Search NASA⌕ Search

SEARCH · Search NASA

Results for “Cluster Computing”

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 523 records · Page 29
All sourcesNASA NTRSDOE OSTI
DOE OSTI2016

Multi-node and Multi-core Performance Studies of a Monte Carlo code RMC

Over the past 15 years the architecture of high-performance computing systems has become increasingly heterogeneous with individual computing cores being subdivided into nodes and even auxiliary co-processing units such as graphics processing units or the Intel many integrated core co-processors. In fact, future computer platforms is moving toward a larger number of nodes and cores with lower memory available per node. Today, several production Monte Carlo codes use some types of hybrid parallelism strategy that makes use of shared-memory parallelism within a computing node and message-passing parallelism between nodes, which enables the codes to utilize available computing resources as much as possible, both among and within computing nodes on a cluster. The so-called hybrid parallelization is also possible with Reactor Monte Carlo Code RMC by activating both MPI and OpenMP parallel algorithms together. On HPC' which typically consists of several or several-ten thousand nodes equipping 16-32 computing intra-cores per node, simulation of RMC can be carried out in the hybrid parallelization with the MPI inter-node parallelization and the OpenMP intra-node parallelization. RMC has adopted a hybrid message-passing and shared-memory approach to parallelism, which enables the code to utilize available computing resources, both among and within computing nodes on a cluster. Shared-memory parallelism is especially valuable for large-model simulations because it enables multiple compute cores within a node to share common model information such as the geometry description, material, cross section and tally data, thus reducing the memory usage per processor. The calculation on the Tianhe-2 supercomputer took 6144 cores, resulting in a rate of 173,314,000 neutrons/minute for BEAVRS benchmark. While pin-wise full core H-M benchmark burnup calculations with millions of depletion regions was selected using MPI/OpenMP parallelism, the speed of 6,144 cores reach 1,843,200 burnable regions per minute. Keeping up with the continuing trend for increases in the number of computing resources within modern HPC systems is a significant challenge for teams that develop and maintain high-performance MC radiation transport solvers. order to address this challenge, RMC has adopted a hybrid message-passing and shared-memory approach to parallelism, which enables the code to utilize available computing resources, both among and within computing nodes on a cluster. Shared-memory parallelism is especially valuable for large-model simulations because it enables multiple compute cores within a node to share common model information such as the geometry description, material and composition definitions, and cross section data, thus reducing the memory usage per processor. The parallel performance has been studied by using two complex and representative benchmarks, H-M and BEAVRS benchmark. All simulations were run on Intel Xeon E5-2692 12 core, 2.2 GHz processors at MilkyWay-2 Super Computer with up to 12,288 available computing cores. While employing MPI parallelism, excellent parallel efficiency is achieved even above 12,000 processors and RMC reaches the parallel efficiency of 96% on 12,288 cores for the pin-by-pin computation of the BEAVRS benchmark. While BEAVRS criticality benchmark was selected using MPI/OpenMP parallelism, the speed of 6,144 cores reach 173,314,000 particles per minute. While pin-wise full core H-M benchmark burnup calculations with millions of depletion regions was selected using MPI/OpenMP parallelism, the speed of 6,144 cores reach 1,843,200 burnable regions per minute. Results from these studies demonstrate that RMC, using the hybrid parallel model described in this paper, is able to handle pin-wise full core burnup calculations with millions of depletion regions and scales well through thousands of processors. (authors)

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2017

Computing eddy-driven effective diffusivity using Lagrangian particles

A novel method to derive effective diffusivity from Lagrangian particle trajectory data sets is developed and then analyzed relative to particle-derived meridional diffusivity for eddy-driven mixing in an idealized circumpolar current. Quantitative standard dispersion- and transport-based mixing diagnostics are defined, compared and contrasted to motivate the computation and use of effective diffusivity derived from Lagrangian particles. We compute the effective diffusivity by first performing scalar transport on Lagrangian control areas using stored trajectories computed from online Lagrangian In-situ Global High-performance particle Tracking (LIGHT) using the Model for Prediction Across Scales Ocean (MPAS-O). Furthermore, the Lagrangian scalar transport scheme is compared against an Eulerian scalar transport scheme. Spatially-variable effective diffusivities are computed from resulting time-varying cumulative concentrations that vary as a function of cumulative area. The transport-based Eulerian and Lagrangian effective diffusivity diagnostics are found to be qualitatively consistent with the dispersion-based diffusivity. All diffusivity estimates show a region of increased subsurface diffusivity within the core of an idealized circumpolar current and results are within a factor of two of each other. The Eulerian and Lagrangian effective diffusivities are most similar; smaller and more spatially diffused values are obtained with the dispersion-based diffusivity computed with particle clusters.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2016

