基于自回避行走数值模拟的并行计算

提出一种蒙特卡洛方法并行计算高分子的单链性质,高分子链采用自回避行走方法生成初始状态,链节点的随机运动通过键长涨落方法进行模拟。自回避链的配分函数没有解析解,研究高分子链的性质往往要计算大量的样本。利用样本之间的独立性进行并行计算达到线性加速比,使蒙特卡洛模拟高分子链的并行计算时间缩短到科学计算可以接受的时间范围。     

基于SMP集群的激光化学反应模拟效率分析*

基于半经典分子动力学模型,在SMP集群中实现激光化学反应双层并行模拟系统。结合粗粒度的原子分解算法和细粒度的矩阵并行乘法实现激光化学反应模拟中力计算部分的并行化,分析粒度划分对半经典分子动力学模拟并行效率的影响。在SMP集群中测试表明,采用128个处理器模拟由500个C原子构成的分子体系,并行效率可达70%。在CPU数量固定的情况下,SMP节点内的细粒度的并行对提高半经典分子动力学模拟并行效率影响较大。该系统能够模拟大分子体系的激光化学反应,在提高加速比的同时保证计算资源的利用效率,满足激光化学反应模拟需求。     

多核计算机上的并行计算

研究了多核计算机上0penMP＋Vc＋＋编程模式的并行程序,并在双核和四核计算机上分别使用传统算法和并行算法计算数列求和、矩阵乘积及矩阵Cholesky分解。试验表明,传统串行程序只能利用多核计算机的一个核资源,而采用OpenMP程序的并行效率很高。     

PowerEdge 1750微机集群并行性能测试及实例分析

随着地震勘探新技术的发展和应用,地震数据量和处理量已变得越来越大。微机集群凭借其良好的性价比和高效的运算速度已逐渐成为地震数据处理的主要平台。本文以PowerEdge1750微机集群为例,首先对其体系结构进行了分析,然后给出了雅可比和矩阵乘积两个基准MPI程序的测试,最后以计算量较大的相干体算法为例,进行了实际并行计算与应用。文中给出的并行计算结果与性能分析,为推广应用该类集群提供了借鉴。     

淮河沙颍河支流水质预报模型的参数率定

水质模型的参数率定一直是水质预报的难点。由于要求模型计算结果准确，故必须对每个参数反复多次计算才能给出一个较为可用的参数值。用传统的串行计算方法来处理参数率定问题，其计算时间过长，无法快速给出准确的模型。文章利用已有的数据和经验，根据沙颖河具体的水质和应用特点，在国内首次提出了采用基于并行计算技术的高性能计算方法，对沙颖河支流水质模型进行多参数率定的方案，并在曙光系列并行机上得以实现。结果表明：其率定的准确性大大优于常规率定，并缩短了整个模型的建模时间，由此为模型的快速应用提供了可靠的参数依据。     

基于GPU的数学形态学运算并行加速研究

数学形态学运算是一种高度并行的运算,其计算量大而又如此广泛地应用于对实时性要求较高的诸多重要领域。为了提高数学形态学运算的速度,提出了一种基于CUDA架构的GPU并行数学形态学运算。文章详细描述了GPU硬件架构和CUDA编程模型,并给出了GPU腐蚀并行运算的详细实现过程以及编程过程中为充分利用GPU资源所需要注意的具体问题。实验结果表明,GPU并行数学形态学运算速度可达到几个数量级的提高。     

谱方法求解水声传播问题的优化与并行

水声传播数值计算的效率是各类水声学应用关心的核心因素之一,谱方法作为求解微分方程的一种数值方法,具有精度高、收敛速度快等优点,因此,近年来利用简正波-谱方法求解水声传播方程引起了许多学者的关注;然而,谱方法计算量更大,计算时间更长,在求解大范围海域声传播问题时,难以满足实时性的需求.因此,需要借助现代高性能计算机系统,...     

Parallel evolutionary algorithms can achieve super-linear performance

One of the main reasons for using parallel evolutionary algorithms (PEAs) is to obtain efficient algorithms with an execution time much lower than that of their sequential counterparts in order, e.g., to tackle more complex problems. This naturally leads to measuring the speedup of the PEA. PEAs have sometimes been reported to provide super-linear performances for different problems, parameterizations, and machines. Super-linear speedup means that using “m” processors leads to an algorithm that runs more than “m” times faster than the sequential version. However, reporting super-linear speedup is controversial, especially for the “traditional” research community, since some non-orthodox practices could be thought of being the cause for this result. Therefore, we begin by offering a taxonomy for speedup, in order to clarify what is being measured. Also, we analyze the sources for such a scenario in this paper. Finally, we study an assorted set of results. Our conclusion is that super-linear performance is possible for PEAs, theoretically and in practice, both in homogeneous and in heterogeneous parallel machines.     

Path Integration on a Quantum Computer

We study path integration on a quantum computer that performs quantum summation. We assume that the measure of path integration is Gaussian, with the eigenvalues of its covariance operator of order j^-k with k>1. For the Wiener measure occurring in many applications we have k=2. We want to compute an -approximation to path integrals whose integrands are at least Lipschitz. We prove: Path integration on a quantum computer is tractable. Path integration on a quantum computer can be solved roughly ^-1 times faster than on a classical computer using randomization, and exponentially faster than on a classical computer with a worst case assurance. The number of quantum queries needed to solve path integration is roughly the square root of the number of function values needed on a classical computer using randomization. More precisely, the number of quantum queries is at most 4.46 ^-1. Furthermore, a lower bound is obtained for the minimal number of quantum queries which shows that this bound cannot be significantly improved. The number of qubits is polynomial in ^-1. Furthermore, for the Wiener measure the degree is 2 for Lipschitz functions, and the degree is 1 for smoother integrands. PACS: 03.67.Lx; 31.15Kb; 31.15.-p; 02.70.-c     

超立方多处理机上大型线性方程组并行迭代求解算法

给出了超立方多处理机系统上大型线性方程组并行迭代求解算法设计及其运行时间复杂性分析,并在并行虚拟环境（ＰＶＭ环境）下做了数值试验,求出了在多台工作站ＳＵＮ４上的运行时间及运行加速比。试验结果表明,算法在超立方上有很好的运行效果。