs. Well known Moore’s Law, speaking of doubling the number of transistors on a chip at regular intervals (going in hand with reduction of manufacturing costs), proved work well over years (initially assumed rate of eighteen months has been slightly extended to two years). Such a trend, due to the technological constraints cannot be everlasting; right now it can be already observed as it slows down. Limitations in minimum size of the individual components (transistors) and a total power draw of a system, forced to change the direction of the technological development. Instead of boost the clock of a processor, it was decided to multiply its number in a chip. Thanks to clustering of processor cores in a single chip that utilise fast shared cache memory, we still can observe considerable performance boost.

关键词： heterogeneous parallel computing HTML5 computing in a web browser OpenCL/WebCL

来源：评论

学校读者我要写书评

暂无评论

Adaptive parallel computing on heterogeneous networks with mpC

引用

parallel computing 2002年第10期28卷 1369-1407页

作者： Lastovetsky, A Univ Coll Dublin Dept Comp Sci Dublin 4 Ireland

The paper presents a new advanced version of the mpC parallel language. The language was designed specially for programming high-performance parallel computations on heterogeneous networks of computers. The advanced version allows the programmer to define at runtime all the main features of the underlying parallel algorithm, which have an impact on the application execution performance, namely, the total number of participating parallel processes, the total volume of compulations to be performed by each of the processes, the total volume of data to be transferred between each pair of the processes, and how exactly the processes interact during the execution of the algorithm. Such an abstraction of parallel algorithm is called a network type in mpC. Given a network type, the programmer can define a network object of this type and describe in details all the computations and communications to be performed on the network object. The mpC programming system uses the information extracted from the network-type definition together with information about the actual performance of the executing network to map the processes of the parallel program to this network in such a way that leads to its better execution time. In addition, the programmer can use a special operator, time of, which predicts the total time of the algorithm execution on the underlying hardware without its real execution. That feature allows the programmer to write such a parallel program that can follow different parallel algorithms to solve the same problem, making choice at runtime depending on the particular executing network and its actual performance. The paper describes both the language model of parallel algorithm and the model of executing parallel environment used by the mpC programming system. It also discusses principles of the implementation of mapping mpC network objects to the computing network. (C) 2002 Elsevier Science B.V. All rights reserved.

关键词： parallel programming language parallel programming system heterogeneous network of computers heterogeneous parallel computing

来源：评论

学校读者我要写书评

暂无评论

TOWARDS MEGACITY-SCALE WIND FLOW SIMULATIONS ON MANY-CORE CPU-ACCELERATOR SYSTEMS

引用

SIAM JOURNAL ON SCIENTIFIC computing 2025年第2期47卷 B402-B427页

作者： Xu, Lei Zhang, Qingyang Lu, Kai Yan, Zhengzheng Gong, Chunye Li, Shengguo Chen, Lin Chen, Xinhai Lin, Xuchuan Liu, Jie Wang, Zheng Chen, Rongliang Chinese Acad Sci Shenzhen Inst Adv Technol Shenzhen 518055 Guangdong Peoples R China State Key Lab High Performance Comp Changsha 410000 Hunan Peoples R China China Earthquake Adm Inst Engn Mech Harbin 150080 Heilongjiang Peoples R China Univ Leeds Sch Comp West Yorkshire LS1 Leeds England

