Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed parallelization”

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

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

At least 1,351 records · Page 75

Predictive Inverse Model for Advective Heat Transfer in a Short–Circuited Fracture: Dimensional Analysis, Machine Learning, and Field Demonstration

Identifying fluid flow maldistribution in planar geometries is a well–established problem in subsurface science/engineering. Of particular importance to the thermal performance of enhanced (or “engineered”) geothermal systems is identifying the existence of nonuniform (i.e., heterogeneous) permeability and subsequently predicting advective heat transfer. Here, machine learning via a genetic algorithm (GA) identifies the spatial distribution of an unknown permeability field in a two–dimensional Hele–Shaw geometry (i.e., parallel plates). The inverse problem is solved by minimizing the L2 norm between simulated residence time distribution (RTD) and measurements of an inert tracer breakthrough curve (BTC) (C–Dot nanoparticle). Principal component analysis (PCA) of spatially correlated permeability fields enabled reduction of the parameter space by more than a factor of 10 and restricted the inverse search to reservoir–scale permeability variations. Thermal experiments and tracer tests conducted at the mesoscale Altona Field Laboratory (AFL) demonstrate that the method accurately predicts the effects of extreme flow channeling on heat transfer in a single bedding–plane rock fracture. However, this is only true when the permeability distributions provide adequate matches to both tracer RTD and frictional pressure loss. Without good agreement to frictional pressure loss, it is still possible to match a simulated RTD to measurements, but subsequent predictions of heat transfer are grossly inaccurate. Here, the results of this study suggest that it is possible to anticipate the thermal effects of flow maldistribution, but only if both simulated RTDs and frictional pressure loss between fluid inlets and outlets are in good agreement with measurements.

42 ENGINEERING↗

Improving CASINO performance for models with large number of electrons

Quantum Monte Carlo calculations have at their core algorithms based on statistical ensembles of multidimensional random walkers which are straightforward to use on parallel computers. Nevertheless some computations have reached the limit of the memory resources for models with more than 1000 electrons because of the need to store a large amount of electronic orbitals related data. Besides that, for systems with large number of electrons, it is interesting to study if the evolution of one configuration of random walkers can be done faster in parallel. We present a comparative study of two ways to solve these problems: (1) distributed orbital data done with MPI or Unix inter-process communication tools, (2) second level parallelism for configuration computation.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Performance analysis of distributed symmetric sparse matrix vector multiplication algorithm for multi-core architectures

In this article, sparse matrix vector multiply (SpMVM) is an important kernel that frequently arises in high performance computing applications. Due to its low arithmetic intensity, several approaches have been proposed in literature to improve its scalability and efficiency in large scale computations. In this paper, our target systems are high end multi-core architectures and we use messaging passing interface + open multiprocessing hybrid programming model for parallelism. We analyze the performance of recently proposed implementation of the distributed symmetric SpMVM, originally developed for large sparse symmetric matrices arising in ab initio nuclear structure calculations. We also study important features of this implementation and compare with previously reported implementations that do not exploit underlying symmetry. Our SpMVM implementations leverage the hybrid paradigm to efficiently overlap expensive communications with computations. Our main comparison criterion is the "CPU core hours" metric, which is the main measure of resource usage on supercomputers. We analyze the effects of topology-aware mapping heuristic using simplified network load model. Furthermore, we have tested the different SpMVM implementations on two large clusters with 3D Torus and Dragonfly topology. Our results show that the distributed SpMVM implementation that exploits matrix symmetry and hides communication yields the best value for the "CPU core hours" metric and significantly reduces data movement overheads.

