Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed parallelization”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 919 records · Page 51

Parallel transport of long mean-free-path plasma along open magnetic field lines: Parallel heat flux

In a long mean-free-path plasma where temperature anisotropy can be sustained, the parallel heat flux has two components with one associated with the parallel thermal energy and the other the perpendicular thermal energy. Due to the large deviation of the distribution function from local Maxwellian in an open field line plasma with low collisionality, the conventional perturbative calculation of the parallel heat flux closure in its local or non-local form is no longer applicable. Here, a non-perturbative calculation is presented for a collisionless plasma in a two-dimensional flux expander bounded by absorbing walls. Specifically, closures of previously unfamiliar form are obtained for ions and electrons, which relate two distinct components of the species parallel heat flux to the lower order fluid moments such as density, parallel flow, parallel and perpendicular temperatures, and the field quantities such as the magnetic field strength and the electrostatic potential. The plasma source and boundary condition at the absorbing wall enter explicitly in the closure calculation. Although the closure calculation does not take into account wave-particle interactions, the results based on passing orbits from steady-state collisionless drift-kinetic equation show remarkable agreement with fully kinetic-Maxwell simulations. As an example of the physical implications of the theory, the parallel heat flux closures are found to predict a surprising observation in the kinetic-Maxwell simulation of the 2D magnetic flux expander problem, where the parallel heat flux of the parallel thermal energy flows from low to high parallel temperature region.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Rarefaction windows in a high-power impulse magnetron sputtering plasma

The velocity distribution function of the sputtered particles in the direction parallel to the planar magnetron cathode is studied by spatially- and time-resolved laser-induced fluorescence spectroscopy in a short-duration (20 μs) high-power impulse magnetron sputtering discharge. The experimental evidence for the neutral and ionized sputtered particles to have a constant (saturated) velocity at the end of the plasma on-time is demonstrated. The velocity component parallel to the target surface reaches the values of about 5 km/s for Ti atoms and ions, which is higher that the values typically measured in the direct current sputtering discharges before. The results point out on the presence of a strong gas rarefaction significantly reducing the sputtered particles energy dissipation during a certain time interval at the end of the plasma pulse, referred to as “rarefaction window” in this work. The obtained results agree with and essentially clarify the dynamics of HiPIMS discharge studied during the plasma off-time previously in the work: N. Britun, Appl. Phys. Lett. 99, 131504 (2011)

36 MATERIALS SCIENCE↗

Multidimensional electron beam-plasma instabilities in the relativistic regime

The interest in relativistic beam-plasma instabilities has been greatly rejuvenated over the past two decades by novel concepts in laboratory and space plasmas. Recent advances in this long-standing field are here reviewed from both theoretical and numerical points of view. The primary focus is on the two-dimensional spectrum of unstable electromagnetic waves growing within relativistic, unmagnetized, and uniform electron beam-plasma systems. Although the goal is to provide a unified picture of all instability classes at play, emphasis is put on the potentially dominant waves propagating obliquely to the beam direction, which have received little attention over the years. First, the basic derivation of the general dielectric function of a kinetic relativistic plasma is recalled. Next, an overview of two-dimensional unstable spectra associated with various beam-plasma distribution functions is given. Both cold-fluid and kinetic linear theory results are reported, the latter being based on waterbag and Maxwell-Juettner model distributions. The main properties of the competing modes (developing parallel, transverse, and oblique to the beam) are given, and their respective region of dominance in the system parameter space is explained. Later sections address particle-in-cell numerical simulations and the nonlinear evolution of multidimensional beam-plasma systems. The elementary structures generated by the various instability classes are first discussed in the case of reduced-geometry systems. Validation of linear theory is then illustrated in detail for large-scale systems, as is the multistaged character of the nonlinear phase. Finally, a collection of closely related beam-plasma problems involving additional physical effects is presented, and worthwhile directions of future research are outlined.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

