Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “parallel, scaling”

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 1,801 records · Page 100

Improving MPI Collective I/O for High Volume Non-Contiguous Requests With Intra-Node Aggregation

Two-phase I/O is a well-known strategy for implementing collective MPI-IO functions. It redistributes I/O requests among the calling processes into a form that minimizes the file access costs. As modern parallel computers continue to grow into the exascale era, the communication cost of such request redistribution can quickly overwhelm collective I/O performance. In this study, this effect has been observed from parallel jobs that run on multiple compute nodes with a high count of MPI processes on each node. Furthermore, to reduce the communication cost, we present a new design for collective I/O by adding an extra communication layer that performs request aggregation among processes within the same compute nodes. This approach can significantly reduce inter-node communication contention when redistributing the I/O requests. We evaluate the performance and compare it with the original two-phase I/O on Cray XC40 parallel computers (Theta and Cori) with Intel KNL and Haswell processors. Using I/O patterns from two large-scale production applications and an I/O benchmark, we show our proposed method effectively reduces the communication cost and hence maintains the scalability for a large number of processes.

97 MATHEMATICS AND COMPUTING↗

MARMOT update for oxide fuel modeling

This report summarizes the lower-length-scale research and development progresses in FY16 at Idaho National Laboratory in developing mechanistic materials models for oxide fuels, in parallel to the development of the MARMOT code which will be summarized in a separate report. This effort is a critical component of the microstructure based fuel performance modeling approach, supported by the Fuels Product Line in the Nuclear Energy Advanced Modeling and Simulation (NEAMS) program. The progresses can be classified into three categories: 1) development of materials models to be used in engineering scale fuel performance modeling regarding the effect of lattice defects on thermal conductivity, 2) development of modeling capabilities for mesoscale fuel behaviors including stage-3 gas release, grain growth, high burn-up structure, fracture and creep, and 3) improved understanding in material science by calculating the anisotropic grain boundary energies in UO$_2$ and obtaining thermodynamic data for solid fission products. Many of these topics are still under active development. They are updated in the report with proper amount of details. For some topics, separate reports are generated in parallel and so stated in the text. The accomplishments have led to better understanding of fuel behaviors and enhance capability of the MOOSE-BISON-MARMOT toolkit.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

ON SPECTRAL BREAKS IN THE POWER SPECTRA OF MAGNETIC FLUCTUATIONS IN FAST SOLAR WIND BETWEEN 0.3 AND 0.9 AU

We analyze the radial variation of the power spectra of the magnetic field from 0.3 to about 0.9 AU, using Helios 2 spacecraft measurements in the fast solar wind. The time resolution of the magnetic field data allows us to study the power spectra up to 2 Hz. Generally, the corresponding spectral break frequency f{sub b} and the Doppler-shifted frequencies, which are related to the proton gyroradius and inertial scales, are close to a frequency f of about 0.5 Hz at a distance of 1 AU from the Sun. However, studying the radial evolution of the power spectra offers us the possibility to distinguish between those scales. Recent Ulysses observations show that, while the proton scales vary, f{sub b} stays nearly constant with the heliocentric distance R. In our study we confirm that f{sub b} varies within a small interval of [0.2, 0.4] Hz only, as R varies from 0.3 to 0.9 AU. Moreover, if we assume parallel propagating fluctuations (with respect to the solar wind flow or background magnetic field), we can show that none of the proton scales are coincident with the break scale. If, however, we take into account the two-dimensional nature of the turbulent fluctuations, then we can show that the spatial scale corresponding to f{sub b} (R) does follow the proton inertial scale, {lambda}{sub p}(R), but not the proton gyroradius scale, {rho}{sub p}(R), as a function of heliocentric distance. These observations indicate that the spectral break at the proton inertial scale might be related to the Hall effect, or be controlled by the ion-cyclotron damping of obliquely propagating fluctuations or the formation of current sheets scaling like {lambda}{sub p}, which could be responsible for ion heating through magnetic reconnection.