Photoconductivities from band states and a dissipative electron dynamics: Si(111) without and with adsorbed Ag clusters

A new general computational procedure is presented to obtain photoconductivities starting from atomic structures, combining ab initio electronic energy band states with populations from density matrix theory, and implemented for a specific set of materials based on Si crystalline slabs and their nanostructured surfaces without and with adsorbed Ag clusters. The procedure accounts for charge mobility in semiconductors in photoexcited states, and specifically electron and hole photomobilities at Si(111) surfaces with and without adsorbed Ag clusters using ab initio energy bands and orbitals generated from a generalized gradient functional, however with excited energy levels modified to provide correct bandgaps. Photoexcited state populations for each band and carrier type were generated using steady state solution of a reduced density matrix which includes dissipative medium effects. The present calculations provide photoexcited electronic populations and photoinduced mobilities resulting from applied electric fields and obtained from the change of driven electron energies with their electronic momentum. Extensive results for Si slabs with 8 layers, without and with adsorbed Ag clusters, show that the metal adsorbates lead to substantial increases in the photomobility and photoconductivity of electrons and holes.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

How Cloud is Accelerating Research at NREL

This presentation coincides with AWS's announcement of their new Parallel Computing Service (PCS) which allows for easy creation of HPC-style clusters in their AWS cloud computing platform. I helped them beta test this service before it was made generally available in August. AWS asked if we would be interested in discussing our experience with the PCS service, and our experience with HPC workloads in the cloud in general, so this slideshow discusses a brief history of scientific computing at NREL and shares a bit of our experiences and approach to utilizing cloud services for HPC-style workloads.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1994

Convergence rate enhancement of navier-stokes codes on clustered grids

Our Sensitivity-Based Minimal Residual (SBMR) method which is based on our earlier Distributed Minimal Residual (DMR) method allows each component of the solution vector in a system of equations to have its own convergence speed. Our global SBMR method was found to consistently outperform the DMR method while requiring considerably less computer memory. Recently, we have developed and tested a new Line SBMR or LSBMR method and a Time-Step-Scaling (TSS) method that are even more robust and computationally efficient than our global SBMR method, especially on highly clustered computational grids in laminar and turbulent flow computations.

Choi, Kwang-Yoon↗
DOE OSTI2021

Oxidation Dynamics of Supported Catalytic Cu Clusters: Coupling to Fluxionality

Copper oxide nanoclusters have a wide range of catalytic applications, such as the selective oxidation of hydrocarbons. O 2 binding to the catalyst, activation, and release upon reagent oxidation are key events in these catalytic chemistries. Furthermore, these events are expected to be accompanied by significant structural changes of the Cu clusters, because O atoms integrate into the cluster, rather than bind to its surface. Topping the complexity of the problem, partially oxidized Cu clusters are known to exhibit strong fluxionality and feature diverse and interconverting structures and oxygen contents in conditions of oxidative dehydrogenation (ODH). Hence, a significant dynamic coupling between the “hot” O 2 molecule impacting the cluster at reaction temperatures and the cluster fluxionality can be expected. In this work, we focus on the dynamics of dioxygen integration into a partially oxidized Cu cluster supported on hydroxylated amorphous alumina–a system recently reported to be an exceptionally selective catalyst for cyclohexane ODH with very little CO produced, whose mechanistic underpinnings are of utmost interest. The statistics over a swarm of adsorption and scattering trajectories where O 2 hits various sites on the cluster at reaction temperature shows that the O 2 binding does not only follow the minimal energy paths. O 2 also rarely integrates into the cluster in a single step and instead first binds to a single Cu atom via either an η 1 -O 2 or an η 2 -O 2 mode. Surprisingly, this step often has a higher barrier than the subsequent O 2 integration and dissociation, which in turn take multiple steps and complete the oxidation process. Dynamic trajectories starting from the key transition state of integration of the adsorbed O 2 can also lead to different intermediate structures during or right after the dissociation, due to the energy released from the transition state and the thermal intracluster effects. From these activated O 2 chemisorbed structures, O 2 dissociation occurs with moderate barriers (~0.5 eV), producing multiple final oxidized Cu 4 O 4 states. Hence, a diversity of reaction profiles for the attack of supported Cu cluster by O 2 emerges due to the dynamic effects, with implications for mechanisms, kinetic models, and catalyst design principles.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2016

