多核CPU和GPU加速分子动力学模拟

在多核中央处理器(CPU)—图形处理器(GPU)异构并行体系结构上,采用OpenMP和计算统一设备架构(CUDA)编程实现了基于AMBER力场的蛋白质分子动力学模拟程序。通过合理地将程序划分为CPU单线程、CPU多线程和GPU多线程执行部分,高效地利用了计算机的处理能力。性能测试结果表明,相对于优化后的CPU串行计算,多核CPU-GPU异构并行计算模型有强大的性能优势,特别是将占整个程序执行时间90%的作用力的计算移植到GPU上执行,获得了最高可达12倍的计算加速比。     

Modeling laser induced lubricant depletion in heat-assisted-magnetic recording systems using a multiple-layered disk structure

In this paper, we model the depletion of lubricant from a disk surface subject to heating by a scanning laser in a heat assisted magnetic recording (HAMR) system. A multi-layer disk structure is used consisting of the substrate (either glass or aluminum), the CoFe based soft magnetic under- layer, a Ru based intermediate layer, a CoCrPt based recording layer, the diamond-like-carbon layer, and the lubricant film. The thickness and material properties of the different layers are shown to play an important role in the conduction of heat from the top layer to the bottom layer and, consequently, in the lubricant depletion process due to heating by a scanning laser. The results show that it is critical to include realistic multi-layer disk structures in HAMR lubricant depletion modeling.     

Algorithm optimization in molecular dynamics simulation

Establishing the neighbor list to efficiently calculate the inter-atomic forces consumes the majority of computation time in molecular dynamics (MD) simulation. Several algorithms have been proposed to improve the computation efficiency for short-range interaction in recent years, although an optimized numerical algorithm has not been provided. Based on a rigorous definition of Verlet radius with respect to temperature and list-updating interval in MD simulation, this paper has successfully developed an estimation formula of the computation time for each MD algorithm calculation so as to find an optimized performance for each algorithm. With the formula proposed here, the best algorithm can be chosen based on different total number of atoms, system average density and system average temperature for the MD simulation. It has been shown that the Verlet Cell-linked List (VCL) algorithm is better than other algorithms for a system with a large number of atoms. Furthermore, a generalized VCL algorithm optimized with a list-updating interval and cell-dividing number is analyzed and has been verified to reduce the computation time by 30∼60% in a MD simulation for a two-dimensional lattice system. Due to similarity, the analysis in this study can be extended to other many-particle systems.     

分子动力学模拟的柔性对接

分子自动对接技术在过去二十年里取得很大发展和成功,但是仍然面对如何处理分子柔性这样一个难题。这篇综述概要介绍分子柔性对接技术的进展并重点介绍分子动力学模拟技术。     

Micromachining process simulation using a continuous cellularautomata method

In this paper, we present results on the development of an anisotropic crystalline etching simulation (AGES) program based on a continuous cellular automata (CA) model. The program provides improved spatial resolution and accuracy compared with the conventional and stochastic CA methods. Implementation of a three-dimensional (3-D) dynamic CA technique provides increased simulation speed and reduced memory requirement. The first AGES software based on common personal computer platforms has been realized for simulation of micromachining processes and visualizing results in 3-D space. The software is uniquely capable of simulating the resultant profile following a series of micromachining steps, including surface passivation, reactive ion etching, as well as wet chemical bulk etching. A novel method for accurately obtaining the etch-rate diagram of anisotropic etching using both experimental and numerical techniques has been developed     

Petascale molecular dynamics simulation using the fast multipole method on K computer

In this paper, we report all-atom simulations of molecular crowding — a result from the full node simulation on the “K computer”, which is a 10-PFLOPS supercomputer in Japan. The capability of this machine enables us to perform simulation of crowded cellular environments, which are more realistic compared to conventional MD simulations where proteins are simulated in isolation. Living cells are “crowded” because macromolecules comprise ∼30% of their molecular weight. Recently, the effects of crowded cellular environments on protein stability have been revealed through in-cell NMR spectroscopy. To measure the performance of the “K computer”, we performed all-atom classical molecular dynamics simulations of two systems: target proteins in a solvent, and target proteins in an environment of molecular crowders that mimic the conditions of a living cell. Using the full system, we achieved 4.4 PFLOPS during a 520 million-atom simulation with cutoff of 28 Å. Furthermore, we discuss the performance and scaling of fast multipole methods for molecular dynamics simulations on the “K computer”, as well as comparisons with Ewald summation methods.     

Experimental study of lubricant depletion in heat assisted magnetic recording: different lubricants on HAMR media

Heat assisted magnetic recording (HAMR) is a promising choice to surpass the super-paramagnetic limit in magnetic recording and to allow continued increase in the areal recording density of hard disk drive. However, lubricant depletion on disk surface might be a problem in HAMR due to high temperature during writing process. In this work, depletion of four types of commercial lubricants on our HAMR media has been evaluated under equivalent HAMR conditions. Linear relationships between the lubricant depletion depth and the logarithm of the equivalent laser heating time have been established for these lubricants. The depletion depths of the lubricants at different laser heating durations are predicated. It is found that the performance of the four evaluated lubricants is acceptable in respect of lubricant depletion depth by hypothesizing the total heating duration for a given point on the media over the lifetime of the drive is 10⁶ ns.     