ANISOTROPIC THREE-DIMENSIONAL FOCUSED TRANSPORT OF SOLAR ENERGETIC PARTICLES IN THE INNER HELIOSPHERE

We investigate the combined effects of solar energetic particle propagation, parallel and perpendicular to the large-scale magnetic field in the solar wind. Numerical methods employing stochastic differential equations are used incorporating pitch-angle diffusion, focusing, and pitch-angle-dependent diffusion perpendicular to the magnetic field. We compute spatial distributions of approx100 keV electrons and 4 MeV protons in the inner heliosphere, assuming impulsive injection near the Sun over a limited range of solar longitude and latitude. In addition, spatial distributions and intensity-time profiles for various combinations of the parallel and perpendicular mean free path, with different assumptions for the dependence of lambda{sub perpendicular} on the radial distance and pitch angle, are investigated. We find that realistic results can be obtained when we assume that the perpendicular mean free path scales in the inner heliosphere with the gyroradius of the particles. Step-like decreases of particle intensities as frequently observed in impulsive events at 1 AU can be reproduced for a ratio of lambda{sub perpendicular}/lambda{sub ||} a few times 10{sup -5}.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A Semi-Implicit Lattice Method for Simulating Flow

We propose a new semi-implicit lattice numerical method for modeling fluid flow that depends only on local primitive variable information (density, pressure, velocity) and not on relaxed upstream distribution function values. This method has the potential for reducing parallel communication and permitting larger time steps compared to the lattice Boltzmann method. The lid-driven cavity is modeled to demonstrate the accuracy of the method.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Transport of radial heat flux and second sound in fusion plasmas

Simple flux-gradient relations that involve time delay and radial coupling are discussed. Such a formulation leads to a rather simple description of avalanches and may explain breaking of gyroBohm transport scaling. The generalization of the flux-gradient relation (i.e., constitutive relation), which involve both time delay and spatial coupling, is derived from drift-kinetic equation, leading to kinetic definitions of constitutive elements such as the flux of radial heat flux. This allows numerical simulations to compute these cubic quantities directly. The formulation introduced here can be viewed as an extension of turbulence spreading to include the effect of spreading of cross-phase as well as turbulence intensity, combined in such a way to give the flux. The link between turbulence spreading and entropy production is highlighted. An extension of this formulation to general quasi-linear theory for the distribution function in the phase space of radial position and parallel velocity is also discussed.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

An MPI + $X$ implementation of contact global search using Kokkos

This paper describes an approach that seeks to parallelize the spatial search associated with computational contact mechanics. In contact mechanics, the purpose of the spatial search is to find “nearest neighbors,” which is the prelude to an imprinting search that resolves the interactions between the external surfaces of contacting bodies. In particular, we are interested in the contact global search portion of the spatial search associated with this operation on domain-decomposition-based meshes. Specifically, we describe an implementation that combines standard domain-decomposition-based MPI-parallel spatial search with thread-level parallelism (MPI-X) available on advanced computer architectures (those with GPU coprocessors). Our goal is to demonstrate the efficacy of the MPI-X paradigm in the overall contact search. Standard MPI-parallel implementations typically use a domain decomposition of the external surfaces of bodies within the domain in an attempt to efficiently distribute computational work. This decomposition may or may not be the same as the volume decomposition associated with the host physics. The parallel contact global search phase is then employed to find and distribute surface entities (nodes and faces) that are needed to compute contact constraints between entities owned by different MPI ranks without further inter-rank communication. Key steps of the contact global search include computing bounding boxes, building surface entity (node and face) search trees and finding and distributing entities required to complete on-rank (local) spatial searches. To enable source-code portability and performance across a variety of different computer architectures, we implemented the algorithm using the Kokkos hardware abstraction library. While we targeted development towards machines with a GPU accelerator per MPI rank, we also report performance results for OpenMP with a conventional multi-core compute node per rank. Results here demonstrate a 47 % decrease in the time spent within the global search algorithm, comparing the reference ACME algorithm with the GPU implementation, on an 18M face problem using four MPI ranks. As a result, while further work remains to maximize performance on the GPU, this result illustrates the potential of the proposed implementation.

97 MATHEMATICS AND COMPUTING↗

Distributed merge trees

Improved simulations and sensors are producing datasets whose increasing complexity exhausts our ability to visualize and comprehend them directly. To cope with this problem, we can detect and extract significant features in the data and use them as the basis for subsequent analysis. Topological methods are valuable in this context because they provide robust and general feature definitions. As the growth of serial computational power has stalled, data analysis is becoming increasingly dependent on massively parallel machines. To satisfy the computational demand created by complex datasets, algorithms need to effectively utilize these computer architectures. The main strength of topological methods, their emphasis on global information, turns into an obstacle during parallelization. We present two approaches to alleviate this problem. We develop a distributed representation of the merge tree that avoids computing the global tree on a single processor and lets us parallelize subsequent queries. To account for the increasing number of cores per processor, we develop a new data structure that lets us take advantage of multiple shared-memory cores to parallelize the work on a single node. Finally, we present experiments that illustrate the strengths of our approach as well as help identify future challenges. Copyright © 2013 ACM 978-1-4503-1922-5/13/ 02...$15.00.

