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 1,711 records · Page 95

Jet array impingement with crossflow-correlation of streamwise resolved flow and heat transfer distributions

Correlations for heat transfer coefficients for jets of circular offices and impinging on a surface parallel to the jet orifice plate are presented. The air, following impingement, is constrained to exit in a single direction along the channel formed by the jet orifice plate and the heat transfer (impingement) surface. The downstream jets are subjected to a crossflow originating from the upstream jets. Impingement surface heat transfer coefficients resolved to one streamwise jet orifice spacing, averaged across the channel span, are correlated with the associated individual spanwise orifice row jet and crossflow velocities, and with the geometric parameters.

Florschuetz, L. W.↗

Non-stoichiometric material transfer in the pulsed laser deposition of LaAlO3

Inequivalent angular distributions have been found for La and Al in the ablation plume from LaAlO3 single crystal targets using a KrF laser during pulsed laser deposition. Angular distributions and stoichiometries in the condensate were measured and reveal decidedly non-stoichiometric transfer from target to substrate over most of the angular range. Composition varied dramatically for plume angles parallel to the long axis of the laser spot with the on-axis position exhibiting a peak in the La/Al atom ratio at ~1.5. The distributions were more diffuse in the perpendicular direction. Stoichiometric LaAlO3 was found in the condensate only at an extreme off-axis position.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Multi-fault Tolerance for Cartesian Data Distributions

Faults are expected to play an increasingly important role in how algorithms and applications are designed to run on future extreme-scale sys- tems. Algorithm-based fault tolerance (ABFT) is a promising approach that involves modications to the algorithm to recover from faults with lower over- heads than replicated storage and a signicant reduction in lost work compared to checkpoint-restart techniques. Fault-tolerant linear algebra (FTLA) algo- rithms employ additional processors that store parities along the dimensions of a matrix to tolerate multiple, simultaneous faults. Existing approaches as- sume regular data distributions (blocked or block-cyclic) with the failures of each data block being independent. To match the characteristics of failures on parallel computers, we extend these approaches to mapping parity blocks in several important ways. First, we handle parity computation for generalized Cartesian data distributions with each processor holding arbitrary subsets of blocks in a Cartesian-distributed array. Second, techniques to handle corre- lated failures, i.e., multiple processors that can be expected to fail together, are presented. Third, we handle the colocation of parity blocks with the data blocks and do not require them to be on additional processors. Several al- ternative approaches, based on graph matching, are presented that attempt to balance the memory overhead on processors while guaranteeing the same fault tolerance properties as existing approaches that assume independent fail- ures on regular blocked data distributions. The evaluation of these algorithms demonstrates that the additional desirable properties are provided by the pro- posed approach with minimal overhead.

Ali, Nawab↗

An MCNPX Monte Carlo model of a discrete spot scanning proton beam therapy nozzle

Purpose: The purposes of this study were to validate a discrete spot scanning proton beam nozzle using the Monte Carlo (MC) code MCNPX and use the MC validated model to investigate the effects of a low-dose envelope, which surrounds the beam's central axis, on measurements of integral depth dose (IDD) profiles. Methods: An accurate model of the discrete spot scanning beam nozzle from The University of Texas M. D. Anderson Cancer Center (Houston, Texas) was developed on the basis of blueprints provided by the manufacturer of the nozzle. The authors performed simulations of single proton pencil beams of various energies using the standard multiple Coulomb scattering (MCS) algorithm within the MCNPX source code and a new MCS algorithm, which was implemented in the MCNPX source code. The MC models were validated by comparing calculated in-air and in-water lateral profiles and percentage depth dose profiles for single pencil beams with their corresponding measured values. The models were then further tested by comparing the calculated and measured three-dimensional (3-D) dose distributions. Finally, an IDD profile was calculated with different scoring radii to determine the limitations on the use of commercially available plane-parallel ionization chambers to measure IDD. Results: The distance to agreement, defined as the distance between the nearest positions of two equivalent distributions with the same value of dose, between measured and simulated ranges was within 0.13 cm for both MCS algorithms. For low and intermediate pencil beam energies, the MC simulations using the standard MCS algorithm were in better agreement with measurements. Conversely, the new MCS algorithm produced better results for high-energy single pencil beams. The IDD profile calculated with cylindrical tallies with an area equivalent to the area of the largest commercially available ionization chamber showed up to 7.8% underestimation of the integral dose in certain depths of the IDD profile. Conclusions: The authors conclude that a combination of MCS algorithms is required to accurately reproduce experimental data of single pencil beams and 3-D dose distributions for the scanning beam nozzle. In addition, the MC simulations showed that because of the low-dose envelope, ionization chambers with radii as large as 4.08 cm are insufficient to accurately measure IDD profiles for a 221.8 MeV pencil beam in the scanning beam nozzle.

