Search results for "Computational Science"
showing 10 items of 124 documents
COMPARISON OF CPML IMPLEMENTATIONS FOR THE GPU-ACCELERATED FDTD SOLVER
2011
Three distinctively difierent implementations of convolu- tional perfectly matched layer for the FDTD method on CUDA enabled graphics processing units are presented. All implementations store ad- ditional variables only inside the convolutional perfectly matched lay- ers, and the computational speeds scale according to the thickness of these layers. The merits of the difierent approaches are discussed, and a comparison of computational performance is made using complex real-life benchmarks.
Computational issues of an electromagnetics transient meshless method
2019
In this paper we refer to the computational issues in solving Maxwell’ s curl equations without using any connectivity among the points in which the problem domain is discretized. The adopted procedure is able to approximate the electric and magnetic vector fields making use of the derivatives of a kernel function at points arranged in the computational domain. In order to improve the numerical accuracy, dealing with irregular data distribution or data located near the boundary, a suitable strategy is considered. The computational core of the overall process requires elementary linear algebra operations. In the paper the method is presented and the discussion is revolved to the computationa…
A Geant4 simulation package for the sage spectrometer
2012
International audience; A comprehensive Geant4 simulation was built for the SAGE spectrometer. The simulation package includes the silicon and germanium detectors, the mechanical structure and the electromagnetic fields present in SAGE. This simulation can be used for making predictions through simulating experiments and for comparing simulated and experimental data to better understand the underlying physics.
New QM/MM implementation of the DFTB3 method in the gromacs package.
2015
The approximate density-functional tight-binding theory method DFTB3 has been implemented in the quantum mechanics/molecular mechanics (QM/MM) framework of the Gromacs molecular simulation package. We show that the efficient smooth particle–mesh Ewald implementation of Gromacs extends to the calculation of QM/MM electrostatic interactions. Further, we make use of the various free-energy functionalities provided by Gromacs and the PLUMED plugin. We exploit the versatility and performance of the current framework in three typical applications of QM/MM methods to solve biophysical problems: (i) ultrafast proton transfer in malonaldehyde, (ii) conformation of the alanine dipeptide, and (iii) el…
A framework for vertex reconstruction in the ATLAS experiment at LHC
2010
In anticipation of the first LHC data to come, a considerable effort has been devoted to ensure the efficient reconstruction of vertices in the ATLAS detector. This includes the reconstruction of photon conversions, long lived particles, secondary vertices in jets as well as finding and fitting of primary vertices. The implementation of the corresponding algorithms requires a modular design based on the use of abstract interfaces and a common Event Data Model. An enhanced software framework addressing various physics applications of vertex reconstruction has been developed in the ATLAS experiment. Presented in this paper are the general principles of this framework. A particular emphasis is…
GPU accelerated Monte Carlo simulations of lattice spin models
2011
We consider Monte Carlo simulations of classical spin models of statistical mechanics using the massively parallel architecture provided by graphics processing units (GPUs). We discuss simulations of models with discrete and continuous variables, and using an array of algorithms ranging from single-spin flip Metropolis updates over cluster algorithms to multicanonical and Wang-Landau techniques to judge the scope and limitations of GPU accelerated computation in this field. For most simulations discussed, we find significant speed-ups by two to three orders of magnitude as compared to single-threaded CPU implementations.
The on-line coupled atmospheric chemistry model system MECO(n) – Part 5: Expanding the Multi-Model-Driver (MMD v2.0) for 2-way data exchange includin…
2018
Abstract. As part of the Modular Earth Submodel System (MESSy), the Multi-Model-Driver (MMD v1.0) was developed to couple online the regional Consortium for Small-scale Modeling (COSMO) model into a driving model, which can be either the regional COSMO model or the global European Centre Hamburg general circulation model (ECHAM) (see Part 2 of the model documentation). The coupled system is called MECO(n), i.e., MESSy-fied ECHAM and COSMO models nested n times. In this article, which is part of the model documentation of the MECO(n) system, the second generation of MMD is introduced. MMD comprises the message-passing infrastructure required for the parallel execution (multiple programme mul…
Visualization in the integrated SimPhoNy multiscale simulation framework
2018
Abstract We describe three distinct approaches to visualization for multiscale materials modelling research. These have been developed with the framework of the SimPhoNy FP7 EU-project, and complement each other in their requirements and possibilities. All have been integrated via wrappers to one or more of the simulation approaches within the SimPhoNy project. In this manuscript we describe and contrast their features. Together they cover visualization needs from electronic to macroscopic scales and are suited to simulations made on personal computers, workstations or advanced High Performance parallel computers. Examples as well as recommendations for future calculations are presented.
Lattice Boltzmann Simulations at Petascale on Multi-GPU Systems with Asynchronous Data Transfer and Strictly Enforced Memory Read Alignment
2015
The lattice Boltzmann method is a well-established numerical approach for complex fluid flow simulations. Recently general-purpose graphics processing units have become accessible as high-performance computing resources at large-scale. We report on implementing a lattice Boltzmann solver for multi-GPU systems that achieves 0.69 PFLOPS performance on 16384 GPUs. In addition to optimizing the data layout on the GPUs and eliminating the halo sites, we make use of the possibility to overlap data transfer between the host CPU and the device GPU with computing on the GPU. We simulate flow in porous media and measure both strong and weak scaling performance with the emphasis being on a large scale…
Real-time computation of parameter fitting and image reconstruction using graphical processing units
2016
Abstract In recent years graphical processing units (GPUs) have become a powerful tool in scientific computing. Their potential to speed up highly parallel applications brings the power of high performance computing to a wider range of users. However, programming these devices and integrating their use in existing applications is still a challenging task. In this paper we examined the potential of GPUs for two different applications. The first application, created at Paul Scherrer Institut (PSI), is used for parameter fitting during data analysis of μ SR (muon spin rotation, relaxation and resonance) experiments. The second application, developed at ETH, is used for PET (Positron Emission T…