Search NASA⌕ Search

SEARCH · Search NASA

Results for “Vectorized algorithm”

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 973 records · Page 54

Final report for “Extreme-scale Algorithms and Solver Resilience”

This is a joint project with principal investigators at Oak Ridge National Laboratory, Sandia National Laboratories, the University of California at Berkeley, and the University of Tennessee. Our part of the project involves developing performance models for highly scalable algorithms and the development of latency tolerant iterative methods. During this project, we extended our performance models for the Multigrid method for solving large systems of linear equations and conducted experiments with highly scalable variants of conjugate gradient methods that avoid blocking synchronization. In addition, we worked with the other members of the project on alternative techniques for resilience and reproducibility. We also presented an alternative approach for reproducible dot-products in parallel computations that performs almost as well as the conventional approach by separating the order of computation from the details of the decomposition of vectors across the processes.

97 MATHEMATICS AND COMPUTING↗

A finite element formulation for deformation twinning induced strain localization in polycrystal magnesium alloys

Deformation twinning induces shear strain localization in hexagonal close-packed crystals and is critical for the material’s ductility and failure. Cracks often occur at twin-twin or twin-grain boundary intersections and propagate along twin bands. However, most crystal plasticity models for deformation twinning are based on a “pseudo-slip” approach and do not capture the localized deformation associated with the formation of each discrete twin band. The few exceptions are discrete twin models that involve very complex numerical algorithms and are often compromised in accuracy due to the numerical convergence. These factors make the discrete twin models hard to adopt. This paper proposes a modification to the conventional finite element weak form, to fully incorporate a twin-induced heterogeneous deformation that does not depend on the “pseudo-slip” assumption. The model starts by splitting the deformation gradient into elastic-slip-twinning components. The twin-induced deformation gradient component is computed separately by solving a microstructural evolution problem and then implemented into finite element weak form by constructing a global “twin-force” vector. The constitutive update (e.g., in the user-defined material subroutine, or UMAT, for ABAQUS) therefore avoids dealing with the twinning and recovers to the form of a regular slip-based crystal plasticity model. The results presented here indicate that the twin-induced strain localization and the associated stress-reversal phenomena near the twin band were naturally captured in the model, which was validated against an in-situ synchrotron X-ray micro-diffraction experiment.

36 MATERIALS SCIENCE↗

Sensitivity of back‐projection algorithm (BPA) synthetic aperture radar (SAR) image formation to initial position, velocity, and attitude navigation errors

Abstract The Back‐Projection Algorithm (BPA) is a time‐domain‐matched filtering technique to form synthetic aperture radar (SAR) images. To produce high‐quality BPA images, precise navigation data for the radar platform must be known. Errors in position, velocity, or attitude result in improperly formed images that are corrupted by shifting and blurring. The contribution of this paper is the development of analytical expressions that characterise the relationship between navigation errors and image formation errors from an inertial navigation point of view, where trajectory estimation errors in position, velocity, and attitude propagate through time and cause compounding errors in the vehicle state vector. These analytical expressions are verified via simulated image formation and real‐data image formation.

Lindstrom, Colton↗

Calculation of grain boundary normals directly from 3D microstructure images

The determination of grain boundary normals is an integral part of the characterization of grain boundaries in polycrystalline materials. These normal vectors are difficult to quantify due to the discretized nature of available microstructure characterization techniques. The most common method to determine grain boundary normals is by generating a surface mesh from an image of the microstructure, but this process can be slow, and is subject to smoothing issues. A new technique is proposed, utilizing first order Cartesian moments of binary indicator functions, to determine grain boundary normals directly from a voxelized microstructure image. In order to validate the accuracy of this technique, the surface normals obtained by the proposed method are compared to those generated by a surface meshing algorithm. Specifically, the local divergence between the surface normals obtained by different variants of the proposed technique and those generated from a surface mesh of a synthetic microstructure constructed using a marching cubes algorithm followed by Laplacian smoothing is quantified. Next, surface normals obtained with the proposed method from a measured 3D microstructure image of a Ni polycrystal are used to generate grain boundary character distributions (GBCD) for Σ3 and Σ9 boundaries, and compared to the GBCD generated using a surface mesh obtained from the same image. Finally, the results show that the proposed technique is an efficient and accurate method to determine voxelized fields of grain boundary normals.

