MRCI程序的并行化

设计了基于空穴-粒子对称的并行化组态相互作用程序,并在一个有12个结点的计算机集群上测试了程序的运行.测试表明,程序的并行化基本上达到了预期的结果,8个结点的加速比达到6.33,在12个结点上成功地完成了5.8亿自旋匹配组态的超大规模MRCI计算.     

SDA 7: A modular and parallel implementation of the simulation of diffusional association software

The simulation of diffusional association (SDA) Brownian dynamics software package has been widely used in the study of biomacromolecular association. Initially developed to calculate bimolecular protein–protein association rate constants, it has since been extended to study electron transfer rates, to predict the structures of biomacromolecular complexes, to investigate the adsorption of proteins to inorganic surfaces, and to simulate the dynamics of large systems containing many biomacromolecular solutes, allowing the study of concentration‐dependent effects. These extensions have led to a number of divergent versions of the software. In this article, we report the development of the latest version of the software (SDA 7). This release was developed to consolidate the existing codes into a single framework, while improving the parallelization of the code to better exploit modern multicore shared memory computer architectures. It is built using a modular object‐oriented programming scheme, to allow for easy maintenance and extension of the software, and includes new features, such as adding flexible solute representations. We discuss a number of application examples, which describe some of the methods available in the release, and provide benchmarking data to demonstrate the parallel performance. © 2015 The Authors. Journal of Computational Chemistry Published by Wiley Periodicals, Inc.     

密度矩阵重正化群的异构并行优化

密度矩阵重正化群方法（DMRG）在求解一维强关联格点模型的基态时可以获得较高的精度，在应用于二维或准二维问题时，要达到类似的精度通常需要较大的计算量与存储空间.本文提出一种新的DMRG异构并行策略，可以同时发挥计算机中央处理器（CPU）和图形处理器（GPU）的计算性能.针对最耗时的哈密顿量对角化部分，实现了数据的分布式存储，并且给出了CPU和GPU之间的负载平衡策略.以费米Hubbard模型为例，测试了异构并行程序在不同DMRG保留状态数下的运行表现，并给出了相应的性能基准.应用于4腿梯子时，观测到了高温超导中常见的电荷密度条纹，此时保留状态数达到10⁴，使用的GPU显存小于12 GB.     

密度矩阵重正化群的异构并行优化

密度矩阵重正化群方法（DMRG）在求解一维强关联格点模型的基态时可以获得较高的精度,在应用于二维或准二维问题时,要达到类似的精度通常需要较大的计算量与存储空间.本文提出一种新的DMRG异构并行策略,可以同时发挥计算机中央处理器（CPU）和图形处理器（GPU）的计算性能.针对最耗时的哈密顿量对角化部分,实现了数据的分布式存储,并且给出了CPU和GPU之间的负载平衡策略.以费米Hubbard模型为例,测试了异构并行程序在不同DMRG保留状态数下的运行表现,并给出了相应的性能基准.应用于4腿梯子时,观测到了高温超导中常见的电荷密度条纹,此时保留状态数达到10⁴,使用的GPU显存小于12 GB.     

The double-tree method: An O(n) unsteady aerodynamic lifting surface method

Two new methods for reducing the computational cost of the unsteady vortex lattice method are developed. These methods use agglomeration to construct time-saving tree structures by approximating the effect of either a group of vortex rings or query points. A case study shows that combining the two new O(n·log n) tree methods together results in an O(n) method, called the double-tree method. Other case studies show that the trade-off between accuracy and speed can be easily and reliably controlled by the agglomeration cutoff distance. For a flat plate with 5 × 200 panels analyzed over 20 time steps, the double-tree method is 7 times faster than the unsteady vortex lattice method with a <5% difference in the force distribution and total lift coefficient. The case studies suggest that the computational benefit will increase for the same level of accuracy if the size of the problem is increased, making the method beneficial for full-aircraft analysis within optimization or dynamic load analysis, where the computational cost of the unsteady vortex lattice method can be large.     