Electric dipole polarizability from first principles calculations

The electric dipole polarizability quantifies the low-energy behavior of the dipole strength and is related to critical observables such as the radii of the proton and neutron distributions. Its computation is challenging because most of the dipole strength lies in the scattering continuum. In our paper we combine integral transforms with the coupled-cluster method and compute the dipole polarizability using bound-state techniques. Furthermore, employing different interactions from chiral effective field theory, we confirm the strong correlation between the dipole polarizability and the charge radius, and study its dependence on three-nucleon forces. Finally, we find good agreement with data for the 4 He, 40 Ca, and 16 O nuclei, and predict the dipole polarizability for the rare nucleus 22 O.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
NASA NTRS2005

A Fast Implementation of the ISODATA Clustering Algorithm

Clustering is central to many image processing and remote sensing applications. ISODATA is one of the most popular and widely used clustering methods in geoscience applications, but it can run slowly, particularly with large data sets. We present a more efficient approach to ISODATA clustering, which achieves better running times by storing the points in a kd-tree and through a modification of the way in which the algorithm estimates the dispersion of each cluster. We also present an approximate version of the algorithm which allows the user to further improve the running time, at the expense of lower fidelity in computing the nearest cluster center to each point. We provide both theoretical and empirical justification that our modified approach produces clusterings that are very similar to those produced by the standard ISODATA approach. We also provide empirical studies on both synthetic data and remotely sensed Landsat and MODIS images that show that our approach has significantly lower running times.

Memarsadeghi, Nargess↗
NASA NTRS2007

A Fast Implementation of the Isodata Clustering Algorithm

Clustering is central to many image processing and remote sensing applications. ISODATA is one of the most popular and widely used clustering methods in geoscience applications, but it can run slowly, particularly with large data sets. We present a more efficient approach to IsoDATA clustering, which achieves better running times by storing the points in a kd-tree and through a modification of the way in which the algorithm estimates the dispersion of each cluster. We also present an approximate version of the algorithm which allows the user to further improve the running time, at the expense of lower fidelity in computing the nearest cluster center to each point. We provide both theoretical and empirical justification that our modified approach produces clusterings that are very similar to those produced by the standard ISODATA approach. We also provide empirical studies on both synthetic data and remotely sensed Landsat and MODIS images that show that our approach has significantly lower running times.

Memarsadeghi, Nargess↗
DOE OSTI2017

Kaposi’s sarcoma–associated herpesvirus stably clusters its genomes across generations to maintain itself extrachromosomally

Genetic elements that replicate extrachromosomally are rare in mammals; however, several human tumor viruses, including the papillomaviruses and the gammaherpesviruses, maintain their plasmid genomes by tethering them to cellular chromosomes. We have uncovered an unprecedented mechanism of viral replication: Kaposi’s sarcoma–associated herpesvirus (KSHV) stably clusters its genomes across generations to maintain itself extrachromosomally. To identify and characterize this mechanism, we developed two complementary, independent approaches: live-cell imaging and a predictive computational model. The clustering of KSHV requires the viral protein, LANA1, to bind viral genomes to nucleosomes arrayed on both cellular and viral DNA. Clustering affects both viral partitioning and viral genome numbers of KSHV. The clustering of KSHV plasmids provides it with an effective evolutionary strategy to rapidly increase copy numbers of genomes per cell at the expense of the total numbers of cells infected.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2014

The next generation Virgo cluster survey. VIII. The spatial distribution of globular clusters in the Virgo cluster