Huge-scale molecular dynamics simulation of multibubble nuclei

We have developed molecular dynamics codes for a short-range interaction potential that adopt both the flat-MPI and MPI/OpenMP hybrid parallelizations on the basis of a full domain decomposition strategy. Benchmark simulations involving up to 38.4 billion Lennard-Jones particles were performed on Fujitsu PRIMEHPC FX10, consisting of 4800 SPARC64 IXfx 1.848 GHz processors, at the Information Technology Center of the University of Tokyo, and a performance of 193 teraflops was achieved, which corresponds to a 17.0% execution efficiency. Cavitation processes were also simulated on PRIMEHPC FX10 and SGI Altix ICE 8400EX at the Institute of Solid State Physics of the University of Tokyo, which involved 1.45 billion and 22.9 million particles, respectively. Ostwald-like ripening was observed after the multibubble nuclei. Our results demonstrate that direct simulations of multiscale phenomena involving phase transitions from the atomic scale are possible and that the molecular dynamics method is a promising method that can be applied to petascale computers.     

A discussion of the tribological phenomena under minimal load using molecular dynamics

This study explored the tribological phenomena under minimal load using molecular dynamics. In the simulation, <NVT> ensemble average standards and the Condensed-phase Optimized Molecular Potential for Atomistic Simulation Studies potential energy function were used. Regarding materials, this study used (5, 5) carbon nanotubes (CNTs) to move the copper atoms in order to obtain the coefficient of friction (COE) between the single-wall CNTs and copper atoms before analyzing the tribological phenomena under minimal load and changes in friction. This study found that, under an extremely tiny load, the direction of the resistance of the relative sliding is not always in the opposite direction of the movement direction. The influence of environment temperatures on the COE was that the environment temperature increases, the number of fluctuation also increases, but the COE reduces.     

分子动力学并行算法的优化与应用

比较了小型机群中模拟小规模粒子系统采用原子分解和区域分解算法的模拟时间.针对原子分解算法,优化了各进程粒子的受力计算,尤其对严重影响并行效率的全局通信进行了优化,最终使得并行效率提高了30%以上.使用该优化算法对固态氩的导热系数进行了计算,模拟结果与实验值比较吻合,大大缩短了时间.     

Experimental study of lubricant depletion in heat assisted magnetic recording: effects of laser heating duration and temperature

Heat assisted magnetic recording (HAMR) is a promising approach to overcome the superparamagnetic limit in magnetic recording and enable large increases in the storage density of hard disk drives. However, it is expected that HAMR causes lubricant depletion problem on disk surface under the high temperature in the heating assisted writing process. In this study, the effects of laser heating duration and laser heating temperature on lubricant depletion are studied experimentally. It is found that lubricant depletion depth is linearly proportional to the logarithm of laser heating duration. Lubricant depletion can be mitigated effectively by lowering laser heating temperature. Lubricant depletion depth over the life-time of the drive is also predicted.     

基于分子动力学的超晶格薄膜的导热系数影响因素分析

采用非平衡态分子动力学方法模拟了超晶格薄膜的热传导性能,分别计算了组成超晶格薄膜的两种材料的导热系数.从粒子质量、势阱深度、晶格常数的变化等方面讨论了其对薄膜两种材料的切向导热系数的影响.模拟结果表明:不同粒子质量的两种材料组成的超晶格薄膜中,二者的导热性能相互间影响并不大.而势阱深度的改变则两种材料的导热性能会相互产生影响,但在晶格不匹配时,超晶格薄膜的界面出现了应变,两者的导热系数均会产生较大的变化.同时还发现,薄膜的宽度不同时,大质量比的材料的导热系数趋于一致.     

分子动力学模拟蛋白质溶液吸附过程构象的变化

计算机模拟作为一种工具在药物分子设计、蛋白质工程、药物筛选等方面逐渐广泛应用起来。为了从分子水平上理解蛋白质吸附的机理，本文采用了刚体模型对聚十赖氨酸在固体表面吸附进行了分子动力学模拟。采用立方周期性边界条件，模拟在NVT条件下进行，各刚体的起始速度按Maxwell取样。初步研究了模拟过程中蛋白质构象的变化，跟踪了吸附过程中二面角φ和ψ的变化。研究结果表明，吸附过程中蛋白质二级结构发生了变化，C末端二级结构的变化最为明显。     

Membrane protein structure quality in molecular dynamics simulation

Our goal was to assess the relationship between membrane protein quality, output from protein quality checkers and output from molecular dynamics (MD) simulations. Membrane transport proteins are essential for a wide range of cellular processes. Structural features of integral membrane proteins are still under-explored due to experimental limitations in structure determination. Computational techniques can be used to exploit biochemical and medium resolution structural data, as well as sequence homology to known structures, and enable us to explore the structure-function relationships in several transmembrane proteins. The quality of the models produced is vitally important to obtain reliable predictions. An examination of the relationship between model stability in molecular dynamics (MD) simulations derived from RMSD (root mean squared deviation) and structure quality assessment from various protein quality checkers was undertaken. The results were compared to membrane protein structures, solved at various resolution, by either X-ray or electron diffraction techniques. The checking programs could predict the potential success of MD in making functional conclusions. MD stability was shown to be a good indicator for the quality of structures. The quality was also shown to be dependent on the resolution at which the structures were determined.     