36 MATERIALS SCIENCE↗

Sparse Coding for N-Gram Feature Extraction and Training for File Fragment Classification

File fragment classification is an important step in the task of file carving in digital forensics. In file carving, files must be reconstructed based on their content as a result of their fragmented storage on disk or in memory. Existing methods for classification of file fragments typically use hand-engineered features such as byte histograms or entropy measures. In this paper, we propose an approach using sparse coding that enables automated feature extraction. Sparse coding, or sparse dictionary learning, is an unsupervised learning algorithm, and is capable of extracting features based simply on how well those features can be used to reconstruct the original data. With respect to file fragments, we learn sparse dictionaries for n-grams, continuous sequences of bytes, of different sizes. These dictionaries may then be used to estimate n-gram frequencies for a given file fragment, but for significantly larger n-gram sizes than are typically found in existing methods which suffer from combinatorial explosion. To demonstrate the capability of our sparse coding approach, we used the resulting features to train standard classifiers such as support vector machines (SVMs) over multiple file types. Experimentally, we achieved significantly better classification results with respect to existing methods, especially when the features were used in supplement to existing hand-engineered features.

97 MATHEMATICS AND COMPUTING↗

Artificial Intelligence

Artificial intelligence is the study of intelligence agents as demonstrated by machines. It is an interdisciplinary field involving computer science as well as, various kinds of engineering and science, for example, robotics, bio-medical engineering, that accentuates automation of human acts and intelligence through machines. AI represents state-of-the-art use of machines to bring about algorithmic computation and understanding of tasks that include learning, problem solving, mapping, perception, and reasoning. Given the data and a description of its properties and relations between objects of interest, AI methods can perform the aforementioned tasks. Widely applied AI capabilities, e.g. learning, are now achievable at large scale through machine learning (ML), large volumes of data and specialized computational machines. ML encompasses learning without any kind of supervision (unsupervised learning) and learning with full supervision (supervised learning). Widely applied supervised learning techniques include deep learning and other machine learning methods that require less data than deep learning e.g. support vector machines, random forests. Unsupervised learning examples include dictionary learning, independent component analysis, and autoencoders. For application tasks with less labeled data, both supervised and unsupervised techniques can be adapted in a semi-supervised manner to produce accurate models and to increase the size of the labeled training data.

97 MATHEMATICS AND COMPUTING↗

Nonlinear Matrix Approximation with Radial Basis Function Components

We introduce and investigate matrix approximation by decomposition into a sum of radial basis function (RBF) components. An RBF component is a generalization of the outer product between a pair of vectors, where an RBF function replaces the scalar multiplication between individual vector elements. Even though the RBF functions are positive definite, the summation across components is not restricted to convex combinations and allows us to compute the decomposition for any real matrix that is not necessarily symmetric or positive definite. We formulate the problem of seeking such a decomposition as an optimization problem with a nonlinear and non-convex loss function. Several modern versions of the gradient descent method, including their scalable stochastic counterparts, are used to solve this problem. We provide extensive empirical evidence of the effectiveness of the RBF decomposition and that of the gradient-based fitting algorithm. While being conceptually motivated by singular value decomposition (SVD), our proposed nonlinear counterpart outperforms SVD by drastically reducing the memory required to approximate a data matrix with the same L2 error for a wide range of matrix types. For example, it leads to 2 to 6 times memory save for Gaussian noise, graph adjacency matrices, and kernel matrices. Moreover, this proximity-based decomposition can offer additional interpretability in applications that involve, e.g., capturing the inner low-dimensional structure of the data, retaining graph connectivity structure, and preserving the acutance of images.

Rebrova, Elizaveta↗

Inverse transformation: unleashing spatially heterogeneous dynamics with an alternative approach to XPCS data analysis

