Search NASA⌕ Search

SEARCH · Search NASA

Results for “Time 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 1,549 records · Page 86
All sourcesNASA NTRSDOE OSTI
DOE OSTI2019

High Order Discontinuous Galerkin Method and Applications (Final Technical Report)

This project aims at the developments and improvements of high order accurate discontinuous Galerkin finite element methods for solving partial differential equations arising from DOE applications. Algorithm development, improvement, analysis, implementation and applications aspects are addressed. The partial differential equations under study involve solutions with shocks or other discontinuities, rapidly changing regions, or multiscale solutions. These include the so-called convection dominated cases, such as the hyperbolic equations (e.g. Euler equations of gas dynamics), parabolic equations with small viscosities (e.g. high Reynolds number Navier-Stokes equations), KdV type equations with small dispersion describing water waves (small third derivative terms), time-dependent bi-harmonic equations and equations with hyper-viscosity (small fourth or higher even derivative terms), problems with highly oscillatory coefficients, or problems for which the macro-scale PDE models may not be correct in localized regions and a micro-scale model must be used instead. Furthermore, we are interested in the long time solutions of such equations with waves traveling a long distance, such as the electromagnetic waves. The requirement on numerical methods is thus high order accuracy, low dissipation/dispersion errors over long time, and a clean and sharp discontinuity resolution. It would also be highly desirable to have the numerical methods to be flexible for complicated geometry and boundary conditions, and to be easy in an adaptive and massively parallel environment. The discontinuous Galerkin finite element method has this potential, however the design and analysis of the discontinuous Galerkin method for each specific class of partial differential equations would require detailed research.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Domaine-Specific Runtime to Orchestrate Computation on Heterogeneous Platforms

Task-based runtime systems in the past have sought to exploit inherent asynchronicity in the application execution to reduce overall runtime. In the last decade focus shifted to supporting the heterogeneity that is increasingly prevalent in high-performance computing systems. However, existing task-based runtime systems, being general, come with a challenging set of issues such as the complexity of abstractions and overheads. And they still leave much of the burden of exposing the parallelism on the application developers who also have to fit their applications to the runtime systems’ interfaces. In this paper we take a different approach for applications to target heterogeneous systems through domain-specific run-times. Our design aspires to leverage the domain-specific knowledge of a focused class of scientific simulations to pragmatically orchestrate computations in the simulations.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2014

3D and 4D magnetic susceptibility tomography based on complex MR images

Magnetic susceptibility is the physical property for T2*-weighted magnetic resonance imaging (T2*MRI). The invention relates to methods for reconstructing an internal distribution (3D map) of magnetic susceptibility values, .chi. (x,y,z), of an object, from 3D T2*MRI phase images, by using Computed Inverse Magnetic Resonance Imaging (CIMRI) tomography. The CIMRI technique solves the inverse problem of the 3D convolution by executing a 3D Total Variation (TV) regularized iterative convolution scheme, using a split Bregman iteration algorithm. The reconstruction of .chi. (x,y,z) can be designed for low-pass, band-pass, and high-pass features by using a convolution kernel that is modified from the standard dipole kernel. Multiple reconstructions can be implemented in parallel, and averaging the reconstructions can suppress noise. 4D dynamic magnetic susceptibility tomography can be implemented by reconstructing a 3D susceptibility volume from a 3D phase volume by performing 3D CIMRI magnetic susceptibility tomography at each snapshot time.

47 OTHER INSTRUMENTATION↗
DOE OSTI2017

Parallel magnetic field suppresses dissipation in superconducting nanostrips

The motion of Abrikosov vortices in type-II superconductors results in a finite resistance in the presence of an applied electric current. Elimination or reduction of the resistance via immobilization of vortices is the "holy grail" of superconductivity research. Common wisdom dictates that an increase in the magnetic field escalates the loss of energy since the number of vortices increases. Here we show that this is no longer true if the magnetic field and the current are applied parallel to each other. Our experimental studies on the resistive behavior of a superconducting Mo0.79Ge0.21 nanostrip reveal the emergence of a dissipative state with increasing magnetic field, followed by a pronounced resistance drop, signifying a reentrance to the superconducting state. Large-scale simulations of the 3D time-dependent Ginzburg-Landau model indicate that the intermediate resistive state is due to an unwinding of twisted vortices. When the magnetic field increases, this instability is suppressed due to a better accommodation of the vortex lattice to the pinning configuration. Our findings show that magnetic field and geometrical confinement can suppress the dissipation induced by vortex motion and thus radically improve the performance of superconducting materials.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2016

