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.

SourceResearch connectionFrom yearThrough year

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

At least 343 records · Page 19

All sources NASA NTRS DOE OSTI

NASA NTRS2005

Testing the Archivas Cluster (Arc) for Ozone Monitoring Instrument (OMI) Scientific Data Storage

The Ozone Monitoring Instrument (OMI) launched on NASA's Aura Spacecraft, the third of the major platforms of the EOS program on July 15,2004. In addition to the long term archive and distribution of the data from OM1 through the Goddard Earth Science Distributed Active Archive Center (GESDAAC), we are evaluating other archive mechanisms that can archive the data in a more immediately available method where it can be used for futher data production and analysis. In 2004, Archivas, Inc. was selected by NASA s Small Business Innovative Research (SBIR) program for the development of their Archivas Cluster (ArC) product. Arc is an online disk based system utilizing self-management and automation on a Linux cluster. Its goal is to produce a low cost solution coupled with the ease of management. The OM1 project is an application partner of the SBIR program, and has deployed a small cluster (5TB) based on the beta Archwas software. We performed extensive testing of the unit using production OM1 data since launch. In 2005, Archivas, Inc. was funded in SBIR Phase II for further development, which will include testing scalability with the deployment of a larger (35TB) cluster at Goddard. We plan to include Arc in the OM1 Team Leader Computing Facility (TLCF) hosting OM1 data for direct access and analysis by the OMI Science Team. This presentation will include a brief technical description of the Archivas Cluster, a summary of the SBIR Phase I beta testing results, and an overview of the OMI ground data processing architecture including its interaction with the Phase II Archivas Cluster and hosting of OMI data for the scientists.

Tilmes, Curt↗

NASA NTRS1994

Emission lines from condensing intracluster gas

We present computations of the emission-line spectra produced when hot gas cools nonuniformly. During inhomogeneous cooling, soft X-ray/extreme ultraviolet radiation from 10(exp 5) to 10(exp 7) K gas photoionizes coexisting clouds that have already cooled to 10(exp 4) K. Thus, strong emission lines emerge from both the high-ionization cooling gas and embedded low-ionization knots. This mechanism, which ought to operate naturally within the sonic radii of cooling flows, can generate optical emission-line spectra similar to those observed at the centers of many clusters of galaxies thought to contain such flows. We have computed the X-ray, ultraviolet (UV), optical, and infrared (IR) emission-line fluxes expected under a variety of intracluster conditions so that the self-irradiation hypothesis can be compared thoroughly with observations. If this picture is found to be adequate, the models can also be used to measure mass flow rates and inflow velocities in the vicinities of cooling-flow nebulae and to constrain the column densities and covering factors of condensed material at the centers of these clusters.

Voit, G. Mark↗

NASA NTRS1972

The estimation of masses of individual galaxies in clusters of galaxies.

Three different methods of estimating masses are discussed. The 'density method' is based on the analysis of the density distribution of galaxies around the object whose mass is to be found. The 'bound-galaxy method' gives estimates of the mass of a double, triple, or quadruple system from analysis of the orbital motion of the components. The 'virial method' utilizes the formulas derived for the second method to obtain estimates of the virial-theorem masses of whole clusters, and thus to obtain upper limits on the mass of an individual galaxy in a cluster. The analytic formulas are developed and compared with computer experiments, and some applications are given.

Wolf, R. A.↗

NASA NTRS1982

Theoretical studies of the electronic structure of small metal clusters

Theoretical studies of the electronic structure of metal clusters, in particular clusters of Group IIA and IIB atoms were conducted. Early in the project it became clear that electron correlation involving d orbitals plays a more important role in the binding of these clusters than had been previously anticipated. This necessitated that computer codes for calculating two electron integrals and for constructing the resulting CI Hamiltonions be replaced with newer, more efficient procedures. Program modification, interfacing and testing were performed. Results of both plans are reported.