Morozov, D↗

Distributed Merge Trees

Improved simulations and sensors are producing datasets whose increasing complexity exhausts our ability to visualize and comprehend them directly. To cope with this problem, we can detect and extract significant features in the data and use them as the basis for subsequent analysis. Topological methods are valuable in this context because they provide robust and general feature definitions. As the growth of serial computational power has stalled, data analysis is becoming increasingly dependent on massively parallel machines. To satisfy the computational demand created by complex datasets, algorithms need to effectively utilize these computer architectures. The main strength of topological methods, their emphasis on global information, turns into an obstacle during parallelization. We present two approaches to alleviate this problem. We develop a distributed representation of the merge tree that avoids computing the global tree on a single processor and lets us parallelize subsequent queries. To account for the increasing number of cores per processor, we develop a new data structure that lets us take advantage of multiple shared-memory cores to parallelize the work on a single node. Finally, we present experiments that illustrate the strengths of our approach as well as help identify future challenges.

97 MATHEMATICS AND COMPUTING↗

In Situ X-Ray Observations of Dendritic Fragmentation During Directional Solidification of a Sn-Bi Alloy

Dendrite fragmentation is an important phenomenon in microstructural development during solidification. For instance, it plays a key role in initiating the columnar-to-equiaxed transition (CET). Here, we use x-ray radiography to study dendrite fragmentation rate in a Sn-39.5 wt.% Bi alloy during directional solidification. Experiments were performed in which solidification was parallel and anti-parallel to gravity, leading to significantly different fragmentation rates. We quantify the distribution of fragmentation rate as a function of distance from the solidification front, time in the mushy zone, and volume fraction of solid. While the observed fragmentation rate can be high, there is no evidence of a CET, illustrating that it requires more than just fragmentation to occur.

Gibbs, John W.↗

Partitioning Models for Scaling Parallel Sparse Matrix-Matrix Multiplication

We investigate outer-product--parallel, inner-product--parallel, and row-by-row-product--parallel formulations of sparse matrix-matrix multiplication (SpGEMM) on distributed memory architectures. For each of these three formulations, we propose a hypergraph model and a bipartite graph model for distributing SpGEMM computations based on one-dimensional (1D) partitioning of input matrices. Here, we also propose a communication hypergraph model for each formulation for distributing communication operations. The computational graph and hypergraph models adopted in the first phase aim at minimizing the total message volume and balancing the computational loads of processors, whereas the communication hypergraph models adopted in the second phase aim at minimizing the total message count and balancing the message volume loads of processors. That is, the computational partitioning models reduce the bandwidth cost and the communication hypergraph models reduce the latency cost. Our extensive parallel experiments on up to 2048 processors for a wide range of realistic SpGEMM instances show that although the outer-product--parallel formulation scales better, the row-by-row-product--parallel formulation is more viable due to its significantly lower partitioning overhead and competitive scalability. For computational partitioning models, our experimental findings indicate that the proposed bipartite graph models are attractive alternatives to their hypergraph counterparts because of their lower partitioning overhead. Finally, we show that by reducing the latency cost besides the bandwidth cost through using the communication hypergraph models, the parallel SpGEMM time can be further improved up to 32%.