79 ASTRONOMY AND ASTROPHYSICS↗

Ion-temperature-gradient sensitivity of the hydrodynamic instability caused by shear in the magnetic-field-aligned plasma flow

The cross-magnetic-field (i.e., perpendicular) profile of ion temperature and the perpendicular profile of the magnetic-field-aligned (parallel) plasma flow are sometimes inhomogeneous for space and laboratory plasma. Instability caused either by a gradient in the ion-temperature profile or by shear in the parallel flow has been discussed extensively in the literature. In this paper, (1) hydrodynamic plasma stability is investigated, (2) real and imaginary frequency are quantified over a range of the shear parameter, the normalized wavenumber, and the ratio of density-gradient and ion-temperature-gradient scale lengths, and (3) the role of inverse Landau damping is illustrated for the case of combined ion-temperature gradient and parallel-flow shear. We find that increasing the ion-temperature gradient reduces the instability threshold for the hydrodynamic parallel-flow shear instability, also known as the parallel Kelvin-Helmholtz instability or the D'Angelo instability. We also find that a kinetic instability arises from the coupled, reinforcing action of both free-energy sources. For the case of comparable electron and ion temperature, we illustrate analytically the transition of the D'Angelo instability to the kinetic instability as (a) the shear parameter, (b) the normalized wavenumber, and (c) the ratio of density-gradient and ion-temperature-gradient scale lengths are varied and we attribute the changes in stability to changes in the amount of inverse ion Landau damping. We show that near a normalized wavenumber k{sub ⊥}ρ{sub i} of order unity (i) the real and imaginary values of frequency become comparable and (ii) the imaginary frequency, i.e., the growth rate, peaks.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Spatial scales and temporal scales in the theory of magnetohydrodynamic turbulence

In the study of incompressible magnetohydrodynamic (MHD) turbulence the correlation lengths along the directions parallel and perpendicular to the local mean magnetic field may be defined by iso-energy surfaces derived from second order structure functions. The correlation time of the turbulence may be defined in a similar manner by means of a temporal second order structure function. Moreover, there is a natural correspondence between the lengthscales of fluctuations with a given energy and the timescale of fluctuations of the same energy so that each iso-energy surface is associated with a unique pair of parallel and perpendicular correlation lengths and a unique correlation time. In the case when the magnetic Prandtl number is unity, Pr{sub m{identical_to}{nu}}/{eta}=1, it is shown that the correlation time {tau} associated with an iso-energy contour of energy E is equal to the energy cascade time of the turbulence in the sense that E/{tau}{approx}{epsilon}, where {epsilon} is the energy cascade rate. For balanced MHD turbulence, turbulence with vanishing cross-helicity, both the Goldreich and Sridhar theory and Boldyrev's theory are shown to have this same property. An attempt is made to generalize these ideas to imbalanced MHD turbulence, turbulence with nonvanishing cross-helicity. It is shown that if the normalized cross-helicity {sigma}{sub c} is a constant in the inertial range, independent of wavenumber, then one obtains a model in which the energy cascade times of the two Elsasser fields are equal throughout the inertial range, that is, {tau}{sup +}/{tau}{sup -}=1. This result appears to be contradicted by numerical simulations and, therefore, the generalization to imbalanced turbulence discussed here requires further development; this is an important unsolved problem.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Nyx: A MASSIVELY PARALLEL AMR CODE FOR COMPUTATIONAL COSMOLOGY

We present a new N-body and gas dynamics code, called Nyx, for large-scale cosmological simulations. Nyx follows the temporal evolution of a system of discrete dark matter particles gravitationally coupled to an inviscid ideal fluid in an expanding universe. The gas is advanced in an Eulerian framework with block-structured adaptive mesh refinement; a particle-mesh scheme using the same grid hierarchy is used to solve for self-gravity and advance the particles. Computational results demonstrating the validation of Nyx on standard cosmological test problems, and the scaling behavior of Nyx to 50,000 cores, are presented.

79 ASTRONOMY AND ASTROPHYSICS↗

High-performance computing in water resources hydrodynamics

In this work, we present a vision of future water resources hydrodynamics codes that can fully utilize the strengths of modern high-performance computing. The advances to computing power, formerly driven by the improvement of central processing unit processors, now focus on parallel computing and, in particular, the use of graphics processing units (GPUs). However, this shift to a parallel framework requires refactoring the code to make efficient use of the data as well as changing even the nature of the algorithm that solves the system of equations. These concepts along with other features such as the precision for the computations, dry regions management, and input/output data are analyzed in this paper. A 2D multi-GPU flood code applied to a large-scale test case is used to corroborate our statements and ascertain the new challenges for the next-generation parallel water resources codes.