Urban wind flow simulation, based on numerical methods, serves as a powerful tool for understanding the intricate interactions between urban structures and atmospheric conditions. The Lattice Boltzmann method (LBM) stands out as a popular choice for simulating urban wind flow. However, traditional LBM approaches face limitations in terms of scalability on large parallel computing systems and their ability to support high-resolution wind flow simulations across vast megacities spanning hundreds of square kilometers. In response to these challenges, we introduce THLB (Tianhe lattice Boltzmann), a purpose-built LBM simulator tailored for large-scale urban wind flow simulations. THLB streamlines the preprocessing of extensive simulation data through an innovative scheme that automatically identifies flow regions along irregular boundaries. Additionally, THLB integrates a novel processing pipeline and employs parallel optimization techniques, enhancing scalability and performance for large-scale LBM simulations. Our assessment of THLB involves conducting wind flow simulations within a megacity covering an area of 50km \times 40km at an impressive one-meter simulation resolution, featuring 150,000 buildings. This simulation represents the most extensive urban wind flow analysis to date, comprising over two trillion simulation lattices. We gauge THLB's performance on the Tianhe new-generation supercomputer, harnessing more than 155 million heterogeneous cores. Our experimental results demonstrate exceptional performance and scalability, achieving a peak computation throughput of 24,553.43 G Lattices Updates Per Second (GLUPS), setting a new state-of-the-art benchmark for LBM simulations. Despite the inherent challenges of large-scale LBM simulations, our approach showcases robust scalability, delivering 90.48\% and 69.91\% of weak and strong scaling efficiency, respectively.

关键词： lattice Boltzmann method heterogeneous parallel computing high-performance computing city-wide meteorological micro-scale computational wind engineering

来源：评论

学校读者我要写书评

暂无评论

Fast Atmospheric Aerosol Size and Shape Imaging Instrument: Design, Calibration, and Intelligent Interaction

引用

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 2025年 74卷

作者： Dong, Li Han, Yong Hu, Maohai Zhang, Yurong Zhou, Qicheng Sun Yat Sen Univ Sch Atmospher Sci Adv Sci & Technol Space Atmospher Phys Grp ASAG Zhuhai 519082 Peoples R China Nanjing Univ Sci & Technol Sch Elect & Opt Engn Nanjing 210094 Peoples R China

Atmospheric aerosol particles have a significant impact on radiation, climate, and human health, with their size and shape being fundamental physical parameters for atmospheric change research. Due to the widespread effects and applications of aerosol particles, the direct measurement of aerosol size and shape has become crucial. Nevertheless, several challenges persist in aerosol measurement instruments, including limited resolution, complex operation, poor synchronization, and inaccurate inversion methods. Therefore, we developed a new scientific instrument and corresponding image intelligent interaction system, whose name is the fast atmospheric aerosol size and shape imaging instrument (FASI). The instrument is designed for transmission imaging that contains a light source, imaging chamber, microscope objective, tube lens, extension tube, camera, etc. Before the operation, the FASI calibrates background field, pixel size, characteristic gray value (CGV), and depth of field (DOF) based on image processing. During intelligent interaction, the FASI extracts aerosol particles by image denoising and edge detection, and then uses our proposed defocus and duplicate particle detection algorithms for secondary screening of aerosols. Aerosol size and shape parameters are measured in parallel by the central processing unit (CPU) and the graphics processing unit (GPU) using heterogeneous computation. Polystyrene latex (PSL) calculations and quantitative experiments indicate that FASI can accurately detect 0.5-20 mu m aerosol particles. In particular, the FASI measures aerosol particles supplied by an aerosol generator, dryer, and neutralizer, demonstrating that the aerosol size distribution range of oil solutions (0.5-3.5 mu m) is narrower than that of aqueous solutions (0.5-7.5 mu m). For all samples, 92.12% of aerosols have an aspect ratio (AR) exceeding 1, and the shape of these nonspherical aerosols varies greatly from each other. The evaluations of computational effici

关键词： Aerosols Particle measurements Aerosol shape aerosol size computer vision heterogeneous parallel computing intelligent instrument Aerosol shape aerosol size computer vision heterogeneous parallel computing intelligent instrument

来源：评论

学校读者我要写书评

暂无评论

Programming heterogeneous parallel Machines Using Refactoring and Monte-Carlo Tree Search