97 MATHEMATICS AND COMPUTING↗

An experimental comparison of triggered and random pulse train uncertainties

In this paper we present an experimental comparison of signal-triggered and randomly triggered based analysis algorithms of neutron multiplicity data. Traditional shift register type signal-triggered multiplicity analysis of singles, doubles and triples rates is compared with analysis using randomly triggered gates. Two methods of random gate generation are explored - non-overlapping gates (Feyrunan approach) and periodic overlapping gates (fast accidentals). Using californium sources with low, medium and high rate in combination with AmLi sources (as a surrogate for plutonium) we investigate relative standard deviation (RSD) of data in order to determine if there are parameter spaces in which one of the measurement methods should be preferred. Neutron correlation analysis is a commonly used NDA technique to assay plutonium mass. The data can be collected in two distinct ways: using signal-triggered or randomly triggered counting gates. Analysis algorithms were developed for both approaches to determine singles (S), doubles (D) and triples (7) rates from the measured sample. Currently the most commonly implemented technique to collect neutron coincidence data utilizes shift register based electronics. Shift register uses signal-triggered counting gates to generate foreground multiplicity distribution of correlated+accidental events and a random gate (opened after a predefined long delay following the signal trigger) to generate background multiplicity distribution of accidental events. Modern shift registers include fast accidental option to sample data with a fixed clock frequency. This way a set of overlapping gates is used to generate background multiplicity distributions in order to improve the measurement precision. In parallel to shift register approach the Feynman variance technique is frequently used, which utilizes set of consecutive non-overlapping gates. In general, different user communities (e.g. safeguards, nuclear material accountancy, emergency response) have used only one of the two analysis methods for the nuclear material assay. The aim of this study is to provide a systematic comparison of the precision of the measured S, D, T rates and {sup 240}Pu effective mass obtained using the above mentioned pulse train sampling techniques. In order to perform this task, a LANL developed list mode based data acquisition system is used, where the entire pulse train is recorded and subsequently analyzed. The list mode acquisition brings an essential advantage for this type of comparison, since the very same pulse train can be analyzed using signal-triggered as well as randomly triggered counting gates. The aim of this study is not only to compare the precision of signal-triggered versus random triggered sampling techniques, but also to investigate the influence of fast accidental sampling on the precision of signal-triggered results. In addition the different random sampling techniques used in safeguards are investigated. For this purpose we implement two types of random sampling - non-overlapping gates (Feynrnan approach) and periodic overlapping gates (fast accidentals). In the following sections the equations utilized in the pulse train analysis are described, experimental setup and measurement techniques are discussed and finally the results are summarized and discussed.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Massively scalable workflows for quantum chemistry: BigChem and ChemCloud

Electronic structure theory, i.e., quantum chemistry, is the fundamental building block for many problems in computational chemistry. Here we present a new distributed computing framework (BigChem), which allows for an efficient solution of many quantum chemistry problems in parallel. BigChem is designed to be easily composable and leverages industry-standard middleware (e.g., Celery, RabbitMQ, and Redis) for distributed approaches to large scale problems. BigChem can harness any collection of worker nodes, including ones on cloud providers (such as AWS or Azure), local clusters, or supercomputer centers (and any mixture of these). BigChem builds upon MolSSI packages, such as QCEngine to standardize the operation of numerous computational chemistry programs, demonstrated here with Psi4, xtb, geomeTRIC, and TeraChem. BigChem delivers full utilization of compute resources at scale, offers a programable canvas for designing sophisticated quantum chemistry workflows, and is fault tolerant to node failures and network disruptions. We demonstrate linear scalability of BigChem running computational chemistry workloads on up to 125 GPUs. Finally, we present ChemCloud, a web API to BigChem and successor to TeraChem Cloud. ChemCloud delivers scalable and secure access to BigChem over the Internet.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Mitigating Cascading Outages in Severe Weather Using Simulation-Based Optimization

Severe weather events can trigger cascading power outages and lead to significant losses. In this work, we investigate cascading outage mitigation under severe weather conditions. Given day-ahead weather forecasts and component failure models, we aim to identify a set of power lines that can be hardened to minimize the expected impact of potential cascading outages. Since the expected load shedding cannot be expressed as an explicit function of line hardening decisions and system states, we developed a cascading outage simulator to estimate the expected value of load shedding under various initial weather-related disruption scenarios generated using a weather forecast. To avoid massive enumeration of all possible combinations of line hardening decisions and reduce the simulation efforts, we employed an efficient simulation-based optimization approach that quickly identifies the (near) optimal line hardening decisions in the presence of both large simulation noises due to the highly variable initial disturbances and system states, and significant randomness in the subsequent cascades. Furthermore, the algorithm is also able to utilize parallel computing to dramatically reduce computation time to support decision making in preparation for severe weather conditions. We performed a case study on the Northeast Power Coordinating Council (NPCC) 140-bus system model to demonstrate that our approach can significantly improve power grid resilience to adverse weather events.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Multithreaded parallelization of the energy and analytic gradient in the fragment molecular orbital method

The fragment molecular orbital method in GAMESS is parallelized in a multithreaded OpenMP implementation combined with the MPI version of the two-level generalized distributed data interface. The energy and analytic gradient in gas phase and the polarizable continuum model of solvation are parallelized in this hybrid three-level scheme, achieving a large memory footprint reduction and a high parallel efficiency on Intel Xeon Phi processors. The parallel efficiency is demonstrated on the Stampede2 and Theta supercomputers using up to 2048 nodes (262 144 threads).