97 MATHEMATICS AND COMPUTING↗

Wetness index based on landscape position and topography ($\mathrm{WILT}$): Modifying TWI to reflect landscape position

Water and land resource management planning benefits greatly from accurate prediction and understanding of the spatial distribution of wetness. The topographic wetness index (TWI) was conceived to predict relative surface wetness, and thus hydrologic responsiveness, across a watershed based on the assumption that shallow slope-parallel flow is a major driver of the movement and distribution of soil water. The index has been extensively used in modeling of landscape characteristics responsive to wetness, and some studies have shown the TWI performs well in landscapes where interflow is a dominant process. However, groundwater flow dominates the hydrology of low-slope landscapes with high subsurface conductivities, and the TWI assumptions are not likely to perform well in such environments. For groundwater dominated systems, we propose a hybrid wetness index (Wetness Index based on Landscape position and Topography, WILT) that inversely weights the upslope contributing area by the distance to the nearest surface water feature and the depth to groundwater. When explicit depth to groundwater data are not available, height above and separation from surface water features can act as surrogates for proximity to groundwater. Here, the resulting WILT map provides a more realistic spatial distribution of relative wetness across a low-slope Coastal Plain landscape as demonstrated by improved prediction of hydric soils, depth to groundwater, nitrogen and carbon concentrations in the A horizon of the soil profile, and sensitivity to DEM scale.

60 APPLIED LIFE SCIENCES↗

Impact of computer advances on future finite elements computations

Research performed over the past 10 years in engineering data base management and parallel computing is discussed, and certain opportunities for research toward the next generation of structural analysis capability are proposed. Particular attention is given to data base management associated with the IPAD project and parallel processing associated with the Finite Element Machine project, both sponsored by NASA, and a near term strategy for a distributed structural analysis capability based on relational data base management software and parallel computers for a future structural analysis system.

Fulton, Robert E.↗

Design considerations for parallel graphics libraries

Applications which run on parallel supercomputers are often characterized by massive datasets. Converting these vast collections of numbers to visual form has proven to be a powerful aid to comprehension. For a variety of reasons, it may be desirable to provide this visual feedback at runtime. One way to accomplish this is to exploit the available parallelism to perform graphics operations in place. In order to do this, we need appropriate parallel rendering algorithms and library interfaces. This paper provides a tutorial introduction to some of the issues which arise in designing parallel graphics libraries and their underlying rendering algorithms. The focus is on polygon rendering for distributed memory message-passing systems. We illustrate our discussion with examples from PGL, a parallel graphics library which has been developed on the Intel family of parallel systems.

Crockett, Thomas W.↗

Magnetospheric Multiscale Observations of the Electron Diffusion Region of Large Guide Field Magnetic Reconnection

We report observations from the Magnetospheric Multiscale (MMS) satellites of a large guide field magnetic reconnection event. The observations suggest that two of the four MMS spacecraft sampled the electron diffusion region, whereas the other two spacecraft detected the exhaust jet from the event. The guide magnetic field amplitude is approximately 4 times that of the reconnecting field. The event is accompanied by a significant parallel electric field (E(sub parallel lines) that is larger than predicted by simulations. The high-speed (approximately 300 km/s) crossing of the electron diffusion region limited the data set to one complete electron distribution inside of the electron diffusion region, which shows significant parallel heating. The data suggest that E(sub parallel lines) is balanced by a combination of electron inertia and a parallel gradient of the gyrotropic electron pressure.

Eriksson, S.↗

Field-Programmable Gate Array Computer in Structural Analysis: An Initial Exploration

This paper reports on an initial assessment of using a Field-Programmable Gate Array (FPGA) computational device as a new tool for solving structural mechanics problems. A FPGA is an assemblage of binary gates arranged in logical blocks that are interconnected via software in a manner dependent on the algorithm being implemented and can be reprogrammed thousands of times per second. In effect, this creates a computer specialized for the problem that automatically exploits all the potential for parallel computing intrinsic in an algorithm. This inherent parallelism is the most important feature of the FPGA computational environment. It is therefore important that if a problem offers a choice of different solution algorithms, an algorithm of a higher degree of inherent parallelism should be selected. It is found that in structural analysis, an 'analog computer' style of programming, which solves problems by direct simulation of the terms in the governing differential equations, yields a more favorable solution algorithm than current solution methods. This style of programming is facilitated by a 'drag-and-drop' graphic programming language that is supplied with the particular type of FPGA computer reported in this paper. Simple examples in structural dynamics and statics illustrate the solution approach used. The FPGA system also allows linear scalability in computing capability. As the problem grows, the number of FPGA chips can be increased with no loss of computing efficiency due to data flow or algorithmic latency that occurs when a single problem is distributed among many conventional processors that operate in parallel. This initial assessment finds the FPGA hardware and software to be in their infancy in regard to the user conveniences; however, they have enormous potential for shrinking the elapsed time of structural analysis solutions if programmed with algorithms that exhibit inherent parallelism and linear scalability. This potential warrants further development of FPGA-tailored algorithms for structural analysis.