Jordan, K. D.↗

NASA NTRS1987

Strake-generated vortex interactions for a fighter-like configuration

A combination of the Euler, and thin-layer Navier-Stokes equations were solved for the flowfield around a wing-strake-fuselage configuration similar to the F-16A aircraft. The flowfield was divided into multiple blocks using a zoning program, and the TNSWF solution program, based on an implicit solution algorithm, updates each subregion of the flowfield sequentially. The total grid contained approximately 300,000 grid points, and clustering was effected normal to all aircraft surfaces. The computed solutions agreed well quantitatively with wind-tunnel pressure distribution data, and they agreed well qualitatively with flow visualization from wind- and water-tunnel tests.

Reznick, Steven G.↗

NASA NTRS1987

Simulation of transonic viscous flow over a fighter-like configuration including inlet

The simulation of transonic viscous flow over a modified F-16A including inlet is presented. A zonal approach is utilized which allows appropriate clustering suitable for viscous calculations on all solid surfaces. Computational efficiency is enhanced by solving the thin-layer Navier-Stokes equations in viscous zones adjacent to the aircraft and the Euler equations on those coarse zones away from the aircraft. The flow conditions for this transonic case are freestream Mach number of 0.9, angle of attack of 4.12 deg, and a Reynolds number (based on root chord) of 4.5 million. A total of 19 zones are utilized yielding a total of 350,000 grid points. This case required about 3000 iterations to reduce the residual by three orders, which takes about 10 hr of CPU time on the Cray X-MP/48 computer. Pressure distributions on the wing and on cross sections through the inlet region compare favorably with the experimental data for this transonic case.

Flores, Jolen↗

NASA NTRS1990

The morphology of star clusters in the SMC

The projected ellipticities of 34 populous SMC star clusters have been derived by means of PDS 1010A scans and a computer interactive method of reduction implemented on an Apollo 570 workstation. A pair of J and R plates taken with the 1.2 m UK Schmidt telescope in Australia were used. Radial ellipticity variations within individual globular clusters seem to be a common phenomenon for the SMC clusters, similar to that observed in the LMC clusters where the innerparts are more elliptical than the outer ones in 95 percent of the cases. The derived ellipticities which correspond to the innermost part of the cluster at radial distances near to half-mass radii have been found to be statistically more elliptical than those of the LMC, known to be more elliptical than those of the Galaxy. The dynamical masses of the clusters seem to correlate with ellipticities supporting the hypothesis that, either the gravitational field of the parent galaxy being a dominant factor affect slower the shape of the high mass clusters and/or the most massive clusters, being dynamically younger, retain their original shape.

Kontizas, E.↗

NASA NTRS1999

Shared Memory Parallelization of an Implicit ADI-type CFD Code

A parallelization study designed for ADI-type algorithms is presented using the OpenMP specification for shared-memory multiprocessor programming. Details of optimizations specifically addressed to cache-based computer architectures are described and performance measurements for the single and multiprocessor implementation are summarized. The paper demonstrates that optimization of memory access on a cache-based computer architecture controls the performance of the computational algorithm. A hybrid MPI/OpenMP approach is proposed for clusters of shared memory machines to further enhance the parallel performance. The method is applied to develop a new LES/DNS code, named LESTool. A preliminary DNS calculation of a fully developed channel flow at a Reynolds number of 180, Re(sub tau) = 180, has shown good agreement with existing data.

Hauser, Th.↗

NASA NTRS2010

Comprehensive T-Matrix Reference Database: A 2007-2009 Update

The T-matrix method is among the most versatile, efficient, and widely used theoretical techniques for the numerically exact computation of electromagnetic scattering by homogeneous and composite particles, clusters of particles, discrete random media, and particles in the vicinity of an interface separating two half-spaces with different refractive indices. This paper presents an update to the comprehensive database of T-matrix publications compiled by us previously and includes the publications that appeared since 2007. It also lists several earlier publications not included in the original database.