MODELING THE RISE OF FIBRIL MAGNETIC FIELDS IN FULLY CONVECTIVE STARS

Many fully convective stars exhibit a wide variety of surface magnetism, including starspots and chromospheric activity. The manner by which bundles of magnetic field traverse portions of the convection zone to emerge at the stellar surface is not especially well understood. In the solar context, some insight into this process has been gleaned by regarding the magnetism as consisting partly of idealized thin flux tubes (TFTs). Here we present the results of a large set of TFT simulations in a rotating spherical domain of convective flows representative of a 0.3 M {sub ⊙} main-sequence star. This is the first study to investigate how individual flux tubes in such a star might rise under the combined influence of buoyancy, convection, and differential rotation. A time-dependent hydrodynamic convective flow field, taken from separate 3D simulations calculated with the anelastic equations, impacts the flux tube as it rises. Convective motions modulate the shape of the initially buoyant flux ring, promoting localized rising loops. Flux tubes in fully convective stars have a tendency to rise nearly parallel to the rotation axis. However, the presence of strong differential rotation allows some initially low-latitude flux tubes of moderate strength to develop rising loops that emerge in the near-equatorial region. Magnetic pumping suppresses the global rise of the flux tube most efficiently in the deeper interior and at lower latitudes. The results of these simulations aim to provide a link between dynamo-generated magnetic fields, fluid motions, and observations of starspots for fully convective stars.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Efficient multi-scale representation of visual objects using a biologically plausible spike-latency code and winner-take-all inhibition

Deep neural networks have surpassed human performance in key visual challenges such as object recognition, but require a large amount of energy, computation, and memory. In contrast, spiking neural networks (SNNs) have the potential to improve both the efficiency and biological plausibility of object recognition systems. Here we present a SNN model that uses spike-latency coding and winner-take-all inhibition (WTA-I) to efficiently represent visual stimuli using multi-scale parallel processing. Mimicking neuronal response properties in early visual cortex, images were preprocessed with three different spatial frequency (SF) channels, before they were fed to a layer of spiking neurons whose synaptic weights were updated using spike-timing-dependent-plasticity. We investigate how the quality of the represented objects changes under different SF bands and WTA-I schemes. We demonstrate that a network of 200 spiking neurons tuned to three SFs can efficiently represent objects with as little as 15 spikes per neuron. Furthermore, studying how core object recognition may be implemented using biologically plausible learning rules in SNNs may not only further our understanding of the brain, but also lead to novel and efficient artificial vision systems.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2014

Landau quantization in the spinning cosmic string spacetime

We analyze the quantum phenomenon arising from the interaction of a spinless charged particle with a rotating cosmic string, under the action of a static and uniform magnetic field parallel to the string. We calculate the energy levels of the particle in the non-relativistic approach, showing how these energies depend on the parameters involved in the problem. In order to do this, we solve the time independent Schrödinger equation in the geometry of the spinning cosmic string, taking into account that the coupling between the rotation of the spacetime and the angular momentum of the particle is very weak, such that makes sense to apply the Schrödinger equation in a curved background whose metric has an off diagonal term which involves time and space. It is also assumed that the particle orbits sufficiently far from the boundary of the region of closed timelike curves which exist around this topological defect. Finally, we find the Landau levels of the particle in the presence of a spinning cosmic string endowed with internal structure, i.e., having a finite width and uniformly filled with both material and vacuum energies. - Highlights: • Solution of the wave equation characterizing the problem. • Energy levels of the particle in spacetime of the structureless string. • Expression for an analogous of the quadratic Zeeman effect. • Energy levels of the particle in spacetime of the string with internal structure. • Evidence of the string structure by the internal existence of the vacuum energy.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2023