Vortex‐in‐cell method combined with a boundary element method for incompressible viscous flow analysis

In this study, an immersed boundary vortex‐in‐cell (VIC) method for simulating the incompressible flow external to two‐dimensional and three‐dimensional bodies is presented. The vorticity transport equation, which is the governing equation of the VIC method, is represented in a Lagrangian form and solved by the vortex blob representation of the flow field. In the present scheme, the treatment of convection and diffusion is based on the classical fractional step algorithm. The rotational component of the velocity is obtained by solving Poisson's equation using an FFT method on a regular Cartesian grid, and the solenoidal component is determined from solving an integral equation using the panel method for the convection term, and the diffusion term is implemented by a particle strength exchange scheme. Both the no‐slip and no‐through flow conditions associated with the surface boundary condition are satisfied by diffusing vortex sheet and distributing singularities on the body, respectively. The present method is distinguished from other methods by the use of the panel method for the enforcement of the no‐through flow condition. The panel method completes making use of the immersed boundary nature inherent in the VIC method and can be also adopted for the calculation of the pressure field. The overall process is parallelized using message passing interface to manage the extensive computational load in the three‐dimensional flow simulations. Copyright © 2011 John Wiley & Sons, Ltd.     

磁瓶飞行时间谱仪用复合磁场系统设计

根据磁瓶飞行时间谱仪磁场系统的基本要求,设计了一套符合其使用的复合磁场系统。最高强度达1.2 T的强非均匀磁场部分由永久磁铁和结构经过优化的截圆锥状磁极靴所产生。永久磁铁为N52系列NdFeB,磁极靴材料为高磁导率珀明德合金。强度为1.0×10-3 T的弱均匀引导磁场由半径为3 cm、长度为3 m螺线管提供。磁极靴与螺线管之间的距离为5 cm。     

A Parallelizable Task Offloading Model with Trajectory-Prediction for Mobile Edge Networks

As an emerging computing model, edge computing greatly expands the collaboration capabilities of the servers. It makes full use of the available resources around the users to quickly complete the task request coming from the terminal devices. Task offloading is a common solution for improving the efficiency of task execution on edge networks. However, the peculiarities of the edge networks, especially the random access of mobile devices, brings unpredictable challenges to the task offloading in a mobile edge network. In this paper, we propose a trajectory prediction model for moving targets in edge networks without users’ historical paths which represents their habitual movement trajectory. We also put forward a mobility-aware parallelizable task offloading strategy based on a trajectory prediction model and parallel mechanisms of tasks. In our experiments, we compared the hit ratio of the prediction model, network bandwidth and task execution efficiency of the edge networks by using the EUA data set. Experimental results showed that our model is much better than random, non-position prediction parallel, non-parallel strategy-based position prediction. Where the task offloading hit rate is closed to the user’s moving speed, when the speed is less 12.96 m/s, the hit rate can reach more than 80%. Meanwhile, we we also find that the bandwidth occupancy is significantly related to the degree of task parallelism and the number of services running on servers in the network. The parallel strategy can boost network bandwidth utilization by more than eight times when compared to a non-parallel policy as the number of parallel activities grows.     

Hybrid particle‐field molecular dynamics simulations: Parallelization and benchmarks

The parallel implementation of a recently developed hybrid scheme for molecular dynamics (MD) simulations (Milano and Kawakatsu, J Chem Phys 2009, 130, 214106) where self‐consistent field theory (SCF) and particle models are combined is described. Because of the peculiar formulation of the hybrid method, considering single particles interacting with density fields, the most computationally expensive part of the hybrid particle‐field MD simulation can be efficiently parallelized using a straightforward particle decomposition algorithm. Benchmarks of simulations, including comparisons of serial MD and MD‐SCF program profiles, serial MD‐SCF and parallel MD‐SCF program profiles, and parallel benchmarks compared with efficient MD program GROMACS 4.5.4 are tested and reported. The results of benchmarks indicate that the proposed parallelization scheme is very efficient and opens the way to molecular simulations of large scale systems with reasonable computational costs. © 2012 Wiley Periodicals, Inc.     

Parallelization of four-component calculations. II. Symmetry-driven parallelization of the 4-Spinor CCSD algorithm

Given the importance of the Coupled-cluster (CC) method as an efficient and accurate way to take electron correlation into account, we extend the parallelization technique in the second part of this series also to the 4-Spinor CCSD algorithm implemented in the Dirac-Fock packages DIRAC and MOLFDIR. The present implementation is based on the availability of the transformed molecular two-electron integrals on an external storage medium. The linearity of the CC equations in these two-electron integrals is used in a parallelization strategy that is based on distribution of the two largest integral classes that carry three or four virtual spinor indices. The corresponding partial contributions to the T(1) and T(2) amplitudes are calculated on each node and added using Message Passing Interface (MPI) library calls. Although we did not employ a master/slave principle, one specific node was assigned to also perform the remaining serial parts of the algorithm. In the critical sections considerable savings in storage requirements and computer time could be achieved, and this allows for computations on larger systems in the framework of four-component theory.