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 1,081 records · Page 60

EVIDENCE OF LANDAU AND CYCLOTRON RESONANCE BETWEEN PROTONS AND KINETIC WAVES IN SOLAR WIND TURBULENCE

The wave–particle interaction processes occurring in the solar wind provide crucial information to understand the wave dissipation and simultaneous particle heating in plasma turbulence. One requires observations of both wave fluctuations and particle kinetics near the dissipation range, which have, however, not yet been analyzed simultaneously. Here we show new evidence of wave–particle interactions by combining the diagnosis of wave modes with the analysis of particle kinetics on the basis of measurements from the WIND spacecraft with a high cadence of about 3 s. Solar wind protons appear to be highly dynamic in their velocity distribution consisting of varying anisotropic core and beam components. The basic scenario of solar wind proton heating through wave–particle interaction is suggested to be the following. Left-handed cyclotron resonance occurs continuously, and is evident from the observed proton core velocity distribution and the concurrent quasi-parallel left-handed Alfvén cyclotron waves. Landau and right-handed cyclotron resonances are persistent and indicated by the observed drifting anisotropic beam and the simultaneous quasi-perpendicular right-handed kinetic Alfvén waves in a general sense. The persistence of non-gyrotropic proton distributions may cast new light on the nature of the interaction between particles and waves near and beyond the proton gyro-frequency.

79 ASTRONOMY AND ASTROPHYSICS↗

Multi-water-bag models of ion temperature gradient instability in cylindrical geometry

Ion temperature gradient instabilities play a major role in the understanding of anomalous transport in core fusion plasmas. In the considered cylindrical geometry, ion dynamics is described using a drift-kinetic multi-water-bag model for the parallel velocity dependency of the ion distribution function. In a first stage, global linear stability analysis is performed. From the obtained normal modes, parametric dependencies of the main spectral characteristics of the instability are then examined. Comparison of the multi-water-bag results with a reference continuous Maxwellian case allows us to evaluate the effects of discrete parallel velocity sampling induced by the Multi-Water-Bag model. Differences between the global model and local models considered in previous works are discussed. Using results from linear, quasilinear, and nonlinear numerical simulations, an analysis of the first stage saturation dynamics of the instability is proposed, where the divergence between the three models is examined.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Automatic Parallelization of Numerical Python Applications using the Global Arrays Toolkit

Global Arrays is a software system from Pacific Northwest National Laboratory that enables an efficient, portable, and parallel shared-memory programming interface to manipulate distributed dense arrays. The NumPy module is the de facto standard for numerical calculation in the Python programming language, a language whose use is growing rapidly in the scientific and engineering communities. NumPy provides a powerful N-dimensional array class as well as other scientific computing capabilities. However, like the majority of the core Python modules, NumPy is inherently serial. Using a combination of Global Arrays and NumPy, we have reimplemented NumPy as a distributed drop-in replacement called Global Arrays in NumPy (GAiN). Serial NumPy applications can become parallel, scalable GAiN applications with only minor source code changes. Scalability studies of several different GAiN applications will be presented showing the utility of developing serial NumPy codes which can later run on more capable clusters or supercomputers.

99 GENERAL AND MISCELLANEOUS↗

Ropes: Support for collective opertions among distributed threads

Lightweight threads are becoming increasingly useful in supporting parallelism and asynchronous control structures in applications and language implementations. Recently, systems have been designed and implemented to support interprocessor communication between lightweight threads so that threads can be exploited in a distributed memory system. Their use, in this setting, has been largely restricted to supporting latency hiding techniques and functional parallelism within a single application. However, to execute data parallel codes independent of other threads in the system, collective operations and relative indexing among threads are required. This paper describes the design of ropes: a scoping mechanism for collective operations and relative indexing among threads. We present the design of ropes in the context of the Chant system, and provide performance results evaluating our initial design decisions.

Haines, Matthew↗

Spectral quadrature for the first principles study of crystal defects: Application to magnesium