Mishchenko, Michael I.↗

NASA NTRS1992

A fuzzy clustering algorithm to detect planar and quadric shapes

In this paper, we introduce a new fuzzy clustering algorithm to detect an unknown number of planar and quadric shapes in noisy data. The proposed algorithm is computationally and implementationally simple, and it overcomes many of the drawbacks of the existing algorithms that have been proposed for similar tasks. Since the clustering is performed in the original image space, and since no features need to be computed, this approach is particularly suited for sparse data. The algorithm may also be used in pattern recognition applications.

Krishnapuram, Raghu↗

NASA NTRS2000

PVM Support for Clusters

The latest version of PVM (3.4.3) now contains support for a PC cluster running Linux, also known as a Beowulf system. A PVM user of a computer outside the Beowulf system can add the Beowulf as a single machine.

Beowulf PVM cluster interactive supercomputing↗

NASA NTRS1993

A numerical investigation of a subsonic jet in a crossflow

The flowfield induced by a single, subsonic jet exhausting perpendicularly from a flat plate into a subsonic crossflow has been numerically investigated. The test case was chosen to match available experimental data where the jet Mach number was 0.78, and the freestream Mach number was 0.13. Time-averaged solutions were obtained using the thin-layer Navier-Stokes equations and two overlapping grids. The solutions were sensitive to the radial grid clustering near the edge of the jet and to the far-field boundary conditions. Experimental data comparisons were required to determine the most appropriate jet grid and satisfactory boundary conditions. Globally, the solutions converged in about 6000 iterations. The computational results accurately showed the deflected jet and associated contrarotating vortices. The fine clustered grid in the region upstream of the jet exit allowed the horseshoe vortex in the boundary layer near the jet exit to be captured. Most importantly for aircraft applications, the computed plate pressure distributions compared favorably with the experimental data over most of the surface. However, in the wake region immediately downstream of the jet exit, where there is extensive flow separation, some discrepancies with experimental data were observed. Two turbulence models were used in this study: (1) the zero-equation, two layer Baldwin-Lomax turbulence model; and (2) one-equation Baldwin-Barth turbulence model. The turbulence models gave results which generally compared no better with experimental data than the laminar computation results.

Chiu, Stephen H.↗

NASA NTRS1995

Nonlinear Fluid Computations in a Distributed Environment

The performance of a loosely and tightly-coupled workstation cluster is compared against a conventional vector supercomputer for the solution the Reynolds- averaged Navier-Stokes equations. The application geometries include a transonic airfoil, a tiltrotor wing/fuselage, and a wing/body/empennage/nacelle transport. Decomposition is of the manager-worker type, with solution of one grid zone per worker process coupled using the PVM message passing library. Task allocation is determined by grid size and processor speed, subject to available memory penalties. Each fluid zone is computed using an implicit diagonal scheme in an overset mesh framework, while relative body motion is accomplished using an additional worker process to re-establish grid communication.

Atwood, Christopher A.↗

NASA NTRS1979

Atmospheric effects on cluster analyses

Ground reflected radiance, from which information is extracted through techniques of cluster analyses for remote sensing application, is altered by the atmosphere when it reaches the satellite. Therefore it is essential to understand the effects of the atmosphere on Landsat measurements, cluster characteristics and analysis accuracy. A doubling model is employed to compute the effective reflectivity, observed from the satellite, as a function of ground reflectivity, solar zenith angle and aerosol optical thickness for standard atmosphere. The relation between the effective reflectivity and ground reflectivity is approximately linear. It is shown that for a horizontally homogeneous atmosphere, the classification statistics from a maximum likelihood classifier remains unchanged under these transforms. If inhomogeneity is present, the divergence between clusters is reduced, and correlation between spectral bands increases. Radiance reflected by the background area surrounding the target may also reach the satellite. The influence of background reflectivity on effective reflectivity is discussed.