60 APPLIED LIFE SCIENCES↗

Scaling of observable properties in rotating stars

The spectral energy distribution as a function of inclination is computed using two-dimensional rotating stellar models and NLTE plane parallel stellar atmospheres. These models cover the range from 1.875 M {sub ☉} to 3.0 M {sub ☉}. The deduced effective temperature is determined by B – V computed from the spectral energy distribution, and the deduced luminosity is computed as the integral of the spectral energy distribution over all frequencies, assuming the distance and reddening are known. These deduced quantities are obtained from the observed spectral energy distribution assuming the objects are spherically symmetric, and thus the results are dependent on the inclination. Previous work has shown that the surface properties between two rotating stellar models with the same surface shape scale, and this is also true for the deduced effective temperature and luminosity over this limited mass range.

79 ASTRONOMY AND ASTROPHYSICS↗

Lowering and Runtime Support for Fortran’s Multi-Image Parallel Features using LLVM Flang, PRIF, and Caffeine

This paper provides an overview of the multi-image parallel features in Fortran 2023 and their implementation in the LLVM flang compiler and the Caffeine parallel runtime library. The features of interest support a Single-Program, Multiple-Data (SPMD) programming model based on executing multiple “images”, each of which is a program instance. The features also support a Partitioned Global Address Space (PGAS) in the form of “coarray” distributed data structures. The paper discusses the lowering of multi-image features to the Parallel Runtime Interface for Fortran (PRIF) and the implementation of PRIF in the Caffeine parallel runtime library. This paper also provides an early view into the design of a new multi-image dialect of the LLVM Multi-Level Intermediate Representation (MLIR). We describe validation and testing of the resulting software stack, and demonstrate that performance compares favorably to another open-source compiler and runtime library: GNU Compiler Collection (GCC) gfortran and OpenCoarrays, respectively.

Bonachea, Dan↗

SU-E-J-208: Feasibility Study On Using Small Plastic Phantoms for Auditing Radiation Output of MR-Linac Systems

Purpose: To study feasibility of using small plastic phantoms designed for conventional linac output auditing to measure the output of MR-Linac systems. Methods: For simulations, the CT scan of an IROC(formerly RPC) acrylic block phantom designed for 8 MV beams was imported in a research version of the treatment planning system (Monaco). Dose delivered to three TLDs in the block was calculated with a Monte Carlo algorithm and a beam model based on an MR-linac prototype with and without a magnetic field (B=1.5T). In a large mathematical water phantom, the same beam was used to calculate dose in full scatter conditions. The block factor (F) was calculated as the ratio of the average dose to the block TLDs to the dose at the reference point in the mathematical phantom. For experimental measurement, four IROC blocks were irradiated with the MR-linac prototype, and data were analyzed by IROC. Results: The F factor without a B field was 1.053. When a B field was applied, it changed the dose distribution in the block, especially on the edges. With a B field parallel to the long axes of the TLD, F was 1.038. However, with a perpendicular B field, F factor increased slightly to 1.075. In the IROC report, the output determined with two blocks parallel to the B field was 2.3% higher than the output by the two blocks perpendicular to the B field. The average of all four blocks was within 2% of machine output measured with an ion chamber. Conclusion: It may be feasible to expand the utility of the acrylic block phantoms for radiation output auditing from conventional linacs to MR-linacs. However, the scatter correction factor can change due to the B field and its orientation to the block. More symmetric phantom designs may be less prone to mistakes. We acknowledge research support from Elekta.