In this work, we present an accurate and efficient finite-difference formulation and parallel implementation of Kohn-Sham Density (Operator) Functional Theory (DFT) for non periodic systems embedded in a bulk environment. Specifically, employing non-local pseudopotentials, local reformulation of electrostatics, and truncation of the spatial Kohn-Sham Hamiltonian, and the Linear Scaling Spectral Quadrature method to solve for the pointwise electronic fields in real-space and the non-local component of the atomic force, we develop a parallel finite difference framework suitable for distributed memory computing architectures to simulate non-periodic systems embedded in a bulk environment. Choosing examples from magnesium-aluminum alloys, we first demonstrate the convergence of energies and forces with respect to spectral quadrature polynomial order, and the width of the spatially truncated Hamiltonian. Next, we demonstrate the parallel scaling of our framework, and show that the computation time and memory scale linearly with respect to the number of atoms. Next, we use the developed framework to simulate isolated point defects and their interactions in magnesium-aluminum alloys. Our findings conclude that the binding energies of divacancies, Al solute-vacancy and two Al solute atoms are anisotropic and are dependent on cell size. Furthermore, the binding is favorable in all three cases.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Applicability of the single equivalent point dipole model to represent a spatially distributed bio-electrical source

Although the single equivalent point dipole model has been used to represent well-localised bio-electrical sources, in realistic situations the source is distributed. Consequently, position estimates of point dipoles determined by inverse algorithms suffer from systematic error due to the non-exact applicability of the inverse model. In realistic situations, this systematic error cannot be avoided, a limitation that is independent of the complexity of the torso model used. This study quantitatively investigates the intrinsic limitations in the assignment of a location to the equivalent dipole due to distributed electrical source. To simulate arrhythmic activity in the heart, a model of a wave of depolarisation spreading from a focal source over the surface of a spherical shell is used. The activity is represented by a sequence of concentric belt sources (obtained by slicing the shell with a sequence of parallel plane pairs), with constant dipole moment per unit length (circumferentially) directed parallel to the propagation direction. The distributed source is represented by N dipoles at equal arc lengths along the belt. The sum of the dipole potentials is calculated at predefined electrode locations. The inverse problem involves finding a single equivalent point dipole that best reproduces the electrode potentials due to the distributed source. The inverse problem is implemented by minimising the chi2 per degree of freedom. It is found that the trajectory traced by the equivalent dipole is sensitive to the location of the spherical shell relative to the fixed electrodes. It is shown that this trajectory does not coincide with the sequence of geometrical centres of the consecutive belt sources. For distributed sources within a bounded spherical medium, displaced from the sphere's centre by 40% of the sphere's radius, it is found that the error in the equivalent dipole location varies from 3 to 20% for sources with size between 5 and 50% of the sphere's radius. Finally, a method is devised to obtain the size of the distributed source during the cardiac cycle.

NASA Discipline Cardiopulmonary↗

National Combustion Code Parallel Performance Enhancements

The National Combustion Code (NCC) is being developed by an industry-government team for the design and analysis of combustion systems. The unstructured grid, reacting flow code uses a distributed memory, message passing model for its parallel implementation. The focus of the present effort has been to improve the performance of the NCC code to meet combustor designer requirements for model accuracy and analysis turnaround time. Improving the performance of this code contributes significantly to the overall reduction in time and cost of the combustor design cycle. This report describes recent parallel processing modifications to NCC that have improved the parallel scalability of the code, enabling a two hour turnaround for a 1.3 million element fully reacting combustion simulation on an SGI Origin 2000.

Quealy, Angela↗

Distributed Halide

Many image processing tasks are naturally expressed as a pipeline of small computational kernels known as stencils. Halide is a popular domain-specific language and compiler designed to implement image processing algorithms. Halide uses simple language constructs to express what to compute and a separate scheduling co-language for expressing when and where to perform the computation. This approach has demonstrated performance comparable to or better than hand-optimized code. Until now, however, Halide has been restricted to parallel shared memory execution, limiting its performance for memory-bandwidth-bound pipelines or large-scale image processing tasks. We present an extension to Halide to support distributed-memory parallel execution of complex stencil pipelines. These extensions compose with the existing scheduling constructs in Halide, allowing expression of complex computation and communication strategies. Existing Halide applications can be distributed with minimal changes, allowing programmers to explore the tradeoff between recomputation and communication with little effort. Approximately 10 new of lines code are needed even for a 200 line, 99 stage application. On nine image processing benchmarks, our extensions give up to a 1.4× speedup on a single node over regular multithreaded execution with the same number of cores, by mitigating the effects of non-uniform memory access. The distributed benchmarks achieve up to 18× speedup on a 16 node testing machine and up to 57× speedup on 64 nodes of the NERSC Cori supercomputer.

97 MATHEMATICS AND COMPUTING↗

Semi-automatic process partitioning for parallel computation

On current multiprocessor architectures one must carefully distribute data in memory in order to achieve high performance. Process partitioning is the operation of rewriting an algorithm as a collection of tasks, each operating primarily on its own portion of the data, to carry out the computation in parallel. A semi-automatic approach to process partitioning is considered in which the compiler, guided by advice from the user, automatically transforms programs into such an interacting task system. This approach is illustrated with a picture processing example written in BLAZE, which is transformed into a task system maximizing locality of memory reference.