54 ENVIRONMENTAL SCIENCES↗

Genetic algorithm based task reordering to improve the performance of batch scheduled massively parallel scientific applications

The growth in size of networked high performance computers along with novel accelerator-based node architectures has further emphasized the importance of communication efficiency in high performance computing. The world's largest high performance computers are usually operated as shared user facilities due to the costs of acquisition and operation. Applications are scheduled for execution in a shared environment and are placed on nodes that are not necessarily contiguous on the interconnect. Furthermore, the placement of tasks on the nodes allocated by the scheduler is sub-optimal, leading to performance loss and variability. Here, we investigate the impact of task placement on the performance of two massively parallel application codes on the Titan supercomputer, a turbulent combustion flow solver (S3D) and a molecular dynamics code (LAMMPS). Benchmark studies show a significant deviation from ideal weak scaling and variability in performance. The inter-task communication distance was determined to be one of the significant contributors to the performance degradation and variability. A genetic algorithm-based parallel optimization technique was used to optimize the task ordering. This technique provides an improved placement of the tasks on the nodes, taking into account the application's communication topology and the system interconnect topology. As a result, application benchmarks after task reordering through genetic algorithm show a significant improvement in performance and reduction in variability, therefore enabling the applications to achieve better time to solution and scalability on Titan during production.

97 MATHEMATICS AND COMPUTING↗

Uncovering Access, Reuse, and Sharing Characteristics of I/O-Intensive Files on Large-Scale Production HPC Systems

Large-scale high-performance computing (HPC) applications running on supercomputers produce large amounts of data routinely and store it in files on multi-PB shared parallel storage systems. Unfortunately, storage community has a limited understanding of the access and reuse patterns of these files. This paper investigates the access and reuse patterns of I/O-intensive files on a production-scale supercomputer.

Patel, Tirthak↗

A parallel adaptive mesh refinement algorithm

Over recent years, Adaptive Mesh Refinement (AMR) algorithms which dynamically match the local resolution of the computational grid to the numerical solution being sought have emerged as powerful tools for solving problems that contain disparate length and time scales. In particular, several workers have demonstrated the effectiveness of employing an adaptive, block-structured hierarchical grid system for simulations of complex shock wave phenomena. Unfortunately, from the parallel algorithm developer's viewpoint, this class of scheme is quite involved; these schemes cannot be distilled down to a small kernel upon which various parallelizing strategies may be tested. However, because of their block-structured nature such schemes are inherently parallel, so all is not lost. In this paper we describe the method by which Quirk's AMR algorithm has been parallelized. This method is built upon just a few simple message passing routines and so it may be implemented across a broad class of MIMD machines. Moreover, the method of parallelization is such that the original serial code is left virtually intact, and so we are left with just a single product to support. The importance of this fact should not be underestimated given the size and complexity of the original algorithm.

Quirk, James J.↗

Machine Learning-Driven Conservative-to-Primitive Conversion in Hybrid Piecewise Polytropic and Tabulated Equations of State

We present a novel machine learning (ML)-based method to accelerate conservative-to-primitive inversion, focusing on hybrid piecewise polytropic and tabulated equations of state. Traditional root-finding techniques are computationally expensive, particularly for large-scale relativistic hydrodynamics simulations. To address this, we employ feedforward neural networks (NNC2PS and NNC2PL), trained in PyTorch (2.0+) and optimized for GPU inference using NVIDIA TensorRT (8.4.1), achieving significant speedups with minimal accuracy loss. The NNC2PS model achieves 𝐿 1 and 𝐿 ∞ errors of 4.54 × 10 −7 and 3.44 × 10−6, respectively, while the NNC2PL model exhibits even lower error values. TensorRT optimization with mixed-precision deployment substantially accelerates performance compared to traditional root-finding methods. Specifically, the mixed-precision TensorRT engine for NNC2PS achieves inference speeds approximately 400 times faster than a traditional single-threaded CPU implementation for a dataset size of 1,000,000 points. Ideal parallelization across an entire compute node in the Delta supercomputer (dual AMD 64-core 2.45 GHz Milan processors and 8 NVIDIA A100 GPUs with 40 GB HBM2 RAM and NVLink) predicts a 25-fold speedup for TensorRT over an optimally parallelized numerical method when processing 8 million data points. Moreover, the ML method exhibits sub-linear scaling with increasing dataset sizes. We release the scientific software developed, enabling further validation and extension of our findings. By exploiting the underlying symmetries within the equation of state, these findings highlight the potential of ML, combined with GPU optimization and model quantization, to accelerate conservative-to-primitive inversion in relativistic hydrodynamics simulations.

conservative-to-primitive conversion↗

Anisotropy of the Irreversibility Field for Zr-doped (Y,Gd)Ba 2<\sub>Cu 3<\sub>O<7-x<\sub> Thin Films up to 45 T