60 APPLIED LIFE SCIENCES↗

MOJAVE: Monitoring of jets in active galactic nuclei with VLBA experiments. XI. Spectral distributions

We have obtained milliarcsecond-scale spectral index distributions for a sample of 190 extragalactic radio jets through the Monitoring of Jets in Active Galactic Nuclei with the VLBA Experiments (MOJAVE) project. The sources were observed in 2006 at 8.1, 8.4, 12.1, and 15.4 GHz, and we have determined spectral index maps between 8.1 and 15.4 GHz to study the four-frequency spectrum in individual jet features. We have performed detailed simulations to study the effects of image alignment and (u, v)-plane coverage on the spectral index maps to verify our results. We use the spectral index maps to study the spectral index evolution along the jet and determine the spectral distributions in different locations of the jets. The core spectral indices are on average flat with a mean value of +0.22 ± 0.03 for the sample, while the jet spectrum is in general steep with a mean index of –1.04 ± 0.03. A simple power-law fit is often inadequate for the core regions, as expected if the cores are partially self-absorbed. The overall jet spectrum steepens at a rate of about –0.001 to –0.004 per deprojected parsec when moving further out from the core with flat spectrum radio quasars having significantly steeper spectra (mean –1.09 ± 0.04) than the BL Lac objects (mean –0.80 ± 0.05). However, the spectrum in both types of objects flattens on average by ∼0.2 at the locations of the jet components indicating particle acceleration or density enhancements along the jet. The mean spectral index at the component locations of –0.81 ± 0.02 corresponds to a power-law index of ∼2.6 for the electron energy distribution. We find a significant trend that jet components with linear polarization parallel to the jet (magnetic field perpendicular to the jet) have flatter spectra, as expected for transverse shocks. Compared to quasars, BL Lacs have more jet components with perpendicular magnetic field alignment, which may explain their generally flatter spectra. The overall steepening of the spectra with distance can be explained with radiative losses if the jets are collimating or with the evolution of the high-energy cutoff in the electron spectrum if the jets are conical. This interpretation is supported by a significant correlation with the age of the component and the spectral index, with older components having steeper spectra.

79 ASTRONOMY AND ASTROPHYSICS↗

Coupled ridge waveguide distributed feedback quantum cascade laser arrays

A coupled ridge waveguide quantum cascade laser (QCL) array consisting of fifteen elements with parallel integration was presented. In-phase fundamental mode operation in each element is secured by both the index-guided nature of the ridge and delicate loss management by properly designed geometries of the ridges and interspaces. Single-lobe lateral far-field with a nearly diffraction limited beam pattern was obtained. By incorporating a one-dimensional buried distributed feedback grating, the in-phase-operating coupled ridge waveguide QCL design provides an efficient solution to obtaining high output power and stable single longitudinal mode emission. The simplicity of this structure and fabrication process makes this approach attractive to many practical applications.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Programming approaches for scalability, performance, and portability of combustion physics codes

Here, this paper presents the process, strategy, and results associated with porting a typical combustion physics flow solver to current state-of-the-art and future massively-parallel computer architectures. Major focus is placed on the distinct algorithmic structure of these types of codes and how it can be integrated with modern programming paradigms for heterogeneous platforms (i.e., distributed many-core systems with accelerators). An end-to-end case study is presented that exemplifies the process in a generic manner, which then serves as a clear guide with respect to the strategy and best practices leading to a robust and adaptable framework that performs well, is durable over time, is portable, and requires minimal human-effort. This end is accomplished beginning with the use of a mature, validated, structured, multiblock code framework optimized for application of both Large Eddy Simulation (LES) and Direct Numerical Simulation (DNS). This code has been ported to a variety of platforms over the past decade, including most recently the Oak Ridge Leadership Computing Facility’s “Summit” Platform. The experience gained on these multiple platforms provides general insights and thus the results presented are not specific to any one code or platform other than the overarching trend toward distributed many-core systems with accelerators in order to move toward exascale performance. The resultant performance and scalability of the ported code is demonstrated on a real-world application; a state-of-the-art rotating detonation rocket engine simulation that matches the complex geometry and boundary conditions imposed as part of a companion experimental campaign.