Koelbel, Charles↗

A 2.5 kW cascaded Schwarz converter for 20 kHz power distribution

Because it avoids the high currents in a parallel loaded capacitor, the cascaded Schwarz converter should offer better component utilization than converters with sinusoidal output voltages. The circuit is relatively easy to protect, and it provides a predictable trapezoidal voltage waveform that should be satisfactory for 20-kHz distribution systems. Analysis of the system is enhanced by plotting curves of normalized variables vs. gamma(1), where gamma(1) is proportional to the variable frequency of the first stage. Light-load operation is greatly improved by the addition of a power recycling rectifier bridge that is back biased at medium to heavy loads. Operation has been verified on a 2.5-kW circuit that uses input and output voltages in the same range as those anticipated for certain future spacecraft power systems.

Shetler, Russell E.↗

Parallel implicit unstructured grid Euler solvers

A mesh-vertex finite volume scheme for solving the Euler equations on triangular unstructured meshes is implemented on an MIMD (multiple instruction/multiple data stream) parallel computer. An explicit four-stage Runge-Kutta scheme is used to solve two-dimensional flow problems. A family of implicit schemes is also developed to solve these problems, where the linear system that arises at each time step is solved by a preconditioned GMRES algorithm. Two partitioning strategies are employed, one that partitions triangles and the other that partitions vertices. The choice of the preconditioner in a distributed memory setting is discussed. All the methods are compared both in terms of elapsed times and convergence rates. It is shown that the implicit schemes offer adequate parallelism at the expense of minimal sequential overhead. The use of a global coarse grid to further minimize this overhead is also investigated. The schemes are implemented on a distributed memory parallel computer, the iPSC/860.

Venkatakrishnan, V.↗

Aces4: A Platform for Computational Chemistry Calculations with Extremely Large Block-Sparse Arrays

Aces4 is a parallel programming platform comprising a DSL for Computational Chemistry and its runtime system. It offers a convenient way to express parallelism together with extensive support for extremely large, possibly sparse, distributed arrays. It aids scientists in the creation of performant, scalable, massively parallel programs that can effectively take advantage of leadership class computing systems to address important scientific questions. Aces4 has enabled the development and implementation of new methods in electronic structure theory which are breaking new ground in their ability to perform highly accurate calculations on ever larger molecular systems. In this paper the design of Aces4, which is based on the the Super Instruction Architecture approach, is described. Experimental scaling results for Molecular Cluster Perturbation Theory, a new method enabled by Aces4, and CCSD, a widely used computational chemistry method are given.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Forced motion of an initially stressed rectangular plate - An elasticity solution

An initially stressed rectangular plate is subjected to a transverse, time dependent load. The initial stress field is constant and parallel to two of the plate edges. The transverse load is uniformly distributed over a rectangular area, the sides of which are parallel to the sides of the plate. The problem is formulated within the framework of classical, three-dimensional elasticity theory, and an exact solution is obtained in series form. The case of static stability under pre-stress is also considered.

Reismann, H.↗

An Adaptive Memory Interface Controller for Improving Bandwidth Utilization of Hybrid and Reconfigurable Systems

Emerging applications such as data mining, bioinformatics, knowledge discovery, social network analysis are irregular. They use data structures based on pointers or linked lists, such as graphs, unbalanced trees or unstructures grids, which generates unpredictable memory accesses. These data structures usually are large, but difficult to partition. These applications mostly are memory bandwidth bounded and have high synchronization intensity. However, they also have large amounts of inherent dynamic parallelism, because they potentially perform a task for each one of the element they are exploring. Several efforts are looking at accelerating these applications on hybrid architectures, which integrate general purpose processors with reconfigurable devices. Some solutions, which demonstrated significant speedups, include custom-hand tuned accelerators or even full processor architectures on the reconfigurable logic. In this paper we present an approach for the automatic synthesis of accelerators from C, targeted at irregular applications. In contrast to typical High Level Synthesis paradigms, which construct a centralized Finite State Machine, our approach generates dynamically scheduled hardware components. While parallelism exploitation in typical HLS-generated accelerators is usually bound within a single execution flow, our solution allows concurrently running multiple execution flow, thus also exploiting the coarser grain task parallelism of irregular applications. Our approach supports multiple, multi-ported and distributed memories, and atomic memory operations. Its main objective is parallelizing as many memory operations as possible, independently from their execution time, to maximize the memory bandwidth utilization. This significantly differs from current HLS flows, which usually consider a single memory port and require precise scheduling of memory operations. A key innovation of our approach is the generation of a memory interface controller, which dynamically maps concurrent memory accesses to multiple ports. We present a case study on a typical irregular kernel, Graph Breadth First search (BFS), exploring different tradeoffs in terms of parallelism and number of memories.

