Dynamic load balancing algorithms for replicated data molecular dynamics

Algorithms to enhance parallel performance of molecular dynamics simulations on parallel computers by dynamic load balancing are described. Load balancing is achieved by redistribution of work based on either a history of time spent computing per processor or on the number of pair interactions computed per processor. The two algorithms we detail are designed to yield optimal load balancing on both workstation clusters and parallel supercomputers. We illustrate these methods using a small molecular dynamics kernel developed for the simulation of rigid molecular solvents. In addition, we discuss our observation regarding global communications performance on workstation clusters with a fiber distributed data interface (FDDI) using a high-speed point-to-point switch (Gigaswitch) and the k-ary 3-cube of the Cray T3D. © 1995 by John Wiley & Sons, Inc.     

A highly portable parallel implementation of AMBER4 using the message passing interface standard

We have implemented a portable parallel version of the macromolecular modeling package AMBER4. The message passing paradigm was used. All message passing constructs are compliant with the Message Passing Interface (MPI) standard. The molecular dynamics/minimization module MINMD and the free-energy perturbation module Gibbs have been implemented in parallel on a number of machines, including a Cray T3D, an IBM SP1/SP2, and a collection of networked workstations. In addition, the code has been tested with an MPI implementation from Argonne National Laboratories/Mississippi State University which runs on many parallel machines. The goal of this work is to decrease the amount of time required to perform molecular dynamics simulations. Performance results for a lipid bilayer molecular dynamics simulation on a Cray T3D, an IBM SP1/SP2, and a Cray C90 are compared. © 1995 John Wiley & Sons, Inc.     

Development of a parallel molecular dynamics code on SIMD computers: Algorithm for use of pair list criterion

In recent years several implementations of molecular dynamics (MD) codes have been reported on multiple instruction multiple data (MIMD) machines. However, very few implementations of MD codes on single instruction multiple data (SIMD) machines have been reported. The difficulty in using pair lists of nonbonded interactions is the major problem with MD codes for SIMD machines, such that, generally, the full connectivity computation has been used. We present an algorithm, the global cut-off algorithm (GCA), which permits the use of pair lists on SIMD machines. GCA is based on a probabilistic approach and requires the cut-off condition to be simultaneously verified on all nodes of the machine. The MD code used was taken from the GROMOS package; only the routines involved in the pair lists and in the computation of nonbonded interactions were rewritten for a parallel architecture. The remaining calculations were performed on the host computer. The algorithm has been tested on Quadrics computers for configurations of 32, 128, and 512 processors and for systems of 4000, 8000, 15,000, and 30,000 particles. Quadrics was developed by Istituto Nazionale di Fisica Nucleare (INFN) and marketed by Alenia Spazio. © 1998 John Wiley & Sons, Inc. J Comput Chem 19: 685–694, 1998     

SCF calculations on MIMD type parallel computers

Summary One of the key methods in quantum chemistry, the Hartree-Fock SCF method, is performing poorly on typical vector supercomputers. A significant acceleration of calculations of this type requires the development and implementation of a parallel SCF algorithm. In this paper various parallelization strategies are discussed comparing local and global communication management as well as sequential and distributed Fock-matrix updates. Programs based on these algorithms are bench marked on transputer networks and two IBM MIMD prototypes. The portability of the code is demonstrated with the portation of the initial Helios version to other operating systems like Parallel VM/SP and PARIX. Based on the PVM libraries, a platform-independent version has been developed for heterogeneous workstation clusters as well as for massively parallel computers.     

D-CICADA: A software for conformational PES elucidation on network of workstations

The methodology of conformational potential energy (hyper)surface (PES) elucidation is the subject of this article. The decomposition of the recently developed software CICADA and its implementation in the distributed environment using PVM (parallel virtual machine) is presented. CICADA has been chosen for the parallelization because of its ability to elucidate systematically the low-energy areas of PES in polynomial time. This makes the method applicable on larger systems which are beyond the scope of the grid search. To show the level of parallelization, conformational PES of two molecules, cyclohexane and terminally blocked alanine, have been studied by the distributed version, D-CICADA, and results have been compared to those of the sequential version. D-CICADA was tested on several virtual machines composed of DEC and Sun workstations. The timing shows good efficiency for both the decomposition of the original algorithm and the PVM environment. © 1994 by John Wiley & Sons, Inc.     