Singleterry, Robert C., Jr.↗

Segmental Refinement: A Multigrid Technique for Data Locality

In this paper, we investigate a domain decomposed multigrid technique, termed segmental refinement, for solving general nonlinear elliptic boundary value problems. We extend the method first proposed in 1994 by analytically and experimentally investigating its complexity. We confirm that communication of traditional parallel multigrid is eliminated on fine grids, with modest amounts of extra work and storage, while maintaining the asymptotic exactness of full multigrid. We observe an accuracy dependence on the segmental refinement subdomain size, which was not considered in the original analysis. Finally, we present a communication complexity analysis that quantifies the communication costs ameliorated by segmental refinement and report performance results with up to 64K cores on a Cray XC30.

97 MATHEMATICS AND COMPUTING↗

Implementation of a parallel unstructured Euler solver on the CM-5

An efficient unstructured 3D Euler solver is parallelized on a Thinking Machine Corporation Connection Machine 5, distributed memory computer with vectoring capability. In this paper, the single instruction multiple data (SIMD) strategy is employed through the use of the CM Fortran language and the CMSSL scientific library. The performance of the CMSSL mesh partitioner is evaluated and the overall efficiency of the parallel flow solver is discussed.

Morano, Eric↗

Monte Carlo-Based Method for Parallelizing Quasi-Static Time-Series Power System Simulations: Preprint

Both commercial and open source power system simulators run quasi-static time-series (QSTS) simulations sequentially. For yearlong high-resolution load profiles this sequential execution of load flows incurs long computation times. Parallelizing QSTS simulations on multiple processors is one possible way to reduce the computation time. Parallelization however introduces errors in the simulation results as the initial state of the system at the beginning of each parallelized time period is not known. In this work a Monte Carlo-based approach has been proposed to estimate the initial state of the system for each parallel simulation run to mitigate the errors in the final results. Classical sequential QSTS simulation is the chosen base case against which all other methods have been compared. Results presented in this paper show that the proposed method improves the simulation results considerably. This paper also discusses possible sensitivities of the proposed method to a number of tunable parameters. Additionally, sensitivity analysis results for a number of parameters have also been presented in the results section.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Simulation and Optimizing Groundwater Investigations, Conceptual Design Analysis, Model Calibration and Remedial Design Using Parallel Numerical Methods: Theory and Application - 16578

Optimization of environmental restoration efforts at complex groundwater sites involves the development of a conceptual and often numerical flow and transport models. Typically, a limited number of simulations are done to assess the impact of parameter uncertainty, to evaluate remedial alternatives and develop a remedy, or to modify an existing remedy. A limited number of simulations are performed because each simulation can take upwards of several hours per simulation and until now the ability to conduct the 100's to 1000's of simulations needed to derive an optimal remedy was beyond the reach of the everyday practitioner. HGL has developed and implemented new parallel simulation and optimization methods in their software and has successfully deployed applications on high performance Cloud-based computing platforms. This advancement removes the impediment of burdensome and unpredictable runtime requirements these types of projects have typically exhibited using heuristic trial-and-error or even automated serial computation approaches. A parallel numerical solver has been added to the subsurface flow and transport simulator MODFLOW-SURFACT{sup TM} and has demonstrated reductions in solution time by as much as a factor of 9.8. This is accomplished by distributing the numerical computations of every time step over multiple processors. The integrated parallel simulation and optimization method uses the breadth and depth search approach; iteratively producing sets of tens or 100's of candidate solutions that are evaluated concurrently instead sequentially (i.e. one model evaluation at a time). Solutions within 80-100% of the globally optimal solution are regularly found expending only 5-10% of the usual computational effort using this new parallel optimization method (reducing computation effort by 10 to 20 times). Distributing optimal search runs across multiple processors additionally reduces calendar time to solutions. Practitioners evaluating multiple conceptual site models, model calibration by auto-calibration methods and remedial design by heuristics can enjoy significant reductions in model solution time by using the parallel numerical matrix solver in the physics-based simulator. Combining investigation and design analysis with breadth and depth parallel optimization methods, projects at the fringe of practical optimization, along with projects that were beyond the reach of formal optimization methods are now wholly feasible and tractable optimization problems. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Comparison of experimental surface pressures with theoretical predictions on twin two-dimensional convergent-divergent nozzles