BFS↗

Design and Analysis of 10” Parallel Plate Relief Device

All Cryomodule (CM) and Cryogenic Distribution System (CDS) relieving into Helium Low Pressure (LP) return header, which is connected to compressor suction so, helium can be preserved during small flow relieving event and recirculated to system. However, during worst case scenario, Helium LP header requires a parallel plate relief device to relieve excess pressure from header. To complete the CDS Warm piping header, a new design for a 10 parallel plate relief device is necessary to relieve outside of the tunnel into atmosphere.

Chicas, Kelly↗

Study of Electromagnetic Electron Cyclotron Waves for Kappa Distribution with AC Field in the Magnetosphere of Saturn

Parallel propagating electromagnetic electron cyclotron (EMEC) waves in the extended plasma sheet (~12R{sub S}) and in the outer magnetosphere (~18R{sub S}) of Saturn have been studied. A dispersion relation for parallel propagating relativistic EMEC waves has been applied to the magnetosphere of Saturn, and comparisons have been made with the data of Voyager 1 at these radial distances. The detailed investigations for EMEC waves have been done in the presence of the perpendicular AC electric field, using the kappa distribution function. The relativistic temporal growth rate is calculated by the method of characteristic solution with the data provided by Voyager 1. The effect of the suprathermal electron density, temperature anisotropy, frequency of AC electric field, thermal energy of ions, and relativistic factor on the temporal growth rate of EMEC wave emission has been studied. The simulation results show that the growth of parallel propagating EMEC waves is significantly affected by variations in the temperature anisotropy, electron density, ion thermal energy, and relativistic factor in both the extended plasma sheet and the outer magnetosphere of Saturn. The temperature anisotropy (T⊥/T{sub ║}), ion thermal energy (K{sub B}T{sub ║i}), and electron density (n{sub 0}) have been found to be a major source of free energy for parallel propagating EMEC waves in both regions.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Water bag modeling of a multispecies plasma

We report in the present paper a new modeling method to study multiple species dynamics in magnetized plasmas. Such a method is based on the gyrowater bag modeling, which consists in using a multistep-like distribution function along the velocity direction parallel to the magnetic field. The choice of a water bag representation allows an elegant link between kinetic and fluid descriptions of a plasma. The gyrowater bag model has been recently adapted to the context of strongly magnetized plasmas. We present its extension to the case of multi ion species magnetized plasmas: each ion species being modeled via a multiwater bag distribution function. The water bag modelization will be discussed in details, under the simplification of a cylindrical geometry that is convenient for linear plasma devices. As an illustration, results obtained in the linear framework for ion temperature gradient instabilities are presented, that are shown to agree qualitatively with older works.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Electron Scattering by High-frequency Whistler Waves at Earth’s Bow Shock

Electrons are accelerated to non-thermal energies at shocks in space and astrophysical environments. While different mechanisms of electron acceleration have been proposed, it remains unclear how non-thermal electrons are produced out of the thermal plasma pool. Here, we report in situ evidence of pitch-angle scattering of non-thermal electrons by whistler waves at Earth’s bow shock. On 2015 November 4, the Magnetospheric Multiscale (MMS) mission crossed the bow shock with an Alfvén Mach number ∼11 and a shock angle ∼84°. In the ramp and overshoot regions, MMS revealed bursty enhancements of non-thermal (0.5–2 keV) electron flux, correlated with high-frequency (0.2–0.4 Ω{sub ce}, where Ω{sub ce} is the cyclotron frequency) parallel-propagating whistler waves. The electron velocity distribution (measured at 30 ms cadence) showed an enhanced gradient of phase-space density at and around the region where the electron velocity component parallel to the magnetic field matched the resonant energy inferred from the wave frequency range. The flux of 0.5 keV electrons (measured at 1 ms cadence) showed fluctuations with the same frequency. These features indicate that non-thermal electrons were pitch-angle scattered by cyclotron resonance with the high-frequency whistler waves. However, the precise role of the pitch-angle scattering by the higher-frequency whistler waves and possible nonlinear effects in the electron acceleration process remains unclear.

79 ASTRONOMY AND ASTROPHYSICS↗