Quantum Monte Carlo for large chemical systems: Implementing efficient strategies for petascale platforms and beyond

Various strategies to implement efficiently quantum Monte Carlo (QMC) simulations for large chemical systems are presented. These include: (i) the introduction of an efficient algorithm to calculate the computationally expensive Slater matrices. This novel scheme is based on the use of the highly localized character of atomic Gaussian basis functions (not the molecular orbitals as usually done), (ii) the possibility of keeping the memory footprint minimal, (iii) the important enhancement of single‐core performance when efficient optimization tools are used, and (iv) the definition of a universal, dynamic, fault‐tolerant, and load‐balanced framework adapted to all kinds of computational platforms (massively parallel machines, clusters, or distributed grids). These strategies have been implemented in the QMC=Chem code developed at Toulouse and illustrated with numerical applications on small peptides of increasing sizes (158, 434, 1056, and 1731 electrons). Using 10–80 k computing cores of the Curie machine (GENCI‐TGCC‐CEA, France), QMC=Chem has been shown to be capable of running at the petascale level, thus demonstrating that for this machine a large part of the peak performance can be achieved. Implementation of large‐scale QMC simulations for future exascale platforms with a comparable level of efficiency is expected to be feasible. © 2013 Wiley Periodicals, Inc.     

PEACH 4 with ABINIT-MP: a general platform for classical and quantum simulations of biological molecules

A program package for molecular simulations of biological molecules was developed. The package, "PEACH version 4 with ABINIT-MP version 20021029," was constructed by incorporating ABINIT-MP, a program for the fragment molecular orbital (FMO) method [Chem.Phys. Lett. 313 (1999) 701], into PEACH, a program package for classical molecular dynamics simulations (MD). A few capabilities of the package were demonstrated. First, high parallel efficiency of FMO was demonstrated in a single point calculation of a protein. Second,FMO-MD simulations [Chem. Phys. Lett. 372 (2003) 342] of a peptide were performed with and without explicit solvent, and the simulations showed the influence of the solvent on the electronic state of the peptide.     

Coarse-grained molecular dynamics modeling of strongly associating fluids: thermodynamics, liquid structure, and dynamics of symmetric binary mixture fluids

Symmetric binary mixtures capable of strong association via a highly directional and saturable specific interaction between unlike molecules are investigated by canonical molecular dynamics simulations. The specific interaction of the molecules is defined in a new coarse-grained pair potential that can be applied in continuous molecular dynamics as well as in Monte Carlo simulations. The thermodynamic, structural, and dynamic properties of the associating mixture fluids are investigated as a function of density, temperature, and association strength of the specific interaction. Detailed analysis of the simulation data confirms a two-stage mechanism in the formation of specific bonds with increasing interaction strength, including a fast dimerization process and a subsequent stage of perfecting the bonds. A large heat capacity peak is found during the formation or breaking of the bonds, reflecting the large energy fluctuation introduced by the strong association. The fractions of nonbonded molecules obtained from the simulations as a function of density, temperature, and interaction strength are in excellent agreement with the predictions of Wertheim's thermodynamic perturbation theory. The translational and rotational dynamics of the Tmer mixture are effectively retarded with increasing association strength and are analyzed in terms of autocorrelation functions and a non-Gaussian parameter for the translational dynamics. The lifetimes of molecules in bonded and nonbonded states provide detailed information about the transformation of molecules between the bonded state and the nonbonded state. Finally, simulation sampling problems inherent to strongly interacting systems are easily overcome using the parallel tempering simulation technique. This latter result confirms that with the new continuous coarse-grained simulation potential we have a versatile and flexible interaction potential that can be used with many available molecular dynamics and Monte Carlo algorithms under various ensembles.     

Improving the performance of molecular dynamics simulations on parallel clusters

In this article a procedure is derived to obtain a performance gain for molecular dynamics (MD) simulations on existing parallel clusters. Parallel clusters use a wide array of interconnection technologies to connect multiple processors together, often at different speeds, such as multiple processor computers and networking. It is demonstrated how to configure existing programs for MD simulations to efficiently handle collective communication on parallel clusters with processor interconnections of different speeds.     

Conformational and energetic effects of truncating nonbonded interactions in an aqueous protein dynamics simulation

In a previous aqueous protein dynamics study, we compared the rms deviation relative to the crystal structure for distance-dependent and constant dielectric models with and without a nonbonded cutoff. The structures obtained from a constant dielectric simulation with a cutoff were substantially different from the structures obtained from a distance-dependent dielectric simulation, with and without cutoff, and a constant dielectric model without a cutoff. In fact, structures from the distance-dependent dielectric simulations were insensitive to the nonbonded cutoff and in good agreement with the structures generated from the constant dielectric simulation without a cutoff. In addition, the solute-solvent temperature differential and solvent evaporation artifacts, characteristic of the constant dielectric simulation with a cutoff, were not present for the distance-dependent dielectric simulations. In this current work, we explore whether this dielectric-dependent cutoff-sensitive behavior for a constant dielectric model arises from the discontinuities in the forces at the nonbonded cutoff or from neglecting the structure-stabilizing interactions beyond the nonbonded cutoff. We also examine the origin of the dielectric-dependent artifacts, and its potential influence on the structural disparity. Several protocols for protein dynamics simulations are compared using both constant and distance-dependent dielectric models, including implementation of a switching function and a nonbonded cutoff and two different temperature coupling algorithms. We show that the distance-dependent dielectric model conserves energy in the SPASMS molecular mechanics and dynamics software for the time steps and nonbonded cutoffs commonly used in macromolecule simulations. Although the switching function simulation also conserved energy over a range of commonly used cutoffs, the constant dielectric model with a switching function yielded conformational results more similar to a constant dielectric simulation without a switching function than to a constant dielectric model without a nonbonded cutoff. Therefore, the conformational disparity between the dielectric models arises from neglecting important structure-stabilizing interactions beyond the cutoff, rather than differences in energy conservation. © 1993 John Wiley & Sons, Inc.     