Microphysically modified magnetosonic modes in collisionless, high-β plasmas

With the support of hybrid-kinetic simulations and analytic theory, we describe the nonlinear behaviour of long-wavelength non-propagating (NP) modes and fast magnetosonic waves in high-β collisionless plasmas, with particular attention to their excitation of and reaction to kinetic micro-instabilities. The perpendicularly pressure balanced polarization of NP modes produces an excess of perpendicular pressure over parallel pressure in regions where the plasma β is increased. For mode amplitudes |δB/B 0 |≳0.3, this excess excites the mirror instability. Particle scattering off these micro-scale mirrors frustrates the nonlinear saturation of transit-time damping, ensuring that large-amplitude NP modes continue their decay to small amplitudes. At asymptotically large wavelengths, we predict that the mirror-induced scattering will be large enough to interrupt transit-time damping entirely, isotropizing the pressure perturbations and morphing the collisionless NP mode into the magnetohydrodynamic (MHD) entropy mode. In fast waves, a fluctuating pressure anisotropy drives both mirror and firehose instabilities when the wave amplitude satisfies |δB/B 0 |≳2β -1 . The induced particle scattering leads to delayed shock formation and MHD-like wave dynamics. Taken alongside prior work on self-interrupting Alfvén waves and self-sustaining ion-acoustic waves, our results establish a foundation for new theories of electromagnetic turbulence in low-collisionality, high-β plasmas such as the intracluster medium, radiatively inefficient accretion flows and the near-Earth solar wind.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2023

Three‐Dimensional Magnetic Reconnection Spreading in Current Sheets of Non‐Uniform Thickness

Abstract Magnetic reconnection in naturally occurring and laboratory settings often begins locally and elongates, or spreads, in the direction perpendicular to the plane of reconnection. Previous work has largely focused on current sheets with a uniform thickness, for which the predicted spreading speed for anti‐parallel reconnection is the local speed of the current carriers. We derive a scaling theory of three‐dimensional (3D) spreading of collisionless anti‐parallel reconnection in a current sheet with its thickness varying in the out‐of‐plane direction, both for spreading from a thinner to thicker region and a thicker to thinner region. We derive an expression for calculating the time it takes for spreading to occur for a current sheet with a given profile of its thickness. A key result is that when reconnection spreads from a thinner to a thicker region, the spreading speed in the thicker region is slower than both the Alfvén speed and the speed of the local current carriers by a factor of the ratio of thin to thick current sheet thicknesses. This is important because magnetospheric and solar observations have previously measured the spreading speed to be slower than previously predicted, so the present mechanism might explain this feature. We confirm the theory via a parametric study using 3D two‐fluid numerical simulations. We use the prediction to calculate the time scale for reconnection spreading in Earth's magnetotail during geomagnetic activity. The results are also potentially important for understanding reconnection spreading in solar flares and the dayside magnetopause of Earth and other planets.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2018

First Direct Observation of Runaway-Electron-Driven Whistler Waves in Tokamaks

DIII-D experiments at low density (n e ~10 19 m -3 ) have directly measured whistler waves in the 100– 200 MHz range excited by multi-MeV runaway electrons. Whistler activity is correlated with runaway intensity (hard x-ray emission level), occurs in novel discrete frequency bands, and exhibits nonlinear limitcycle- like behavior. The measured frequencies scale with the magnetic field strength and electron density as expected from the whistler dispersion relation. The modes are stabilized with increasing magnetic field, which is consistent with wave-particle resonance mechanisms. The mode amplitudes show intermittent time variations correlated with changes in the electron cyclotron emission that follow predator-prey cycles. These can be interpreted as wave-induced pitch angle scattering of moderate energy runaways. The tokamak runaway-whistler mechanisms have parallels to whistler phenomena in ionospheric plasmas. The observations also open new directions for the modeling and active control of runaway electrons in tokamaks.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Pseudonymization at Scale: OLCF’s Summit Usage Data Case Study