X-ray photon correlation spectroscopy (XPCS), an extension of dynamic light scattering (DLS) in the X-ray regime, detects temporal intensity fluctuations of coherent speckles and provides scattering-vector-dependent sample dynamics at length scales smaller than DLS. The penetrating power of X-rays enables XPCS to probe the dynamics in a broad array of materials, including polymers, glasses and metal alloys, where attempts to describe the dynamics with a simple exponential fit usually fail. In these cases, the prevailing XPCS data analysis approach employs stretched or compressed exponential decay functions (Kohlrausch functions), which implicitly assume homogeneous dynamics. This paper proposes an alternative analysis scheme based upon inverse Laplace or Gaussian transformation for elucidating heterogeneous distributions of dynamic time scales in XPCS, an approach analogous to the CONTIN algorithm widely accepted in the analysis of DLS from polydisperse and multimodal systems. In conclusion, using XPCS data measured from colloidal gels, it is demonstrated that the inverse transform approach reveals hidden multimodal dynamics in materials, unleashing the full potential of XPCS.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Temporal network alignment via GoT-WAVE

Abstract Motivation Network alignment (NA) finds conserved regions between two networks. NA methods optimize node conservation (NC) and edge conservation. Dynamic graphlet degree vectors are a state-of-the-art dynamic NC measure, used within the fastest and most accurate NA method for temporal networks: DynaWAVE. Here, we use graphlet-orbit transitions (GoTs), a different graphlet-based measure of temporal node similarity, as a new dynamic NC measure within DynaWAVE, resulting in GoT-WAVE. Results On synthetic networks, GoT-WAVE improves DynaWAVE’s accuracy by 30% and speed by 64%. On real networks, when optimizing only dynamic NC, the methods are complementary. Furthermore, only GoT-WAVE supports directed edges. Hence, GoT-WAVE is a promising new temporal NA algorithm, which efficiently optimizes dynamic NC. We provide a user-friendly user interface and source code for GoT-WAVE. Availability and implementation http://www.dcc.fc.up.pt/got-wave/ Supplementary information Supplementary data are available at Bioinformatics online.

Aparício, David↗

COSMIC FLOW FROM TWO MICRON ALL-SKY REDSHIFT SURVEY: THE ORIGIN OF COSMIC MICROWAVE BACKGROUND DIPOLE AND IMPLICATIONS FOR LAMBDACDM COSMOLOGY

We generate the peculiar velocity field for the Two Micron All-Sky Redshift Survey (2MRS) catalog using an orbit-reconstruction algorithm. The reconstructed velocities of individual objects in 2MRS are well correlated with the peculiar velocities obtained from high-precision observed distances within 3000 km s{sup -1}. We estimate the mean matter density to be OMEGA{sub m} = 0.31 +- 0.05 by comparing observed to reconstructed velocities in this volume. The reconstructed motion of the Local Group in the rest frame established by distances within 3000 km s{sup -1} agrees with the observed motion and is generated by fluctuations within this volume, in agreement with observations. Having tested our method against observed distances, we reconstruct the velocity field of 2MRS in successively larger radii, to study the problem of convergence toward the cosmic microwave background (CMB) dipole. We find that less than half of the amplitude of the CMB dipole is generated within a volume enclosing the Hydra-Centaurus-Norma supercluster at around 40 h {sup -1} Mpc. Although most of the amplitude of the CMB dipole seems to be recovered by 120 h {sup -1} Mpc, the direction does not agree and hence we observe no convergence up to this scale. Due to dominant superclusters such as Shapley or Horologium-Reticulum in the southern hemisphere at scales above 120 h {sup -1} Mpc, one might need to go well beyond 200 h {sup -1} Mpc to fully recover the dipole vector. We develop a statistical model which allows us to estimate cosmological parameters from the reconstructed growth of convergence of the velocity of the Local Group toward the CMB dipole motion. For scales up to 60 h {sup -1} Mpc, assuming a Local Group velocity of 627 km s{sup -1}, we estimate OMEGA{sub m} h {sup 2} = 0.11 +- 0.06 and sigma{sub 8} = 0.9 +- 0.4, in agreement with WMAP5 measurements at the 1sigma level. However, for scales up to 100 h {sup -1} Mpc, we obtain OMEGA{sub m} h {sup 2} = 0.08 +- 0.03 and sigma{sub 8} = 1.0 +- 0.4, which agrees at the 1sigma to 2sigma level with WMAP5 results.

