Massively-parallel Supercomputers Research Articles

A New Rusanov-Type Solver with a Local Linear Solution Reconstruction for Numerical Modeling of White Dwarf Mergers by Means Massive Parallel Supercomputers

The results of numerical modeling of white dwarf mergers on massive parallel supercomputers using a AVX-512 technique are presented. A hydrodynamic model of white dwarfs closed by a star equation of state and supplemented by a Poisson equation for the gravitational potential is constructed. This paper presents a modification based on a local linear reconstruction of the solution of the Rusanov scheme for the hydrodynamic equations. This reconstruction makes it possible to considerably decrease the numerical dissipation of the scheme for weak shock waves without any external piecewise polynomial reconstruction. The scheme is efficient for unstructured grids, when it is difficult to construct a piecewise polynomial solution, and also in parallel implementations of structured nested or adaptive grids, when the costs of interprocess interactions increase significantly. As input data, piecewise constant values of the physical variables in the left and right cells of a discontinuity are used. The smoothness of the solution is measured by the discrepancy between the maximum left and right eigenvalues. This discrepancy is used for a local piecewise polynomial reconstruction in the left and right cells. Then the solutions are integrated along the characteristics taking into account the piecewise linear representation of the physical variables. A performance of 234 gigaflops and 33-fold speedup are obtained on two Intel Skylake processors on the cluster NKS-1P of the Siberian Supercomputer Center ICM & MG SB RAS.

A New Parallel Intel Xeon Phi Hydrodynamics Code for Massively Parallel Supercomputers

In this paper, a new hydrodynamics code called gooPhi to simulate astrophysical flows on modern Intel Xeon Phi processors with KNL architecture is presented. A new vector numerical method implemented in the form of a program code for massively parallel architectures is proposed. A detailed description is given and a parallel implementation of the code is made. A performance of 173 gigaflops and 48 speedup are obtained on a single Intel Xeon Phi processor. A 97 per cent scalability is reached with 16 processors.

Near-surface coherent structures explored by large eddy simulation of entire tropical cyclones

Taking advantage of the huge computational power of a massive parallel supercomputer (K-supercomputer), this study conducts large eddy simulations of entire tropical cyclones by employing a numerical weather prediction model, and explores near-surface coherent structures. The maximum of the near-surface wind changes little from that simulated based on coarse-resolution runs. Three kinds of coherent structures appeared inside the boundary layer. The first is a Type-A roll, which is caused by an inflection-point instability of the radial flow and prevails outside the radius of maximum wind. The second is a Type-B roll that also appears to be caused by an inflection-point instability but of both radial and tangential winds. Its roll axis is almost orthogonal to the Type-A roll. The third is a Type-C roll, which occurs inside the radius of maximum wind and only near the surface. It transports horizontal momentum in an up-gradient sense and causes the largest gusts.

Performance evaluation of a throughput-aware framework for ensemble data assimilation: the case of NICAM-LETKF

Abstract. In this paper, we propose the design and implementation of an ensemble data assimilation (DA) framework for weather prediction at a high resolution and with a large ensemble size. We consider the deployment of this framework on the data throughput of file input/output (I/O) and multi-node communication. As an instance of the application of the proposed framework, a local ensemble transform Kalman filter (LETKF) was used with a Non-hydrostatic Icosahedral Atmospheric Model (NICAM) for the DA system. Benchmark tests were performed using the K computer, a massive parallel supercomputer with distributed file systems. The results showed an improvement in total time required for the workflow as well as satisfactory scalability of up to 10 K nodes (80 K cores). With regard to high-performance computing systems, where data throughput performance increases at a slower rate than computational performance, our new framework for ensemble DA systems promises drastic reduction of total execution time.

Modelling of Complexities in PA of Laplace Equation

Dominant trends in parallel and distributed computing (parallel computers, parallel algorithms) aim to connect a number of powerful workstations (WS) based on single personal computers (PC) or symmetrical multiprocessor system (SMP) to solve complex problems in a parallel way. Typical representatives of these trends are network of workstation (NOW) and Grid as a high integrated network of NOW modules. In relation to it there is a bad need of unified modelling in traditionally evolved parallel computing and distributed computing as two separate research disciplines. Likewise current trends in high performance computing (HPC) aim to use NOW modules based on SMP parallel computers as cheaper alternatives to the traditionally used massive parallel multiprocessors or supercomputers and to profit from unifying modelling in parallel and distributed computing too. To exploit the parallel processing capability of such cluster, any application program must be paralleled. A way how to do it (decomposition model) belongs to a most important step in developing process of effective parallel algorithm. This article at first discusses the development of effective (optimised) iterative parallel algorithms (IPA) to solve partial differential equations (PDE). By carefully choosing an illustrative typical problem of PDE (Laplace equation) the proposed research has demonstrated the most important complexities in the process of complete modelling of substantial complexities and their optimisation in order to develop an effective PA. Such approach could be used as illustrative example for a complete modelling and effective solution of even other more complex PDE´s in a parallel way.