A three-dimensional subsonic aerodynamic panel code (VSAERO) was used to predict the effects of upper and lower external nozzle flap geometry on the external afterbody/nozzle pressure coefficient distributions and external nozzle drag of nonaxisymmetric convergent-divergent exhaust nozzles having parallel external sidewalls installed on a generic twin-engine high performance aircraft model. Nozzle static pressure coefficient distributions along the upper and lower surfaces near the model centerline and near the outer edges (corner) of the two surfaces were calculated, and nozzle drag was predicted using these surface pressure distributions. A comparison between the theoretical predictions and experimental wind tunnel data is made to evaluate the utility of the code in calculating the flow about these types of non-axisymmetric afterbody configurations. For free-stream Mach numbers of 0.60 and 0.90, the conditions where the flows were attached on the boattails yielded the best comparison between the theoretical predictions and the experimental data. For the Boattail terminal angles of greater than 15 deg., the experimental data for M = 0.60 and 0.90 indicated areas of separated flow, so the theoretical predictions failed to match the experimental data. Even though calculations of regions of separated flows are within the capabilities of the theoretical method, acceptable solutions were not obtained.

Carlson, J. R.↗

Resonant Interactions Between Protons and Oblique Alfven/Ion-Cyclotron Waves

Resonant interactions between ions and Alfven/ion-cyclotron (A/IC) waves may play an important role in the heating and acceleration of the fast solar wind. Although such interactions have been studied extensively for 'parallel' waves, whose wave vectors k are aligned with the background magnetic field B{sub 0}, much less is known about interactions between ions and oblique A/IC waves, for which the angle theta between k and B{sub 0} is nonzero. In this paper, we present new numerical results on resonant cyclotron interactions between protons and oblique A/IC waves in collisionless low-beta plasmas such as the solar corona. We find that if some mechanism generates oblique high-frequency A/IC waves, then these waves initially modify the proton distribution function in such a way that it becomes unstable to parallel waves. Parallel waves are then amplified to the point that they dominate the wave energy at the large parallel wave numbers at which the waves resonate with the particles. Pitch-angle scattering by these waves then causes the plasma to evolve towards a state in which the proton distribution is constant along a particular set of nested 'scattering surfaces' in velocity space, whose shapes have been calculated previously. As the distribution function approaches this state, the imaginary part of the frequency of parallel A/IC waves drops continuously towards zero, but oblique waves continue to undergo cyclotron damping while simultaneously causing protons to diffuse across these kinetic shells to higher energies. We conclude that oblique A/IC waves can be more effective at heating protons than parallel A/IC waves, because for oblique waves the plasma does not relax towards a state in which proton damping of oblique A/IC waves ceases.

79 ASTRONOMY AND ASTROPHYSICS↗

Wave dispersion in a counterstreaming, relativistic thermal, magnetized, electron-positron plasma

The dispersion equation is analyzed for waves in a strongly magnetized, electron-positron plasma in which counterstreaming electrons and positrons have a relativistic thermal distribution in their respective rest frames, for propagation parallel to the magnetic field. We derive the response tensor for the medium, demonstrate the dispersion curves for different temperatures, and discuss the differences from the cold-plasma case. Application to the case of pulsar magnetospheres is discussed.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Nonlinear whistler instability driven by a beamlike distribution of resonant electrons

Theory and simulation are used to study the instability of a coherent whistler parallel-propagating in a simplified model radiation belt with a background of cold electrons, as well as a ring distribution of energetic electrons. A nonlinear instability is initiated at the location z{sub +}, where the electrons are cyclotron resonant with the wave, on the side of the equator (z=0) where the wave is propagating away from the equator. The instability propagates backward toward the equator, growing both spatially and temporally. As the instability develops, frequency falls in such a way as to keep the electrons nearly resonant with the waves over the entire region 0<z<z{sub +}. The instability causes a sharp drop in the pitch angle of the resonant electrons and eventually saturates with peak amplitude near the equator.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Distributed Macroscopic Traffic Simulation with Open Traffic Models

