检索结果-南通市图书馆

Fast Parallel Cutoff Pair Interactions for Molecular Dynamics on Heterogeneous Systems

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Tsinghua Science and Technology 2012年第3期17卷 265-277页

作者： Qiang Wu canqun yang Tao Tang Kai Lu School of Computer Science National University of Defense TechnologyChangsha 410073China

Heterogeneous systems with both Central Processing Units （CPUs） and Graphics Processing Units （GPUs） are frequently used to accelerate short-ranged Molecular Dynamics （MD） simulations. The most time-consuming task in short-ranged MD simulations is the computation of particle-to-particle interac- tions. Beyond a certain distance, these interactions decrease to zero. To minimize the operations to investi- gate distance, previous works have tiled interactions by employing the spatial attribute, which increases the memory access and GPU computations, hence decreasing performance. Other studies ignore the spatial attribute and construct an all-versus-all interaction matrix, which has poor scalability. This paper presents an improved algorithm. The algorithm first bins particles into voxels according to the spatial attributes, and then tiles the all-versus-all matrix into voxel-versus-voxel sub-matrixes. Only the sub-matrixes between neighbor- ing voxels are computed on the GPU. Therefore, the algorithm reduces the distance examine operations and limits additional memory access and GPU computations. This paper also adopts a multi-level program- ming model to implement the algorithm on multi-nodes of Tianhe-lA. By employing （1） a patch design to ex- ploit parallelism across the simulation domain, （2） a communication overlapping method to overlap the communications between CPUs and GPUs, and （3） a dynamic workload balancing method to adjust the workloads among compute nodes, the implementation achieves a speedup of 4.16x on one NVIDIA Tesla M2050 GPU compared to a 2.93 GHz six-core Intel Xeon X5670 CPU. In addition, it runs 2.41x faster on 256 compute nodes of Tianhe-lA （with two CPUs and one GPU inside a node） than on 256 GPU-excluded nodes.

关键词： cutoff pair interactions molecular dynamics heterogeneous computing GPU computing

MPI Based 3D Plasma Simulation on Xeon Phi

在线全文

学校读者我要写书评

暂无评论

MPI Based 3D Plasma Simulation on Xeon Phi

第十七届计算机工程与工艺年会暨第三届微处理器技术论坛

作者： Yao Wenke yang canqun Wu Qiang Chen Cheng Qi Jin School of Computer Science National University of Defense Technology

3D plasma simulations have been widely used in the study of *** paper ports a particle-in-cell based plasma code,LARED-P,on Xeon Phi,which is a many-core coprocessor with 512-bit wide vector *** port LARED-P on Xeon Phi by employing MPI *** also optimize the implementation through manually *** evaluation shows that our implementation scales well on Xeon ***,the SIMD-based optimization achieves speedups up to 4.61.

关键词： LARED-P Xeon Phi many-core coprocessor 512bit SIMD

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

基于并行构件的动态负载池平衡技术的研究

计算机工程 2007年第13期33卷 62-64页

作者：杜静敖富江杨灿群国防科技大学计算机学院长沙410073

在基于并行构件的系统中,负载平衡技术是充分利用系统资源,提高系统性能的关键技术。该文提出了一种基于并行构件的动态负载池平衡技术,与已有的负载平衡策略不同,该技术综合考虑了通信延迟、容错技术以及数据一致性对负载平衡策略的影... 详细信息

在基于并行构件的系统中,负载平衡技术是充分利用系统资源,提高系统性能的关键技术。该文提出了一种基于并行构件的动态负载池平衡技术,与已有的负载平衡策略不同,该技术综合考虑了通信延迟、容错技术以及数据一致性对负载平衡策略的影响,旨在提高策略的准确性、高效性和健壮性,增强运行时系统的性能。

关键词：负载平衡处理器池通信延迟容错数据一致性

在线全文

学校读者我要写书评

暂无评论

编译指导的构件并行技术

计算机应用与软件 2005年第10期22卷 11-13页

作者：杜静杨灿群敖富江国防科技大学计算机学院湖南长沙410073

提出了一种新的面向科学计算的构件技术——编译指导的构件并行技术,旨在提高构件间的并行度和数据的局部性,避免通信瓶颈。该技术分为两个阶段:域划分和子构件组合。其中域划分利用了编译时的数据依赖分析技术,子构件组合利用了所设计... 详细信息