The anisotropic irreversibility fieldBIrr of twoYBa2Cu3O7 x thin films dopedwith additional rare earth (RE)= (Gd, Y) and Zr and containing strong correlated pins (splayed BaZrO3 nanorods and RE2O3 anoprecipitates) has been measured over a very broad range up to 45 T at temperatures 56 K < T < Tc. We found that the experimental angular dependence of BIrr ( ) does not follow the mass anisotropy scaling BIrr ( ) = BIrr (0)(cos2 + 2 sin2 ) 1/2, where = (mc/mab)1/2 = 5 6 for the RE-doped Ba2Cu3O7 x (REBCO) crystals, mab and mc are the effective masses along the ab plane and the c-axis, respectively, and is the angle between B and the c-axis. For B parallel to the ab planes and to the c-axis correlated pinning strongly enhances BIrr , while at intermediate angles, BIrr ( ) follows the scaling behavior BIrr ( ) (cos2 + 2 RP sin2 ) 1/2 with the effective anisotropy factor RP 3 significantly smaller than the ass anisotropy would suggest. In spite of the strong effects of c-axis BaZrO3 nanorods, we found even greater enhancements of BIrr for fields along the ab planes than for fields parallel to the c-axis, as well as different temperature dependences of the correlated pinning contributions to BIrr for B//ab and B//c. Our results show that the dense and strong pins, which can now be incorporated into REBCO thin films in a controlled way, exert major and diverse effects on the measured vortex pinning anisotropy and the irreversibility field over wide ranges of B and T . In particular, we show that the relative contribution of correlated pinning to BIrr for B//c increases as the temperature increases due to the suppression of thermal fluctuations of vortices by splayed distribution of BaZrO3 nanorods.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

The General Isothermal Oxidation Behavior of Cu-8Cr-4Nb

Oxidation kinetics of Cu-8Cr-4Nb was investigated by TGA (thermogravimetric) exposures between 500 and 900-C (at 25-50 C intervals) and the oxide scale morphologies examined by microscopy and micro-analysis. Because Cu-8Cr-4Nb is comprised of fine Cr2Nb precipitates in a Cu matrix, the results were interpreted by comparison with the behavior of copper (OFHC) and 'NARloy-Z' (a rival candidate material for thrust cell liner applications in advanced rocket engines) under the same conditions. While NARloy-Z and Cu exhibited identical oxidation behavior, Cu-8Cr-4Nb differed markedly in several respects: below approx. 700 C its oxidation rates were significantly lower than those of Cu; At higher temperatures its oxidation rates fell into two categories: an initial rate exceeding that of Cu, and a terminal rate comparable to that of Cu. Differences in oxide morphologies paralleled the kinetic differences at higher temperature: While NARloy-Z and Cu produced a uniform oxide scale of Cu2O inner layer and CuO outer layer, the inner (Cu2O) layer on Cu-8Cr-4Nb was stratified, with a highly porous/spongy inner stratum (responsible for the fast initial kinetics) and a dense/blocky outer stratum (corresponding to the slow terminal kinetics). Single and spinel oxides of Nb and Cr were found at the interface between the oxide scale and Cu-8Cr-4Nb substrate and it appears that these oxides were responsible for its suppressed oxidation rates at the intermediate temperatures. No difference was found between Cu-8Cr-4Nb oxidation in air and in oxygen at 1.0 atm.

Thomas-Ogbuji, L. U.↗

Scale-Dependent Fracture-Matrix Interactions and Their Impact on Radionuclide Transport: Development of efficient particle-tracking methods

Matrix Diffusion and Adsorption within a rock matrix are important mechanisms for retarding transport of radionuclides in fractured rock. Due to computational limitations and difficulties in characterizing complex subsurface systems, diffusive exchange between a fracture network and surrounding rock matrix is often modeled using simplified conceptual representations. There is significant uncertainty in “effective” parameters used in these models, such as the “effective matrix diffusivity”. Often, these parameters are estimated by fitting sparse breakthrough data, and estimated values fall outside meaningful ranges, because simplified interpretive models do not consider complex three-dimensional flow. There is limited understanding of the relationship between the effective parameters and rock mass characteristics including network structure and matrix properties. There is also evidence for an apparent scale-dependence in “effective matrix diffusion” coefficients. These observations raise questions on whether fracture-matrix interaction parameters estimated from small-scale tracer tests can be used for predicting radionuclide fate and transport at the scale of DOE field sites. High-resolution three-dimensional Discrete-Fracture-Network-Matrix (DFNM) models based on well-defined local scale transport equations can help to address some of these questions. Due to tremendous advances in computational technology over the last 10 years, DFNM modeling in relatively large domains is now feasible. The overarching objective of our research is to use DFNM modeling to improve fundamental understanding of how effective parameters in conceptual models are related to fracture network structure and matrix properties. An advanced three-dimensional DFNM model is being developed, which combines upscaled particle-tracking algorithms for fracture-matrix interaction and a parallel fracture-network flow simulator. The particle-tracking algorithms allow complexity in flow fields at different scales, and track transport across fracture-matrix interfaces based on rigorous local approximations to the transport equations. This modeling approach can incorporate aperture variability, multi-scale preferential flow and matrix heterogeneity. We developed efficient particle-tracking methods for handling matrix diffusion and adsorption on fracture walls and demonstrated their efficiency for use within the context of large-scale complex fracture network models with variability in apertures across a network of fractures and within individual fractures.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