We report on a large-scale study of the distribution of globular clusters (GCs) throughout the Virgo cluster, based on photometry from the Next Generation Virgo Cluster Survey (NGVS), a large imaging survey covering Virgo's primary subclusters (Virgo A = M87 and Virgo B = M49) out to their virial radii. Using the g{sub o}{sup ′}, (g' – i') {sub o} color-magnitude diagram of unresolved and marginally resolved sources within the NGVS, we have constructed two-dimensional maps of the (irregular) GC distribution over 100 deg{sup 2} to a depth of g{sub o}{sup ′} = 24. We present the clearest evidence to date showing the difference in concentration between red and blue GCs over the full extent of the cluster, where the red (more metal-rich) GCs are largely located around the massive early-type galaxies in Virgo, while the blue (metal-poor) GCs have a much more extended spatial distribution with significant populations still present beyond 83' (∼215 kpc) along the major axes of both M49 and M87. A comparison of our GC maps to the diffuse light in the outermost regions of M49 and M87 show remarkable agreement in the shape, ellipticity, and boxiness of both luminous systems. We also find evidence for spatial enhancements of GCs surrounding M87 that may be indicative of recent interactions or an ongoing merger history. We compare the GC map to that of the locations of Virgo galaxies and the X-ray intracluster gas, and find generally good agreement between these various baryonic structures. We calculate the Virgo cluster contains a total population of N {sub GC} = 67, 300 ± 14, 400, of which 35% are located in M87 and M49 alone. For the first time, we compute a cluster-wide specific frequency S {sub N,} {sub CL} = 2.8 ± 0.7, after correcting for Virgo's diffuse light. We also find a GC-to-baryonic mass fraction ε {sub b} = 5.7 ± 1.1 × 10{sup –4} and a GC-to-total cluster mass formation efficiency ε {sub t} = 2.9 ± 0.5 × 10{sup –5}, the latter values slightly lower than but consistent with those derived for individual galactic halos. Taken as a whole, our results show that the production of the complex structures in the unrelaxed Virgo cluster core (including the production of the diffuse intracluster light) is an ongoing and continuing process.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

DECT-MULTRA: Dual-Energy CT Image Decomposition With Learned Mixed Material Models and Efficient Clustering

Dual-energy computed tomography (DECT) imaging plays an important role in advanced imaging applications due to its material decomposition capability. Image-domain decomposition operates directly on CT images using linear matrix inversion, but the decomposed material images can be severely degraded by noise and artifacts. This paper proposes a new method dubbed DECT-MULTRA for image-domain DECT material decomposition that combines conventional penalized weighted-least squares (PWLS) estimation with regularization based on a mixed union of learned transforms (MULTRA) model. Our proposed approach pre-learns a union of common-material sparsifying transforms from patches extracted from all the basis materials, and a union of cross-material sparsifying transforms from multi-material patches. The common-material transforms capture the common properties among different material images, while the cross-material transforms capture the cross-dependencies. The proposed PWLS formulation is optimized efficiently by alternating between an image update step and a sparse coding and clustering step, with both of these steps having closed-form solutions. The effectiveness of our method is validated with both XCAT phantom and clinical head data. The results demonstrate that our proposed method provides superior material image quality and decomposition accuracy compared to other competing methods.

47 OTHER INSTRUMENTATION↗
DOE OSTI2020

CLUE: A Fast Parallel Clustering Algorithm for High Granularity Calorimeters in High-Energy Physics

One of the challenges of high granularity calorimeters, such as that to be built to cover the endcap region in the CMS Phase-2 Upgrade for HL-LHC, is that the large number of channels causes a surge in the computing load when clustering numerous digitized energy deposits (hits) in the reconstruction stage. In this article, we propose a fast and fully parallelizable density-based clustering algorithm, optimized for high-occupancy scenarios, where the number of clusters is much larger than the average number of hits in a cluster. The algorithm uses a grid spatial index for fast querying of neighbors and its timing scales linearly with the number of hits within the range considered. We also show a comparison of the performance on CPU and GPU implementations, demonstrating the power of algorithmic parallelization in the coming era of heterogeneous computing in high-energy physics.

Rovere, Marco↗
DOE OSTI2021

Failure of molecular dynamics to provide appropriate structures for quantum mechanical description of the aqueous chloride ion charge-transfer-to-solvent ultraviolet spectrum

The lowest band in the charge-transfer-to-solvent ultraviolet absorption spectrum of aqueous chloride ion is studied by experiment and computation. Interestingly, the experiments indicate that at concentrations up to at least 0.25 M, where calculations indicate ion pairing to be significant, there is no notable effect of ionic strength on the spectrum. The experimental spectra are fitted to aid comparison with computations. Classical molecular dynamic simulations are carried out on dilute aqueous Cl - , Na + , and NaCl, producing radial distribution functions in reasonable agreement with experiment and, for NaCl, clear evidence of ion pairing. Clusters are extracted from the simulations for quantum mechanical excited state calculations. Accurate ab initio coupled-cluster benchmark calculations on a small number of representative clusters are carried out and used to identify and validate an efficient protocol based on time-dependent density functional theory. The latter is used to carry out quantum mechanical calculations on thousands of clusters. The resulting computed spectrum is in excellent agreement with experiment for the peak position, with little influence from ion pairing, but is in qualitative disagreement on the width, being only about half as wide. It is concluded that simulation by classical molecular dynamics fails to provide an adequate variety of structures to explain the experimental CTTS spectrum of aqueous Cl - .