A general purpose parallel molecular dynamics simulation program

We present a general purpose parallel molecular dynamics simulation code. The code can handle NVE, NVT, and NPT ensemble molecular dynamics, Langevin dynamics, and dissipative particle dynamics. Long-range interactions are handled by using the smooth particle mesh Ewald method. The implicit solvent model using solvent-accessible surface area was also implemented. Benchmark results using molecular dynamics, Langevin dynamics, and dissipative particle dynamics are given.

Program summary

Title of program:MM_PARCatalogue identifier:ADXP_v1_0Program summary URL:http://cpc.cs.qub.ac.uk/summaries/ADXP_v1_0Program obtainable from: CPC Program Library, Queen's University of Belfast, N. IrelandComputer for which the program is designed and others on which it has been tested:any UNIX machine. The code has been tested on Linux cluster and IBM p690Operating systems or monitors under which the program has been tested:Linux, AIXProgramming language used:CMemory required to execute with typical data:∼60 MB for a system of atoms Has the code been vectorized or parallelized? parallelized with MPI using atom decomposition and domain decompositionNo. of lines in distributed program, including test data, etc.:171 427No. of bytes in distributed program, including test data, etc.:4 558 773Distribution format:tar.gzExternal routines/libraries used:FFTW free software (http://www.fftw.org)Nature of physical problem:Structural, thermodynamic, and dynamical properties of fluids and solids from microscopic scales to mesoscopic scales.Method of solution:Molecular dynamics simulation in NVE, NVT, and NPT ensemble, Langevin dynamics simulation, dissipative particle dynamics simulation.Typical running time:Table below shows the typical run times for the four test programs.

Benchmark results. The values in the parenthesis are the number of processors used
System	Method	Timing for 100 steps in seconds
256 TIP3P	MD	23.8 (1)
64 DMPC + 1645 TIP3P	MD	890 (1)	528 (2)	326 (4)	209 (8)
8 Aβ16-22	LD	1.02 (1)
23760 Groot-Warren particles	DPD	22.16 (1)

Full-size table

     

烷基芳基磺酸盐界面性质的分子动力学模拟

利用分子动力学方法模拟了阴离子表面活性剂在水/辛烷,水/甲苯,水/辛烷、甲苯混合油相时的动力学特点,结果很好地展示了体系静止时清楚的油水界面,并与无表面活性剂的油水界面进行了比较.利用密度剖面图可说明表面活性剂对不同油相的诱导作用.同时,模拟计算出烷基芳基磺酸盐表面活性剂在油水界面的界面张力、均方根偏差、接触面积等,由计算结果可知,油相为甲苯时能很好的降低其界面张力,油相为辛烷时更易于形成清楚的油水界面且界面强度较大.     

多核并行技术在分子动力学模拟中的应用

为了充分利用多核处理器资源,研究了一种用于分子动力学模拟中的多核并行技术。在多核处理器上利用OpenMP技术实现多线程创建与同步、动态设置子线程的调度运行方式以及负载均衡以减少子线程执行等待时间。通过对不同分子体系结构下的动力学模型测试,得出在不同子线程下并行计算的时间,并且得到了良好的性能加速比。实验结果表明,采用OpenMP并行技术可有效地提高电荷求解过程在分子动力学模拟运算中的时间效率,以及多核计算机资源的利用率。     

Parallel implementation of molecular dynamics simulation for short-ranged interaction

A parallel molecular dynamics simulation method, designed for large-scale problems, employing dynamic spatial domain decomposition for short-ranged molecular interactions is proposed. In this parallel cellular molecular dynamics (PCMD) simulation method, the link-cell data structure is used to reduce the searching time required for forming the cut-off neighbor list as well as for domain decomposition, which utilizes the multi-level graph-partitioning technique. A simple threshold scheme (STS), in which workload imbalance is monitored and compared with some threshold value during the runtime, is proposed to decide the proper time for repartitioning the domain. The simulation code is implemented and tested on the memory-distributed parallel machine, e.g., PC-cluster system. Parallel performance is studied using approximately one million L-J atoms in the condensed, vaporized and supercritical states. Results show that fairly good parallel efficiency at 49 processors can be obtained for the condensed and supercritical states (∼60%), while it is comparably lower for the vaporized state (∼40%).