Sparc64 XIfx: Fujitsu's Next-Generation Processor for High-Performance Computing

Sparc64 XIfx, the latest high-performance computing processor, includes a 34-core processor and achieves 1.1 teraflops of peak performance. This chip is designed for massive parallel supercomputer systems that realize high performance for a wide range of real applications. As groundwork for exascale supercomputing, it has made major technological changes from the previous Sparc64 processor in the instruction set architecture, microarchitecture, memory modules, and embedded interconnect.

Large-Scale Electronic-Structure Calculations in the Real-Space Scheme: Bilayer Graphene and Silicene

ABSTRACTWe have developed our original DFT (density-functional theory) calculation code “RSDFT” using real-space schemes. The code is FFT-free, leading to high-performance computing in massively-parallel supercomputers. The code allows us to deal with systems including huge numbers of atoms from first-principles. We have applied our schemes to clarify atomic and electronic structures of two relevant nano-scale systems: twisted bilayer graphene and silicene on Ag substrate.

Quantum chemistry reloaded

When Per-Olov Löwdin founded the International Journal of Quantum Chemistry in 1967 quantum chemistry was entering the modern era, beginning to explore the potential of computers as well as extending its realm beyond the study of atoms and molecules to small systems of biological interest. Löwdin's vision for the discipline is reflected in his editorial for the inaugural issue of the journal.[1] Coming into existence through the interactions of physics, chemistry, applied mathematics, and what we now would call computer science, modern quantum chemistry has been a naturally interdisciplinary field from the outset, and IJQC's role is to report and facilitate this interplay of theory, computational algorithms, and applications to atoms, molecules, crystals, and biology, in close connection with experiments. In the nearly half a century since that historic editorial, quantum chemistry has experienced continuous evolution, harvesting the most exciting developments from the parent fields. The rise of materials science research, for example, has brought the challenge of extending the predicting power of quantum chemical calculations to ever larger and more complex systems, whereas the study of chemical reactivity and biological processes requires the accurate simulation of atomic dynamics by first-principles. In the same way, the introduction and optimization of new, powerful, and sophisticated spectroscopic techniques calls for increasingly accurate theoretical approaches for their interpretation, and computational algorithms and software must be developed to adapt to new, ever more powerful computer architectures: from massive parallel supercomputers with hundreds of thousands of processors to systems based on graphical chips. Finally, the application of quantum information concepts to chemistry[2] could drive a new beginning for quantum chemistry, similar to that experienced with the advent of electronic computers 50 years ago. Mirroring quantum chemistry's path, IJQC must also evolve to continue to serve its original mission of fostering the exchange of ideas between researchers from all aspects of the discipline. A journal at the service of today's quantum chemistry community demands modern publishing processes, and we have been working hard to improve our manuscript handling and production procedures to ensure that your research is available to the community as soon as possible. These efforts have resulted in the reduction of the acceptance to online publication time to a very competitive 15 working days on average. Our team of professional in-house editors is focused on delivering rapid and rigorous peer-review, keeping well in sight the journal's role in providing interdisciplinary communication and scientific insight, and not just simple numerical results. With our new Reviews and Tutorials series,[3] we plan to echo the versatility and variety of modern quantum chemistry, and by making these articles free to access we hope to give everybody an opportunity to discover new aspects of this stimulating discipline. Moreover, our most exciting articles are now highlighted through the very popular ChemistryViews.org and MaterialsViews.com, Wiley's chemistry and materials science news sites, to guarantee them the great exposure they deserve even beyond the theoretical chemistry community. Finally, we hope that you have already appreciated IJQC's new look, which premiered last year with issue 17/2012. In addition to full-page cover images, highlighting some of the most interesting articles of every issue, we have introduced a graphical table of contents to guide you quickly to the papers of your interest in the journal, both online and in print. We look forward to what the future holds in store for quantum chemistry. Whatever this may be, you will read it on the pages of IJQC. Enjoy the “reloaded” International Journal of Quantum Chemistry !