This paper presents OTM-MPI, an extension of the Open Traffic Models platform (OTM) for running macroscopic traffic simulations in high-performance computing environments. OTM-MPI represents the first open-source, distributed-memory, macroscopic simulation model developed for modern high performance parallel machines and large networks. Macroscopic simulations are appropriate for studying regional traffic scenarios when aggregate trends are of interest, rather than individual vehicle traces. They are also appropriate for studying the routing behavior of classes of vehicles, such as app-informed vehicles. The network partitioning was performed with METIS. Inter-process communication was done with MPI (message-passing interface). Results are provided for two networks: one realistic network which was obtained from Open Street Maps for Chattanooga, TN, and another larger synthetic grid network. The software recorded a speedups of 198x using 256 cores for Chattanooga, and 475x with 1,024 cores for the synthetic network.

macro-scopic traffic simulation↗

Catalytic Consequences of Acid Strength in the Conversion of Methanol to Dimethyl Ether

The effects of acid identity on CH{sub 3}OH dehydration are examined here using density functional theory (DFT) estimates of acid strength (as deprotonation energies, DPE) and reaction energies, combined with rate data on Keggin polyoxometalate (POM) clusters and zeolite H-BEA. Measured first-order (k{sub mono}) and zero-order (k{sub dimer}) CH3OH dehydration rate constants depend exponentially on DPE for POM clusters; the value of k{sub mono} depends more strongly on DPE than k{sub dimer} does. The chemical significance of these rate parameters and the basis for their dependences on acid strength were established by using DFT to estimate the energies of intermediates and transition states involved in elementary steps that are consistent with measured rate equations. We conclude from this treatment that CH{sub 3}OH dehydration proceeds via direct reactions of co-adsorbed CH{sub 3}OH molecules for relevant solid acids and reaction conditions. Methyl cations formed at ion-pair transition states in these direct routes are solvated by H{sub 2}O and CH{sub 3}OH more effectively than those in alternate sequential routes involving methoxide formation and subsequent reaction with CH{sub 3}OH. The stability of ion-pairs, prevalent as intermediates and transition states on solid acids, depends sensitively on DPE because of concomitant correlations between the stability of the conjugate anionic cluster and DPE. The chemical interpretation of k{sub mono} and k{sub dimer} from mechanism-based rate equations, together with thermochemical cycles of their respective transition state formations, show that similar charge distributions in the intermediate and transition state involved in k{sub dimer} cause its weaker dependence on DPE. Values of k{sub mono} involve uncharged reactants and the same ion-pair transition state as k{sub dimer}; these species sense acid strength differently and cause the larger effects of DPE on k{sub mono}. Confinement effects in H-BEA affect the value of k{sub mono} because the different sizes and number of molecules in reactants and transition states selectively stabilize the latter; however, they do not influence k{sub dimer}, for which reactants and transition states of similar size sense spatial constraints to the same extent. This combination of theory and experiment for solid acids of known structure sheds considerable light on the relative contributions from solvation, electrostatic and van der Waals interactions in stabilizing cationic transition states and provides predictive insights into the relative contributions of parallel routes based on the size and charge distributions of their relevant intermediates and transition states. These findings also demonstrate how the consequences of acid strength on measured turnover rates depend on reaction conditions and their concomitant changes in the chemical significance of the rate parameters measured. Moreover, the complementary use of experiment and theory in resolving mechanistic controversies has given predictive guidance about how rate and equilibrium constants, often inextricably combined as measured rate parameters, individually depend on acid strength based on the magnitude and spatial distributions of charges in reactants, products and transition states involved in relevant elementary steps. The unique relations between k{sub mono}, k{sub dimer} and DPE developed here for CH{sub 3}OH dehydration can be applied in practice to assess the acid strength of any solid acid, many of which have unknown structures, preventing reliable calculations of their DPE by theory.

10 SYNTHETIC FUELS↗