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2015

Swelling In Light Water Reactor Internal Components: Insights from Computational Modeling

A modern cluster dynamics model has been used to investigate the materials and irradiation parameters that control microstructural evolution under the relatively low-temperature exposure conditions that are representative of the operating environment for in-core light water reactor components. The focus is on components fabricated from austenitic stainless steel. The model accounts for the synergistic interaction between radiation-produced vacancies and the helium that is produced by nuclear transmutation reactions. Cavity nucleation rates are shown to be relatively high in this temperature regime (275 to 325°C), but are sensitive to assumptions about the fine scale microstructure produced under low-temperature irradiation. The cavity nucleation rates observed run counter to the expectation that void swelling would not occur under these conditions. This expectation was based on previous research on void swelling in austenitic steels in fast reactors. This misleading impression arose primarily from an absence of relevant data. The results of the computational modeling are generally consistent with recent data obtained by examining ex-service components. However, it has been shown that the sensitivity of the model s predictions of low-temperature swelling behavior to assumptions about the primary damage source term and specification of the mean-field sink strengths is somewhat greater that that observed at higher temperatures. Further assessment of the mathematical model is underway to meet the long-term objective of this research, which is to provide a predictive model of void swelling at relevant lifetime exposures to support extended reactor operations.

36 MATERIALS SCIENCE↗
DOE OSTI2015

Low-temperature swelling in LWR internal components: a computational assessment

A modern cluster dynamics model has been used to investigate the materials and irradiation parameters that control microstructural evolution under the relatively low-temperature exposure conditions that are representative of the operating environment for in-core light water reactor components. The focus is on components fabricated from austenitic stainless steel. The model accounts for the synergistic interaction between radiation-produced vacancies and the helium that is produced by nuclear transmutation reactions. Cavity nucleation rates are shown to be relatively high in this temperature regime (275 to 325 °C), but are sensitive to assumptions about the fine-scale microstructure produced under low-temperature irradiation. The cavity nucleation rates observed run counter to the expectation that void swelling would not occur under these conditions. This expectation was based on previous research on void swelling in austenitic steels in fast reactors. This misleading impression arose primarily from an absence of relevant data. The results of the computational modeling are generally consistent with recent data obtained by examining ex-service components. The long-term objective of this research is to provide a predictive model of void swelling at relevant lifetime exposures to support extended reactor operations.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2015

The benzene⋯naphthalene complex: A more challenging system than the benzene dimer for newly developed computational methods

High-level coupled cluster singles, doubles, and perturbative triples [CCSD(T)] computations with up to the aug-cc-pVQZ basis set (1924 basis functions) and various extrapolations toward the complete basis set (CBS) limit are presented for the sandwich, T-shaped, and parallel-displaced benzene⋯naphthalene complex. Using the CCSD(T)/CBS interaction energies as a benchmark, the performance of some newly developed wave function and density functional theory methods has been evaluated. The best performing methods were found to be the dispersion-corrected PBE0 functional (PBE0-D3) and spin-component scaled zeroth-order symmetry-adapted perturbation theory (SCS-SAPT0). The success of SCS-SAPT0 is very encouraging because it provides one method for energy component analysis of π-stacked complexes with 200 atoms or more. Most newly developed methods do, however, overestimate the interaction energies. The results of energy component analysis show that interaction energies are overestimated mainly due to the overestimation of dispersion energy.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2018

HipMCL: a high-performance parallel implementation of the Markov clustering algorithm for large-scale networks

Biological networks capture structural or functional properties of relevant entities such as molecules, proteins or genes. Characteristic examples are gene expression networks or protein–protein interaction networks, which hold information about functional affinities or structural similarities. Such networks have been expanding in size due to increasing scale and abundance of biological data. While various clustering algorithms have been proposed to find highly connected regions, Markov Clustering (MCL) has been one of the most successful approaches to cluster sequence similarity or expression networks. Despite its popularity, MCL’s scalability to cluster large datasets still remains a bottleneck due to high running times and memory demands. In this paper, we present High-performance MCL (HipMCL), a parallel implementation of the original MCL algorithm that can run on distributed-memory computers. We show that HipMCL can efficiently utilize 2000 compute nodes and cluster a network of ~70 million nodes with ~68 billion edges in ~2.4 h. By exploiting distributed-memory environments, HipMCL clusters large-scale networks several orders of magnitude faster than MCL and enables clustering of even bigger networks. Finally, HipMCL is based on MPI and OpenMP and is freely available under a modified BSD license.

59 BASIC BIOLOGICAL SCIENCES↗