Porting and scaling OpenACC applications on massively-parallel, GPU-accelerated supercomputers

An increasing number of massively-parallel supercomputers are based on heterogeneous node architectures combining traditional, powerful multicore CPUs with energy-efficient GPU accelerators. Such systems offer high computational performance with modest power consumption. As the industry trend of closer integration of CPU and GPU silicon continues, these architectures are a possible template for future exascale systems. Given the longevity of large-scale parallel HPC applications, it is important that there is a mechanism for easy migration to such hybrid systems. The OpenACC programming model offers a directive-based method for porting existing codes to run on hybrid architectures. In this paper, we describe our experiences in porting the Himeno benchmark to run on the Cray XK6 hybrid supercomputer. We describe the OpenACC programming model and the changes needed in the code, both to port the functionality and to tune the performance. Despite the additional PCIe-related overheads when transferring data from one GPU to another over the Cray Gemini interconnect, we find the application gives very good performance and scales well. Of particular interest is the facility to launch OpenACC kernels and data transfers asynchronously, which speeds the Himeno benchmark by 5%–10%. Comparing performance with an optimised code on a similar CPU-based system (using 32 threads per node), we find the OpenACC GPU version to be just under twice the speed in a node-for-node comparison. This speed-up is limited by the computational simplicity of the Himeno benchmark and is likely to be greater for more complicated applications.

Viscous flow simulation in a stenosis model using discrete particle dynamics: a comparison between DPD and CFD

Flow and stresses induced by blood flow acting on the blood cellular constituents can be represented to a certain extent by a continuum mechanics approach down to the order of theμm level. However, the molecular effects of, e.g., adhesion/aggregation bonds of blood clotting can be on the order of nm. The coupling of the disparate length and timescales between such molecular levels and macroscopic transport represents a major computational challenge. To address this challenge, a multiscale numerical approach based on discrete particle dynamics (DPD) methodology derived from molecular dynamics (MD) principles is proposed. The feasibility of the approach was firstly tested for its ability to simulate viscous flow conditions. Simulations were conducted in low Reynolds numbers flows (Re = 25-33) through constricted tubes representing blood vessels with various degrees of stenosis. Multiple discrete particles interacting with each other were simulated, with 1.24-1.36 million particles representing the flow domain and 0.4 million particles representing the vessel wall. The computation was carried out on the massive parallel supercomputer NY BlueGene/L employing NAMD-a parallel MD package for high performance computing (HPC). Typical recirculation zones were formed distal to the stenoses. The velocity profiles and recirculation zones were in excellent agreement with computational fluid dynamics (CFD) 3D Navier-Stokes viscous fluid flow simulations and with classic numerical and experimental results by YC Fung in constricted tubes. This feasibility analysis demonstrates the potential of a methodology that widely departs from a continuum approach to simulate multiscale phenomena such as flow induced blood clotting.

Development and Applications of a Novel QM/MM Hybrid Molecular Dynamics Calculation System on Highly Parallel Supercomputer Systems

Quantum mechanical (QM) calculation is now an important tool for investigations of functional mechanisms of biological macromolecules based on their three dimensional and electronic structures. However, the system size which QM calculations can treat is usually up to a few hundred atoms, whereas those of most biological systems of interests are in the range of 1,000 to 1,000,000 including surrounding solvent water. To overcome these difficulties, quantum mechanics/molecular mechanics (QM/MM) calculation has been used as an efficient method, in which the system is divided into QM and MM regions; the active sites to be investigated are assigned as the QM regions, which are described quantum mechanically, and the other regions of the macromolecular systems are assigned as the MM regions, which are described molecular mechanically. To date, many works for developments of efficient/accurate algorithms and their implementations/applications have been performed for QM/MM calculations. In this study, we have developed an interface program to connect conventional but highly-parallelized QM and MM calculation engines running on massively-parallel supercomputers with more than thousands of CPUs. We connected AMBER and GAMESS for molecular dynamics (MD) and QM calculation engines, respectively, which enabled us to perform high-performance QM/MM hybrid MD simulations. Actually, we have evaluated the accuracy and performance of the present system on our supercomputers, the PACS-CS system (14.34 TFlops) and the T2K Tsukuba system (95.39 TFlops) in the Center for Computational Sciences, University of Tsukuba, by comparing the calculated results with experimental data with respect to a metalloprotein (the Cu-bound active center is assigned as the QM region). Furthermore, we have applied it for investigations of environmental effects on the electronic structure of a protein-DNA complex and the reaction mechanisms of cytochrome c oxidase.