The analysis of vast amounts of data and the processing of complex computational jobs have traditionally relied upon high performance computing (HPC) systems, which offer reliable and efficient management of large-scale computational and data resources. Understanding these analyses’ needs is paramount for designing solutions that can lead to better science, and similarly, understanding the characteristics of the user behavior on those systems is important for improving user experiences on HPC systems. A common approach to gathering data about user behavior is to extract workload characteristics from system log data available only to system administrators. Recently at Oak Ridge Leadership Computing Facility (OLCF), however, we unveiled user behavior about the Summit supercomputer by collecting data from a user’s point of view with ordinary Unix commands.In this paper, we discuss the process, challenges, and lessons learned while preparing this dataset for publication and submission to an open data challenge. The original dataset contains personal identifiable information (PII) about the users of OLCF which needed be masked prior to publication, and we determined that anonymization, which scrubs PII completely, destroyed too much of the structure of the data to be interesting for the data challenge. We instead chose to pseudonymize the dataset, which reduced the linkability of the dataset to the users’ identities. Pseudonymization is significantly more computationally expensive than anonymization, and the size of our dataset, which is approximately 175 million lines of raw text, necessitated the development of a parallelized workflow that could be reused on different HPC machines. We demonstrate the scaling behavior of the workflow on two leadership class HPC systems at OLCF, and we show that we were able to bring the overall makespan time from an impractical 20+ hours on a single node down to around 2 hours. As a result of this work, we release the entire pseudonymized dataset and make the workflows and source code publicly available.

Maheshwari, Ketan↗
DOE OSTI2019

Investigation of LSM-YSZ Composite Cathode Performance Degradation with a Multistep Charge Transfer Model

The long-term performance degradation of LSM/YSZ composite cathodes is investigated via calibrated multiphysics simulation with a multistep oxygen reduction reaction (ORR) model and structural coarsening data from a phase field study. Multiphysics simulations, including a multistep ORR mechanism, are first developed for better understanding of the limiting processes in solid oxide fuel cells. The multistep ORR mechanism includes two parallel pathways, surface pathway and bulk pathway, which consist of elementary reaction steps. The multiphysics model is simultaneously calibrated with experimental polarization curves and impedance behavior for various air/fuel supply conditions. To our knowledge, this is the first time that a multistep ORR model is calibrated with such datasets. Next, the calibrated simulations are utilized to simulate a 2D half-cell constructed with measured microstructural data and random heterogeneity. Finally, the long-term performance degradation of the half-cell is predicted by the calibrated multiphysics model coupled with structural coarsening trends simulated using a phase field-based coarsening model. Degradation of both polarization curves and impedance behavior is investigated. Thorough analyses, including changes of contributions from different pathways, the resistance components, and overall reaction order, are performed to provide more insights into cathode performance degradation due to grain coarsening phenomena.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Serial2Parallel

In the era of machine learning, we often need to run the same code/script many times with little or no variations (e.g., performance evaluation, data preprocessing, data generation, hyperparameter tuning, etc.). It is not a problem when you just need to do that a few times, but when the number of repetitions becomes very large, it can be a daunting task. The code “Serial2Parallel” provides an easy way for users to be able to run many numbers of any serial code/scripts in a parallel manner across multiple nodes in an message passing interface (MPI) cluster. The code includes the server program that deals with task pool management and client program that processes task. The server gets the tasks ready and waits for clients' connections. The client code pulls tasks from the server and processes them. The client code will run in parallel.

Sangkeun, MattLee↗
DOE OSTI2023

A PIC Bootstrapping Strategy for Exascale CFD-DEM Simulation

The exascale computing era comes with the release of MFIX-Exa, a new code for studying gas-particle fluidization using CFD-DEM and PIC models on massively parallel, heterogeneous high-performance computing architectures. However, as compute resources continue to grow in scale and complexity, so too does the impetus to use them efficiently. Here, we propose a novel bootstrapping method to minimize neglected simulation time and cast the approach more broadly among a growing field of multi-fidelity methods.