引用

INTERNATIONAL JOURNAL OF parallel PROGRAMMING 2020年第4期48卷 583-602页

作者： Brown, Christopher Janjic, Vladimir Goli, M. McCall, J. Univ St Andrews Sch Comp Sci St Andrews Fife Scotland Robert Gordon Univ Aberdeen Scotland

This paper presents a new technique for introducing and tuning parallelism for heterogeneous shared-memory systems (comprising a mixture of CPUs and GPUs), using a combination of algorithmic skeletons (such as farms and pipelines), Monte-Carlo tree search for deriving mappings of tasks to available hardware resources, and refactoring tool support for applying the patterns and mappings in an easy and effective way. Using our approach, we demonstrate easily obtainable, significant and scalable speedups on a number of case studies showing speedups of up to 41 over the sequential code on a 24-core machine with one GPU. We also demonstrate that the speedups obtained by mappings derived by the MCTS algorithm are within 5-15% of the best-obtained manual parallelisation.

关键词： heterogeneous parallel computing Monte-Carlo tree search Optimisations

来源：评论

学校读者我要写书评

暂无评论

A parallel computing method using blocked format with optimal partitioning for SpMV on GPU

引用

JOURNAL OF COMPUTER AND SYSTEM SCIENCES 2018年 92卷 152-170页

作者： Yang, Wangdong Li, Kenli Li, Keqin Hunan Univ Coll Comp Sci & Elect Engn Changsha 410082 Hunan Peoples R China Hunan City Univ Coll Informat Sci & Engn Yiyang 413000 Hunan Peoples R China Natl Supercomp Ctr Changsha Changsha 410082 Hunan Peoples R China SUNY Coll New Paltz Dept Comp Sci New Paltz NY 12561 USA

For large-scale sparse matrices, SpMV cannot be processed on GPU using the common storage formats because of the memory limitation. In addition, the parallel effect is poor using general formats for the sparse matrices with extremely uneven distribution of non-zero elements, which leads to performance deterioration. This paper presents an optimal partitioning strategy based on the distribution of non-zero elements in a sparse matrix to improve the performance of SpMV, and uses a hybrid format, which mixes CSR and ELL formats, to store the blocks partitioned from the sparse matrix. The hybrid blocked format has better compression effect and more uniform distribution of non-zero elements, which can be suitable for more types of sparse matrices. Our pattitioning strategy is proven to be optimal, which can yield the minimum parallel execution time on GPU. (C) 2017 Elsevier Inc. All rights reserved.

关键词： Blocked format CPU/GPU Dynamic programming heterogeneous parallel computing Partitioning Reordering Sparse matrix-vector multiplication

来源：评论

学校读者我要写书评

暂无评论

A machine learning approach for accelerating DNA sequence analysis

引用

INTERNATIONAL JOURNAL OF HIGH PERFORMANCE computing APPLICATIONS 2018年第3期32卷 363-379页

作者： Memeti, Suejb Pllana, Sabri Linnaeus Univ Dept Comp Sci S-35195 Vaxjo Sweden

The DNA sequence analysis is a data and computationally intensive problem and therefore demands suitable parallel computing resources and algorithms. In this paper, we describe an optimized approach for DNA sequence analysis on a heterogeneous platform that is accelerated with the Intel Xeon Phi. Such platforms commonly comprise one or two general purpose host central processing units (CPUs) and one or more Xeon Phi devices. We present a parallel algorithm that shares the work of DNA sequence analysis between the host CPUs and the Xeon Phi device to reduce the overall analysis time. For automatic worksharing we use a supervised machine learning approach, which predicts the performance of DNA sequence analysis on the host and device and accordingly maps fractions of the DNA sequence to the host and device. We evaluate our approach empirically using real-world DNA segments for human and various animals on a heterogeneous platform that comprises two 12-core Intel Xeon E5 CPUs and an Intel Xeon Phi 7120P device with 61 cores.

关键词： DNA sequence analysis machine learning heterogeneous parallel computing

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：