Accelerating molecular modeling applications with graphics processors

Molecular mechanics simulations offer a computational approach to study the behavior of biomolecules at atomic detail, but such simulations are limited in size and timescale by the available computing resources. State-of-the-art graphics processing units (GPUs) can perform over 500 billion arithmetic operations per second, a tremendous computational resource that can now be utilized for general purpose computing as a result of recent advances in GPU hardware and software architecture. In this article, an overview of recent advances in programmable GPUs is presented, with an emphasis on their application to molecular mechanics simulations and the programming techniques required to obtain optimal performance in these cases. We demonstrate the use of GPUs for the calculation of long-range electrostatics and nonbonded forces for molecular dynamics simulations, where GPU-based calculations are typically 10-100 times faster than heavily optimized CPU-based implementations. The application of GPU acceleration to biomolecular simulation is also demonstrated through the use of GPU-accelerated Coulomb-based ion placement and calculation of time-averaged potentials from molecular dynamics trajectories. A novel approximation to Coulomb potential calculation, the multilevel summation method, is introduced and compared with direct Coulomb summation. In light of the performance obtained for this set of calculations, future applications of graphics processors to molecular dynamics simulations are discussed.     

Performance enhancements for GROMACS nonbonded interactions on BlueGene

Several improvements to the previously optimized GROMACS BlueGene inner loops that evaluate nonbonded interactions in molecular dynamics simulations are presented. The new improvements yielded an 11% decrease in running time for both PME and other kinds of GROMACS simulations that use nonbonded table look-ups. Some other GROMACS simulations will show a small gain.     

Application of MDGRAPE-3, a special purpose board for molecular dynamics simulations, to periodic biomolecular systems

We describe the application of a special purpose board for molecular dynamics simulations, named MDGRAPE-3, to the problem of simulating periodic bio-molecular systems. MDGRAPE-3 is the latest board in a series of hardware accelerators designed to calculate the nonbonding long-range interactions much more rapidly than normal processors. So far, MDGRAPEs were mainly applied to isolated systems, where very many nonbonded interactions were calculated without any distance cutoff. However, in order to regulate the density and pressure during simulations of membrane embedded protein systems, one has to evaluate interactions under periodic boundary conditions. For this purpose, we implemented the Particle-Mesh Ewald (PME) method, and its approximation with distance cutoffs and charge neutrality as proposed by Wolf et al., using MDGRAPE-3. When the two methods were applied to simulations of two periodic biomolecular systems, a single MDGRAPE-3 achieved 30-40 times faster computation times than a single conventional processor did in the both cases. Both methods are shown to have the same molecular structures and dynamics of the systems.     

SIMONA 1.0: An efficient and versatile framework for stochastic simulations of molecular and nanoscale systems

Molecular simulation methods have increasingly contributed to our understanding of molecular and nanoscale systems. However, the family of Monte Carlo techniques has taken a backseat to molecular dynamics based methods, which is also reflected in the number of available simulation packages. Here, we report the development of a generic, versatile simulation package for stochastic simulations and demonstrate its application to protein conformational change, protein–protein association, small-molecule protein docking, and simulation of the growth of nanoscale clusters of organic molecules. Simulation of molecular and nanoscale systems (SIMONA) is easy to use for standard simulations via a graphical user interface and highly parallel both via MPI and the use of graphical processors. It is also extendable to many additional simulations types. Being freely available to academic users, we hope it will enable a large community of researchers in the life- and materials-sciences to use and extend SIMONA in the future. SIMONA is available for download under http://int.kit.edu/nanosim/simona . © 2012 Wiley Periodicals, Inc.     