97 MATHEMATICS AND COMPUTING↗

Distributed parameter modeling of the structural dynamics of the Solar Array Flight Experiment

A distributed-parameter model of the structural dynamics of the space-shuttle-deployed Solar Array Flight Experiment is developed and used to produce estimates of the modal frequencies and mode shapes. A lumped parameter version of the distributed model is used to estimate model characteristics by analyzing the measured responses of 32 targets. To make the modeling more tenable, a distributed parameter system is used to reduce the number of unknown parameters, a modified Newton-Raphson technique is used for rapid convergence, and a parallel processing supercomputer is used for more efficient computation. The performances of computers with a high-speed serial processor and with a high-speed parallel processor are compared. The best results are obtained with the modeling approach in which maximum likelihood estimation is applied to distributed parameter models.

Taylor, L. W., Jr.↗

Non-gyrotropic proton and alpha-particle velocity distributions in the solar wind: TAUS observations and stability analysis

Ion velocity distribution functions have been measured with high time resolution by the TAUS plasma instrument on the PHOBOS mission to Mars in 1989. The unambiguous separation of protons and alpha-particles by TAUS enabled us to study the nonthermal features of their distributions separately and to analyze the stability of the distributions against excitation of waves in the cyclotron-frequency domain. Typical nonthermal features include temperature anisotropies, with T(sub perpendicular) larger than T(sub parallel), and ion beam populations drifting along the local magnetic field direction. Also, distinctly non-gyrotropic alpha-particle velocity distributions were sometimes found. Non-gyrotropy strongly changes the wave dispersion and gives rise to new growing modes, related to the coupling of the standard wave modes existing in gyrotropic plasma. It is found that for the measured non-gyrotropic ion distributions the right-hand polarized wave can also be excited by a temperature anistropy instead of the usual beam drift.

Astudillo, H. F.↗

Bell-Curve Based Evolutionary Optimization Algorithm

The paper presents an optimization algorithm that falls in the category of genetic, or evolutionary algorithms. While the bit exchange is the basis of most of the Genetic Algorithms (GA) in research and applications in America, some alternatives, also in the category of evolutionary algorithms, but use a direct, geometrical approach have gained popularity in Europe and Asia. The Bell-Curve Based Evolutionary Algorithm (BCB) is in this alternative category and is distinguished by the use of a combination of n-dimensional geometry and the normal distribution, the bell-curve, in the generation of the offspring. The tool for creating a child is a geometrical construct comprising a line connecting two parents and a weighted point on that line. The point that defines the child deviates from the weighted point in two directions: parallel and orthogonal to the connecting line, the deviation in each direction obeying a probabilistic distribution. Tests showed satisfactory performance of BCB. The principal advantage of BCB is its controllability via the normal distribution parameters and the geometrical construct variables.

Sobieszczanski-Sobieski, J.↗

torch-einshard v1.0

torch-einshard is a Python library for describing local and distributed PyTorch tensor computations with compact, einsum-like notation. Its expressions name logical axes, specify how they are sharded across a PyTorch DeviceMesh, and represent partial reductions. The library automatically performs contractions, permutations, reshaping, splitting, gathering, reduction, reduce-scatter, and repartitioning while preserving autograd. Additional features include sharding-aware FFTs, tensor rolls, halo exchange, sliding windows, 1D–3D convolutions, uneven-shard handling, parameter initialization and gradient management, and cost-based execution planning. It is designed for scientific machine learning and large-model workloads, including tensor-, sequence-, and spatial-parallel MLPs, attention, convolutions, and spectral operations. Compared with manually combining torch.einsum and distributed collectives, torch-einshard expresses both the mathematical operation and data placement in one readable formula. This reduces boilerplate and synchronization errors, keeps forward and backward communication consistent, and allows the library to select optimized collective strategies without changing model code.