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Shift-and-invert parallel spectral transformation eigensolver: Massively parallel performance for density-functional based tight-binding

The Shift-and-invert parallel spectral transformations (SIPs), a computational approach to solve sparse eigenvalue problems, is developed for massively parallel architectures with exceptional parallel scalability and robustness. The capabilities of SIPs are demonstrated by diagonalization of density-functional based tight-binding (DFTB) Hamiltonian and overlap matrices for single-wall metallic carbon nanotubes, diamond nanowires, and bulk diamond crystals. The largest (smallest) example studied is a 128,000 (2000) atom nanotube for which ~330,000 (~5600) eigenvalues and eigenfunctions are obtained in ~190 (~5) seconds when parallelized over 266,144 (16,384) Blue Gene/Q cores. Weak scaling and strong scaling of SIPs are analyzed and the performance of SIPs is compared with other novel methods. Different matrix ordering methods are investigated to reduce the cost of the factorization step, which dominates the time-to-solution at the strong scaling limit. As a result, a parallel implementation of assembling the density matrix from the distributed eigenvectors is demonstrated.

36 MATERIALS SCIENCE↗

In situ multi-tier auto-ignition detection applied to dual-fuel combustion simulations

Here we use an anomaly detection methodology that is centered on analyzing fourth-order joint moments (co-kurtosis), particularly focusing on its application in auto-ignition of combustion problems with large numbers of species. Unsupervised anomaly detection is challenging to generalize across problem types and domains. A recent technique, centered on analyzing information in the fourth-order joint moment co-kurtosis, has shown promise, especially for high-dimensional scientific data. In this work we present developments to the co-kurtosis based anomaly detection method needed to make it effective and scalable for large-scale distributed scientific data, such as those generated by massively parallel simulations. An in situ co-kurtosis algorithm is employed as the anomaly detection method for identifying ignition kernels in simulations of turbulent combustion. Here, we extend an existing methodology which identifies regions of the domain where anomalies are present, and add another tier of anomaly detection where the individual samples contributing to the anomaly are identified. We apply this algorithm on-the-fly to a variety of turbulent reacting flow problems and compare it to the widely used (but significantly more expensive) chemical explosive mode analysis (CEMA). We demonstrate the ability of the method to detect and identify the onset of low and high temperature ignition which can be used for computational steering, as chemical and combustion anomalies occur intermittently at spatio-temporal locations unknown a priori. Finally, we apply our lightweight in situ algorithm to an exascale high-fidelity simulation with a total of 2.4 Trillion degrees of freedom, performed using an adaptive mesh refinement solver. Furthermore, through a scalability analysis, we show that the relative computational cost of this in-situ anomaly detection algorithm compared to an iteration of the reacting flow solver is negligible.

97 MATHEMATICS AND COMPUTING↗

Self-sustaining sound in collisionless, high- β plasma

Using analytical theory and hybrid-kinetic numerical simulations, we demonstrate that, in a collisionless plasma, long-wavelength ion-acoustic waves (IAWs) with amplitudes $δn/n_0≳2/β$ (where $β\gg1$ is the ratio of thermal to magnetic pressure) generate sufficient pressure anisotropy to destabilize the plasma to firehose and mirror instabilities. These kinetic instabilities grow rapidly to reduce the pressure anisotropy by pitch-angle scattering and trapping particles, respectively, thereby impeding the maintenance of Landau resonances that enable such waves’ otherwise potent collisionless damping. The result is wave dynamics that evince a weakly collisional plasma: the ion distribution function is near-Maxwellian, the field-parallel flow of heat resembles its Braginskii form (except in regions where large-amplitude magnetic mirrors strongly suppress particle transport), and the relations between various thermodynamic quantities are more ‘fluid-like’ than kinetic. A nonlinear fluctuation–dissipation relation for self-sustaining IAWs is obtained by solving a plasma-kinetic Langevin problem, which demonstrates suppressed damping, enhanced fluctuation levels and weakly collisional thermodynamics when IAWs with $δn/n_0≳2/β$ are stochastically driven. We investigate how our results depend upon the scale separation between the wavelength of the IAW and the Larmor radius of the ions, and discuss briefly their implications for our understanding of turbulence and transport in the solar wind and the intracluster medium of galaxy clusters.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