Kiang, R. K.↗

NASA NTRS1995

Visualization of unsteady computational fluid dynamics

The current computing environment that most researchers are using for the calculation of 3D unsteady Computational Fluid Dynamic (CFD) results is a super-computer class machine. The Massively Parallel Processors (MPP's) such as the 160 node IBM SP2 at NAS and clusters of workstations acting as a single MPP (like NAS's SGI Power-Challenge array) provide the required computation bandwidth for CFD calculations of transient problems. Work is in progress on a set of software tools designed specifically to address visualizing 3D unsteady CFD results in these super-computer-like environments. The visualization is concurrently executed with the CFD solver. The parallel version of Visual3, pV3 required splitting up the unsteady visualization task to allow execution across a network of workstation(s) and compute servers. In this computing model, the network is almost always the bottleneck so much of the effort involved techniques to reduce the size of the data transferred between machines.

Haimes, Robert↗

NASA NTRS1989

NETRA: A parallel architecture for integrated vision systems 2: Algorithms and performance evaluation

In part 1 architecture of NETRA is presented. A performance evaluation of NETRA using several common vision algorithms is also presented. Performance of algorithms when they are mapped on one cluster is described. It is shown that SIMD, MIMD, and systolic algorithms can be easily mapped onto processor clusters, and almost linear speedups are possible. For some algorithms, analytical performance results are compared with implementation performance results. It is observed that the analysis is very accurate. Performance analysis of parallel algorithms when mapped across clusters is presented. Mappings across clusters illustrate the importance and use of shared as well as distributed memory in achieving high performance. The parameters for evaluation are derived from the characteristics of the parallel algorithms, and these parameters are used to evaluate the alternative communication strategies in NETRA. Furthermore, the effect of communication interference from other processors in the system on the execution of an algorithm is studied. Using the analysis, performance of many algorithms with different characteristics is presented. It is observed that if communication speeds are matched with the computation speeds, good speedups are possible when algorithms are mapped across clusters.

Choudhary, Alok N.↗

NASA NTRS2013

Reducing Earth Topography Resolution for SMAP Mission Ground Tracks Using K-Means Clustering

The K-means clustering algorithm is used to reduce Earth topography resolution for the SMAP mission ground tracks. As SMAP propagates in orbit, knowledge of the radar antenna footprints on Earth is required for the antenna misalignment calibration. Each antenna footprint contains a latitude and longitude location pair on the Earth surface. There are 400 pairs in one data set for the calibration model. It is computationally expensive to calculate corresponding Earth elevation for these data pairs. Thus, the antenna footprint resolution is reduced. Similar topographical data pairs are grouped together with the K-means clustering algorithm. The resolution is reduced to the mean of each topographical cluster called the cluster centroid. The corresponding Earth elevation for each cluster centroid is assigned to the entire group. Results show that 400 data points are reduced to 60 while still maintaining algorithm performance and computational efficiency. In this work, sensitivity analysis is also performed to show a trade-off between algorithm performance versus computational efficiency as the number of cluster centroids and algorithm iterations are increased.

ground tracks↗

NASA NTRS2010

Initialization and Restart in Stochastic Local Search: Computing a Most Probable Explanation in Bayesian Networks

For hard computational problems, stochastic local search has proven to be a competitive approach to finding optimal or approximately optimal problem solutions. Two key research questions for stochastic local search algorithms are: Which algorithms are effective for initialization? When should the search process be restarted? In the present work we investigate these research questions in the context of approximate computation of most probable explanations (MPEs) in Bayesian networks (BNs). We introduce a novel approach, based on the Viterbi algorithm, to explanation initialization in BNs. While the Viterbi algorithm works on sequences and trees, our approach works on BNs with arbitrary topologies. We also give a novel formalization of stochastic local search, with focus on initialization and restart, using probability theory and mixture models. Experimentally, we apply our methods to the problem of MPE computation, using a stochastic local search algorithm known as Stochastic Greedy Search. By carefully optimizing both initialization and restart, we reduce the MPE search time for application BNs by several orders of magnitude compared to using uniform at random initialization without restart. On several BNs from applications, the performance of Stochastic Greedy Search is competitive with clique tree clustering, a state-of-the-art exact algorithm used for MPE computation in BNs.

Mengshoel, Ole J.↗