Morozov, Dmitriy [Lawrence Berkeley National Labor↗

A simple method for obtaining a three-dimensional proton distribution function from Voyager plasma data

The main sensor of the Vogager plasma experiment consists of a cluster of three, modulated-grid Faraday cups whose normals are arranged symmetrically about the symmetry axis of the cluster at an angle of 20 degrees to that axis. In interplanetary space, each cup explores the positive ion distribution by accepting particles from contiguous slices in velocity space. The slices are narrow in the direction of the normal to the modulating grid but are broad in planes parallel to that grid. The resulting three sets of measurements can be combined to yield the three-dimensional distribution function in the following way: the distribution function is assumed to be gyrotropic. For each value of speed in a frame of reference moving with the bulk velocity of the solar wind, the variation of the distribution function with angle from the field direction is represented by a series of Legendre polynomials. Effects such as double-streaming and heat flow can be well represented by using only the first three terms of the series which are fully specified by the measurements. Examples of the use of this method in the analysis of Voyager data are shown.

Olbert, S.↗

Probability of relativistic electron trapping by parallel and oblique whistler-mode waves in Earth's radiation belts

We investigate electron trapping by high-amplitude whistler-mode waves propagating at small as well as large angles relative to geomagnetic field lines. The inhomogeneity of the background magnetic field can result in an effective acceleration of trapped particles. Here, we derive useful analytical expressions for the probability of electron trapping by both parallel and oblique waves, paving the way for a full analytical description of trapping effects on the particle distribution. Numerical integrations of particle trajectories allow to demonstrate the accuracy of the derived analytical estimates. For realistic wave amplitudes, the levels of probabilities of trapping are generally comparable for oblique and parallel waves, but they turn out to be most efficient over complementary energy ranges. Trapping acceleration of <100 keV electrons is mainly provided by oblique waves, while parallel waves are responsible for the trapping acceleration of >100 keV electrons.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Particle Acceleration in Relativistic Alfvénic Turbulence

Abstract Strong magnetically dominated Alfvénic turbulence is an efficient engine of nonthermal particle acceleration in a relativistic collisionless plasma. We argue that in the limit of strong magnetization, the type of energy distribution attained by accelerated particles depends on the relative strengths of turbulent fluctuations δ B 0 and the guide field B 0 . If δ B 0 ≪ B 0 , the particle magnetic moments are conserved, and the acceleration is provided by magnetic curvature drifts. Curvature acceleration energizes particles in the direction parallel to the magnetic field lines, resulting in log-normal tails of particle energy distribution functions. Conversely, if δ B 0 ≳ B 0 , interactions of energetic particles with intense turbulent structures can scatter particles, creating a population with large pitch angles. In this case, magnetic mirror effects become important, and turbulent acceleration leads to power-law tails of the energy distribution functions.

79 ASTRONOMY AND ASTROPHYSICS↗

Reducing Concurrency Bottlenecks in Parallel I/O Workloads

To enable high performance parallel checkpointing we introduced the Parallel Log Structured File System (PLFS). PLFS is middleware interposed on the file system stack to transform concurrent writing of one application file into many non-concurrently written component files. The promising effectiveness of PLFS makes it important to examine its performance for workloads other than checkpoint capture, notably the different ways that state snapshots may be later read, to make the case for using PLFS in the Exascale I/O stack. Reading a PLFS file involved reading each of its component files. In this paper we identify performance limitations on broader workloads in an early version of PLFS, specifically the need to build and distribute an index for the overall file, and the pressure on the underlying parallel file system's metadata server, and show how PLFS's decomposed components architecture can be exploited to alleviate bottlenecks in the underlying parallel file system.

COMPUTER ARCHITECTURE↗