Porcu, Roberto↗
DOE OSTI2018

Final Report: Optimal Model Complexity in Geological Carbon Sequestration: A Response Surface Uncertainty Analysis

The critical component of a risk assessment study in evaluating GCS is an analysis of uncertainty in CO2 modeling. In such analyses, direct numerical simulation of CO2 flow and leakage requires many time-consuming model runs. Alternatively, analytical methods have been developed which allow fast and efficient estimation of CO2 storage and leakage, although restrictive assumptions on formation rock and fluid properties are employed. In this study, an intermediate approach is proposed based on the Design of Experiment and Response Surface methodology, which consists of using a limited number of numerical simulations to estimate a prediction outcome as a combination of the most influential uncertain site properties. The methodology can be implemented within a Monte Carlo framework to efficiently assess parameter and prediction uncertainty while honoring the accuracy of numerical simulations. The choice of the uncertain properties is flexible and can include geologic parameters that influence reservoir heterogeneity, engineering parameters that influence gas trapping and migration, and reactive parameters that influence the extent of fluid/rock reactions. The method was tested and verified on modeling long-term CO2 flow, non-isothermal heat transport, and CO2 dissolution storage by coupling two-phase flow with explicit miscibility calculation using an accurate equation of state that gives rise to convective mixing of formation brine variably saturated with CO2. All simulations were performed using three-dimensional high-resolution models including a target deep saline aquifer, overlying caprock, and a shallow aquifer. To evaluate the uncertainty in representing reservoir permeability, sediment hierarchy of a heterogeneous digital stratigraphy was mapped to create multiple irregularly shape stratigraphic models of decreasing geologic resolutions: heterogeneous (reference), lithofacies, depositional environment, and a (homogeneous) geologic formation. To ensure model equivalency, all the stratigraphic models were successfully upscaled from the reference heterogeneous model for bulk flow and transport predictions (Zhang & Zhang, 2015). GCS simulation was then simulated with all models, yielding insights into the level of parameterization complexity that is needed for the accurate simulation of reservoir pore pressure, CO2 storage, leakage, footprint, and dissolution over both short (i.e., injection) and longer (monitoring) time scales. Important uncertainty parameters that impact these key performance metrics were identified for the stratigraphic models as well as for the heterogeneous model, leading to the development of reduced/simplified models at lower characterization cost that can be used for the reservoir uncertainty analysis. All the CO2 modeling was conducted using PFLOTRAN – a massively parallel, multiphase, multi-component, and reactive transport simulator developed by a multi-laboratory DOE/SciDAC (Scientific Discovery through Advanced Computing) project (Zhang et al., 2017, in review). Within the uncertainty analysis framework, increasing reservoir depth were investigated to explore its effect on the uncertainty outcomes and the potential for developing gravity-stable injection with increased storage security (Dai et al., 20126; Dai et al., 2017, in review). Finally, to accurately model CO2 fluid-rock reactions and resulting long-term storage as secondary carbonate minerals, a modified kinetic rate law for general mineral dissolution and precipitation was proposed and verified that is invariant to a scale transformation of the mineral formula weight. This new formulation will lead to more accurate assessment of mineral storage over geologic time scales (Lichtner, 2016).

58 GEOSCIENCES↗
DOE OSTI2018

A Gain Scheduling Wide-Area Damping Controller for the Efficient Integration of Photovoltaic Plant

With photovoltaic generation increasing, the wide-area oscillation has a vital influence on the long-distance power transmission of photovoltaic power. For an efficient integration of photovoltaic generation, a gain scheduling wide-area damping controller (WADC) is designed to guarantee system damping performance under a wide-area signal delay and a changing operating condition, which could be aggravated by the fluctuant and uncertain photovoltaic generation. A linear parameter-varying (LPV) model is proposed to represent all possible operating conditions within varying ranges of the photovoltaic generation and the load. Then, a tensor product (TP) model transformation is applied to transform the LPV model into a TP model. Based on the TP model, a parallel distributed compensation controller (PDC) framework is adopted to design the gain scheduling WADC, which is solved by linear matrix inequalities (LMIs) considering the accommodation to the random interval delay. In addition, a staircase basis function method is proposed to reduce the conservatism of LMIs within the PDC framework. The gain of the proposed WADC will be adjusted according to the operating condition acknowledged through the real-time scheduling and control center, once a wide-area oscillation is detected. Lastly, the performance of the proposed WADC is verified by a 2-area 4-machine system and a 68-bus 16-machine system.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