A Review on the Coalescence of Confined Drops with a Focus on Scaling Laws for the Growth of the Liquid Bridge

The surface–tension-driven coalescence of drops has been extensively studied because of the omnipresence of the phenomenon and its significance in various natural and engineering systems. When two drops come into contact, a liquid bridge is formed between them and then grows in its lateral dimensions. As a result, the two drops merge to become a bigger drop. The growth dynamics of the bridge are governed by a balance between the driving force and the viscous and inertial resistances of involved liquids, and it is usually represented by power–law scaling relations on the temporal evolution of the bridge dimension. Such scaling laws have been well-characterized for the coalescence of unconfined or freely suspended drops. However, drops are often confined by solid or liquid surfaces and thus are a different shape from spheres, which affects their coalescence dynamics. As such, the coalescence of confined drops poses more complicated interfacial fluid dynamics challenges compared to that of unconfined drops. Although there have been several studies on the coalescence of confined drops, they have not been systematically reviewed in terms of the properties and geometry of the confining surface. Thus, we aim to review the current literature on the coalescence of confined drops in three categories: drop coalescence on a solid surface, drop coalescence on a deformable surface, and drop coalescence between two parallel surfaces with a small gap (i.e., Hele-Shaw cell), with a focus on power–law scaling relations, and to suggest challenges and outlooks for future research on the phenomena.

Sangjin Ryu↗

Regional heritability advanced complex trait analysis for GPU and traditional parallel architectures

Motivation: Quantification of the contribution of genetic variation to phenotypic variation for complex traits becomes increasingly computationally demanding with increasing numbers of single-nucleotide polymorphisms and individuals. To meet the challenges in making feasible large-scale studies, we present the REgional heritability advanced complex trait analysis software. Adapted from advanced complex trait analysis (and, in turn, genome-wide complex trait analysis), it is tailored to exploit the parallelism present in modern traditional and graphics processing unit (GPU)-accelerated machines, from workstations to supercomputers. Results: We adapt the genetic relationship matrix estimation algorithm to remove limitations on memory, allowing the analysis of large datasets. Further, we build on this to develop a version of the code able to efficiently exploit GPU-accelerated systems for both the genetic relationship matrix and REstricted maximum likelihood (REML) parts of the analysis, offering substantial speedup over the traditional central processing unit version. We develop the ability to analyze multiple small regions of the genome across multiple compute nodes in parallel, following the ‘regional heritability’ approach. We demonstrate the new software using 1024 GPUs in parallel on one of the world’s fastest supercomputers. Availability: The code is freely available at http://www.epcc.ed.ac.uk/software-products

59 BASIC BIOLOGICAL SCIENCES↗

Microscopic nuclear mass table with high-performance computing

We overview the methodology behind the large-scale mass table calculations based on the nuclear density functional theory (DFT) with Skyrme energy density functionals (EDFs). The calculations employing massively parallel computers are done in a large configuration space of an axially deformed harmonic oscillator. Nuclear mass tables tabulating global nuclear properties such as binding energies, radii, shape deformations, and pairing gaps are obtained for several Skyrme EDFs augmented by a mixed-type pairing functional, using an approximate particle number projection before variation. Specialty visualization tools have been developed to analyze the results. As an illustrative example of our current capabilities, we show some results for a wide range of even-even nuclei with Z ≤ 120 and N ≤ 300.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Study and Development of an Air Conditioning System Operating on a Magnetic Heat Pump Cycle

This report describes the design of a laboratory scale demonstration prototype of an air conditioning system operating on a magnetic heat pump cycle. Design parameters were selected through studies performed by a Kennedy Space Center (KSC) System Simulation Computer Model. The heat pump consists of a rotor turning through four magnetic fields that are created by permanent magnets. Gadolinium was selected as the working material for this demonstration prototype. The rotor was designed to be constructed of flat parallel disks of gadolinium with very little space in between. The rotor rotates in an aluminum housing. The laboratory scale demonstration prototype is designed to provide a theoretical Carnot Cycle efficiency of 62 percent and a Coefficient of Performance of 16.55.

Wang, Pao-Lien↗

Refine your search to explore more results.