提出了一种新的面向科学计算的构件技术——编译指导的构件并行技术,旨在提高构件间的并行度和数据的局部性,避免通信瓶颈。该技术分为两个阶段:域划分和子构件组合。其中域划分利用了编译时的数据依赖分析技术,子构件组合利用了所设计的基于区间数据的模糊聚类算法。

关键词：科学计算并行计算通信瓶颈并行度数据局部性构件技术并行技术编译数据依赖分析模糊聚类算法

在线全文

学校读者我要写书评

暂无评论

编译器前端乘幂运算的实现与优化

计算机工程与应用 2004年第36期40卷 47-49,80页

作者：王锋杨灿群国防科技大学计算机学院长沙410073

乘幂运算是最基本的数学表达式之一,并且被大多数编译器支持,在高性能计算和密码算法中应用十分频繁。为了实现和优化乘幂运算人们提出了许多算法,比如加法链算法。该文描述了在编译器前端一个具体的新的算法,它实现了乘幂运算并且进行... 详细信息

乘幂运算是最基本的数学表达式之一,并且被大多数编译器支持,在高性能计算和密码算法中应用十分频繁。为了实现和优化乘幂运算人们提出了许多算法,比如加法链算法。该文描述了在编译器前端一个具体的新的算法,它实现了乘幂运算并且进行了优化,优化方法结合了静态最优加法链算法和Window算法。这个新的算法已经应用于GNUFortran95编译器(Gfortran),并取得良好效果。

关键词：乘幂运算加法链 Window算法编译器优化前端

在线全文

学校读者我要写书评

暂无评论

基于内存缓存的异步检查点容错技术

计算机研究与发展 2014年第6期51卷 1229-1239页

作者：易会战王锋左克杨灿群杜云飞马亚青国防科学技术大学计算机学院长沙410073 北方车辆研究所北京100072

高性能计算机系统规模越来越大,系统可靠性问题越来越严重.检查点技术是最典型的容错方法,但是因为并行文件系统的性能提高相对缓慢,数据写带宽低,传统检查点方法产生了严峻的性能问题.针对当前计算机系统计算和存储资源丰富,而并行文... 详细信息

高性能计算机系统规模越来越大,系统可靠性问题越来越严重.检查点技术是最典型的容错方法,但是因为并行文件系统的性能提高相对缓慢,数据写带宽低,传统检查点方法产生了严峻的性能问题.针对当前计算机系统计算和存储资源丰富,而并行文件系统写带宽提高相对滞后的特点,提出了基于内存缓存的异步检查点容错技术,传统的检查点技术被划分为两步:检查点文件首先被缓存在计算结点的局部内存,然后使用一个独立的帮助任务将数据拷贝到并行文件系统.利用局部内存带宽高以及帮助任务和计算任务并行执行的特点,新方法极大减小了检查点容错引入的时间开销,模拟和实际程序测试验证了异步检查点容错技术的有效性.

关键词：内存缓存异步检查点容错 JASMIN

基于Imagine的科学程序局部性优化方法

在线全文

学校读者我要写书评

暂无评论

计算机研究与发展 2007年第Z1期44卷 35-39页

作者：杜静贾佳杨学军杨灿群李春江国防科学技术大学计算机学院长沙410073 国防科学技术大学计算机学院长沙410073 国防科学技术大学计算机学院长沙410073 国防科学技术大学计算机学院长沙410073 国防科学技术大学计算机学院长沙410073

基于Imagine体系结构,提出了一种科学程序局部性优化方法,旨在提高流程序的带宽利用率并保证Imagine强大的计算能力.关键技术在于通过对循环的计算变换和数据变换来开发体系结构的优势.对4个典型科学程序的实验表明,该优化能够有效地提... 详细信息

基于Imagine体系结构,提出了一种科学程序局部性优化方法,旨在提高流程序的带宽利用率并保证Imagine强大的计算能力.关键技术在于通过对循环的计算变换和数据变换来开发体系结构的优势.对4个典型科学程序的实验表明,该优化能够有效地提高程序计算密集性且减少索引流,从而增强程序的局部性.