Large‐scale asynchronous and distributed multidimensional replica exchange molecular simulations and efficiency analysis

We describe methods to perform replica exchange molecular dynamics (REMD) simulations asynchronously (ASyncRE). The methods are designed to facilitate large scale REMD simulations on grid computing networks consisting of heterogeneous and distributed computing environments as well as on homogeneous high‐performance clusters. We have implemented these methods on NSF (National Science Foundation) XSEDE (Extreme Science and Engineering Discovery Environment) clusters and BOINC (Berkeley Open Infrastructure for Network Computing) distributed computing networks at Temple University and Brooklyn College at CUNY (the City University of New York). They are also being implemented on the IBM World Community Grid. To illustrate the methods, we have performed extensive (more than 60 ms in aggregate) simulations for the beta‐cyclodextrin‐heptanoate host‐guest system in the context of one‐ and two‐dimensional ASyncRE, and we used the results to estimate absolute binding free energies using the binding energy distribution analysis method. We propose ways to improve the efficiency of REMD simulations: these include increasing the number of exchanges attempted after a specified molecular dynamics (MD) period up to the fast exchange limit and/or adjusting the MD period to allow sufficient internal relaxation within each thermodynamic state. Although ASyncRE simulations generally require long MD periods (>picoseconds) per replica exchange cycle to minimize the overhead imposed by heterogeneous computing networks, we found that it is possible to reach an efficiency similar to conventional synchronous REMD, by optimizing the combination of the MD period and the number of exchanges attempted per cycle. © 2015 Wiley Periodicals, Inc.     

Parallel direct SCF and gradient program for workstation clusters

A parallel direct SCF and gradient program for workstation clusters has been implemented on the basis of the ab initio program package TURBOMOLE. Applications on large molecular systems monitor an appreciable speedup in residence time. © John Wiley & Sons, Inc.     

Simulation of metal-ligand self-assembly into spherical complex m(6)l(8)

Molecular dynamics simulations were performed to study the self-assembly of a spherical complex through metal-ligand coordination interactions. M(6)L(8), a nanosphere with six palladium ions and eight pyridine-capped tridentate ligands, was selected as a target system. We successfully observed the spontaneous formation of spherical shaped M(6)L(8) cages over the course of our simulations, starting from random initial placement of the metals and ligands. To simulate spontaneous coordination bond formations and breaks, the cationic dummy atom method was employed to model nonbonded metal-ligand interactions. A coarse-grained solvent model was used to fill the gap between the time scale of the supramolecular self-assembly and that accessible by common molecular dynamics simulation. The simulated formation process occurred in the distinct three-stage (assembly, evolution, fixation) process that is well correlated with the experimental results. We found that the difference of the lifetime (or the ligand exchange rate) between the smaller-sized incomplete clusters and the completed M(6)L(8) nanospheres is crucially important in their supramolecular self-assembly.     

The COMPASS force field: parameterization and validation for phosphazenes

A new, condensed-phase optimised ab-initio force field, COMPASS, has been developed recently. In this paper, the validation of COMPASS for phosphazenes is presented. The functional forms of this force field are of the consistent force field (CFF) type. Charges and bonded terms were derived from HF/6–31G1 calculations, while the nonbonded parameters (L-J 9-6 vdW potential) were initially transferred from the polymer consistent force field, pcff, and optimised using MD simulations of condensed-phase properties. As a validation of COMPASS, molecular mechanics calculations and molecular dynamics simulations have been made on a number of isolated molecules, liquids, and crystals. The calculated molecular structure, vibration frequencies, conformational properties for isolated molecules, crystal cell parameters and density, liquid density, and heat of evaporation agreed favourably with most experimental data. The special conformational properties of the tetracyclophosphazenes, (NPCI₂)₄ and (NPF₂)₄, in the solid state are discussed based on molecular mechanics and CASTEP ab-initio calculations. The effect of nonbonded cutoff distance and different algorithms for pressure control in NPT simulation was also investigated. Finally, molecular dynamics using the COMPASS force field was used to predict properties of three isomers of high-molecular-weight amorphous poly(dibutoxyphosphazenes). In this case, excellent agreement was achieved between densities and glass transition temperatures obtained from dynamics and experimental data.     

Molecular dynamics on a distributed-memory multiprocessor

Dynamics simulations of molecular systems are notoriously computationally intensive. Using parallel computers for these simulations is important for reducing their turnaround time. In this article we describe a parallelization of the simulation program CHARMM for the Intel iPSC/860, a distributed memory multiprocessor. In the parallelization, the computational work is partitioned among the processors for core calculations including the calculation of forces, the integration of equations of motion, the correction of atomic coordinates by constraint, and the generation and update of data structures used to compute nonbonded interactions. Processors coordinate their activity using synchronous communication to exchange data values. Key data structures used are partitioned among the processors in nearly equal pieces, reducing the memory requirement per node and making it possible to simulate larger molecular systems. We examine the effectiveness of the parallelization in the context of a case study of a realistic molecular system. While effective speedup was achieved for many of the dynamics calculations, other calculations fared less well due to growing communication costs for exchanging data among processors. The strategies we used are applicable to parallelization of similar molecular mechanics and dynamics programs for distributed memory multiprocessors. © 1992 by John Wiley & Sons, Inc.