Quantum mechanical ab initio simulation of the electron screening effect in metal deuteride crystals

In antecedent experiments the electron screening energies of the d+d reactions in metallic environments have been determined to be enhanced by an order of magnitude in comparison to the case of gaseous deuterium targets. The analytical models describing averaged material properties have not been able to explain the experimental results so far. Therefore, a first effort has been undertaken to simulate the dynamics of reacting deuterons in a metallic lattice by means of an ab-initio Hartree-Fock calculation of the total electrostatic force between the lattice and the successively approaching deuterons via path integration. The calculations have been performed for Li and Ta, clearly showing a migration of electrons from host metallic to the deuterium atoms. However, in order to avoid more of the necessary simplifications in the model the utilization of a massive parallel supercomputer would be required.

Particle-in-Cell Simulations with Kinetic Electrons

A new scheme, based on an exact separation between adiabatic and nonadiabatic electron responses, for particle-in-cell (PIC) simulations of drift-type modes is presented. The (linear and nonlinear) elliptic equations for the scalar fields are solved using a multigrid solver. The new scheme yields linear growth rates in excellent agreement with theory and it is shown to conserve energy well into the nonlinear regime. It is also demonstrated that simulations with few electrons are reliable and accurate, suggesting that large-scale, PIC simulations with electron dynamics in toroidal geometry (e.g., tokamaks and stellarators plasmas) are within reach of present-day massively-parallel supercomputers.

Accurate and efficient non-adiabatic quantum molecular dynamics approach for laser–matter interactions

A non-adiabatic quantum molecular dynamics approach for treating the interaction of matter with intense, short-duration laser pulses is developed. This approach, which is parallelized to run on massively-parallel supercomputers, is shown to be both accurate and efficient. Illustrative results are presented for harmonic generation occurring in diatomic molecules using linearly polarized laser pulses.

Parallel 3-D viscoelastic finite difference seismic modelling

Computational power has advanced to a state where we can begin to perform wavefield simulations for realistic (complex) 3-D earth models at frequencies of interest to both seismologists and engineers. On serial platforms, however, 3-D calculations are still limited to small grid sizes and short seismic wave traveltimes. To make use of the efficiency of network computers a parallel 3-D viscoelastic finite difference (FD) code is implemented which allows to distribute the work on several PCs or workstations connected via standard ethernet in an in-house network. By using the portable message passing interface standard (MPI) for the communication between processors, running times can be reduced and grid sizes can be increased significantly. Furthermore, the code shows good performance on massive parallel supercomputers which makes the computation of very large grids feasible. This implementation greatly expands the applicability of the 3-D elastic/viscoelastic finite-difference modelling technique by providing an efficient, portable and practical C-program.

Molecular dynamics simulation of realistic systems

Molecular dynamics simulation is a powerful tool for understanding and predicting thermophysical properties in systems both at equilibrium and away from equilibrium and composed of molecules ranging in complexity from spherically symmetric monatomic molecules to polymers. The widespread utilization of low-cost high performance engineering workstations has permitted the application of molecular dynamics and Monte Carlo simulation techniques to proliferate at an astounding rate. High-performance massively-parallel supercomputers (such as the Intel Paragon and IBM SP/2) permit the simulation of systems with an unprecedented degree of complexity. As computing costs continue to decline, molecular simulation of systems modeled to a high degree of accuracy promises to become an equal partner with experiment for the prediction of thermophysical properties for industrial purposes. Several examples of the use of molecular dynamics simulation to predict the properties of highly realistic systems are described.

Using massively-parallel supercomputers to model stochastic spatial predator-prey systems

Massively parallel supercomputers open the door to extensive study of ecological simulations that in the past have required prohibitive amounts of computation time. An implementation of a stochastic spatial predator-prey model on such a supercomputer is described. A model that can produce Volterra type oscillations, is extended to study the effectiveness of preferential migration rules for predators and prey. A comparison of various predator migration rules, and results of an initial investigation into the effects produced by mutation of predator migration rates are given.