关键词： Imagine 科学程序局部性计算密集性索引流

Anomalous quality factor evolution of induced transparency in a virtualized coupled-oscillator system

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Physics,Mechanics & Astronomy) 2025年第1期68卷 144-151页

作者： Yunhan yang Han Jia Tuo Liu Yuzhen yang Jun yang Key Laboratory of Noise and Vibration Research Institute of AcousticsChinese Academy of SciencesBeijing 100190China University of Chinese Academy of Sciences Beijing 100049China State Key Laboratory of Acoustics Institute of AcousticsChinese Academy of SciencesBeijing 100190China

The phenomenon of induced transparency,characterized by a high quality(Q)factor,holds significant importance in the fields of high-precision sensing,energy storage and wave *** transparency systems are often constrained by the limited experimentally accessible parameter space and the lack of independent control over intrinsic parameters,thereby hindering advanced research on high-Q *** we propose a virtualized coupled-oscillator system that enables reconfigurable induced transparency in the reflection *** virtually incorporating a pair of coupled oscillators into an individual acoustic meta-atom,we achieve the manipulation of induced transparency in a decoupled,wide-ranging,and elaborate *** the diverse evolutions across the intrinsic parameter space,we show experimentally a counter-intuitive increase in the Q-factor of the transparency window,which is attributed to the enhanced dissipation of the bright ***,a constant group delay at the center frequency is observed accompanying this *** work demonstrates an intriguing platform as well as a novel strategy to engineer high-Q induced transparency phenomenon,and paves the way towards applications in programmable acoustic sensors,filters and slow-wave devices.

关键词： induced transparency acoustic meta-atom high quality-factor scattering programmable intrinsic parameter

Rapid reduction of air pollution and short-term exposure risks in China

在线全文

学校读者我要写书评

暂无评论

Journal of Environmental Sciences 2025年第3期149卷 126-138页

作者： Hao Fan Chuanfeng Zhao Yikun yang Xingchuan yang State Key Joint Laboratory of Environmental Simulation and Pollution Control School of EnvironmentBeijing Normal UniversityBeijing 100875China Laboratory for Climate and Ocean-Atmospheric Studies Department of Atmospheric and Oceanic SciencesSchool of PhysicsPeking UniversityBeijing 100871China College of Resource Environment and Tourism Capital Normal UniversityBeijing 100048China

With the continuous control of anthropogenic emissions,China’s air quality has improved significantly in recent *** this background,research on how the short-term exposure risks caused by air pollution in China have changed is *** study utilized hourly concentration data from ground observation stations and the official air quality guidelines of the Ministry of Ecology and Environment of China and the World Health Organization as standards to systematically investigate the spatiotemporal characteristics and short-term exposure risks of air pollution in China from 2015 to *** results indicate that various atmospheric pollutants except for ozone showed a decreasing trend ***,both single pollutant air pollution days(SAPDs)and multiple pollutant air pollution days(MAPDs)showed varying degrees of reduction within 15 and 25 days,*** was dominated mainly by excessive PM_(2.5)and PM_(10)pollutants,while MAPD was dominated mainly by excessive pollutant combinations,including PM_(2.5)+PM_(10),CO+PM_(2.5)+PM_(10),and SO_(2)+PM_(2.5)+PM_(10).As the concentration of atmospheric pollutants decreased,the total excess risk(ER)decreased yearly from 2015 to 2022,but there were significant regional ***,the ER is less than 0.25%in southern China,in the range of 0.25%-0.5%in the North China Plain and some cities in the northeast,and higher than 1%in the *** matter is currently the primary pollutant posing short-term exposure risk in China,especially due to the impact of sandstorm *** study indicates that China’s atmospheric cleaning action is significantly beneficial for reducing health risks.

关键词： Air pollution Emission-reduction Total excess risk Particulate matter China