79 ASTRONOMY AND ASTROPHYSICS↗

Quasioptical modeling of wave beams with and without mode conversion. I. Basic theory

This work opens a series of papers where we develop a general quasi-optical theory for mode-converting electromagnetic beams in plasma and implement it in a numerical algorithm. Here, the basic theory is introduced. We consider a general quasimonochromatic multicomponent wave in a weakly inhomogeneous linear medium with no sources. For any given dispersion operator that governs the wave field, we explicitly calculate the approximate operator that governs the wave envelope ψ to the second order in the geometrical-optics parameter. Then, we further simplify this envelope operator by assuming that the gradient of ψ transverse to the local group velocity is much larger than the corresponding parallel gradient. This leads to a parabolic differential equation for ψ (“quasioptical equation”) on the basis of the geometrical-optics polarization vectors. Scalar and mode-converting vector beams are described on the same footing. Here, we also explain how to apply this model to electromagnetic waves in general. In the next papers of this series, we report successful quasioptical modeling of radio frequency wave beams in magnetized plasma based on this theory.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Sparse Matrix-Vector Multiplication on Multicore and Accelerators

This chapter consolidates recent work on the development of high performance multicore and accelerator-based implementations of sparse matrix-vector multiplication (SpMV). As an object of study, SpMV is an interesting computation for two key reasons. First, it appears widely in applications in scientific and engineering computing, financial and economic modeling, and information retrieval, among others, and is therefore of great practical interest. Secondly, it is both simple to describe but challenging to implement well, since its performance is limited by a variety of factors, including low computational intensity, potentially highly irregular memory access behavior, and a strong input dependence that be known only at run time. Thus, we believe SpMV is both practically important and provides important insights for understanding the algorithmic and implementation principles necessary to making effective use of state-of-the-art systems.

43 PARTICLE ACCELERATORS↗

Quantifying the Impact of Single Bit Flips on Floating Point Arithmetic

In high-end computing, the collective surface area, smaller fabrication sizes, and increasing density of components have led to an increase in the number of observed bit flips. If mechanisms are not in place to detect them, such flips produce silent errors, i.e. the code returns a result that deviates from the desired solution by more than the allowed tolerance and the discrepancy cannot be distinguished from the standard numerical error associated with the algorithm. These phenomena are believed to occur more frequently in DRAM, but logic gates, arithmetic units, and other circuits are also susceptible to bit flips. Previous work has focused on algorithmic techniques for detecting and correcting bit flips in specific data structures, however, they suffer from lack of generality and often times cannot be implemented in heterogeneous computing environment. Our work takes a novel approach to this problem. We focus on quantifying the impact of a single bit flip on specific floating-point operations. We analyze the error induced by flipping specific bits in the most widely used IEEE floating-point representation in an architecture-agnostic manner, i.e., without requiring proprietary information such as bit flip rates and the vendor-specific circuit designs. We initially study dot products of vectors and demonstrate that not all bit flips create a large error and, more importantly, expected value of the relative magnitude of the error is very sensitive on the bit pattern of the binary representation of the exponent, which strongly depends on scaling. Our results are derived analytically and then verified experimentally with Monte Carlo sampling of random vectors. Furthermore, we consider the natural resilience properties of solvers based on the fixed point iteration and we demonstrate how the resilience of the Jacobi method for linear equations can be significantly improved by rescaling the associated matrix.

Elliott, James J.↗

Detecting Process Equipment Failures Using Acoustic Data and Machine Learning

Nuclear power plant (NPP) process equipment such as fans, motors, valves, and pumps generate frequent or continuous noise, and deviations from the normal operational sounds made by this equipment can indicate potential issues. These deviations can be identified via automated acoustic anomaly detection, which involves using acoustic sensors (i.e., microphones) alongside detection algorithms to continuously monitor for changes in acoustic signatures. This task is made challenging by the substantial background noise that exists, such as operators opening and closing doors, manipulating valves, and conversing—in addition to typical plant noises. In collaboration with a nuclear power utility partner, this effort assessed the efficacy of acoustic anomaly detection when using a specific acoustic sensor that compresses data into a fixed set of features that are transferable over a standard Internet of Things communication protocol, thereby improving usability but potentially degrading detection performance. Two methods of performing automated acoustic anomaly detection were evaluated: one-class support vector machine (OC-SVM) and isolation forest (iForest). To enable the use of high-quality acoustic data encompassing both normal and anomalous conditions, the study utilized the publicly available Malfunctioning Industrial Machine Investigation and Inspection dataset, which includes real measured acoustic sensor data for a range of equipment types, model numbers, and signal-to-noise ratios (SNRs), along with a benchmark set of detection results. Using this dataset, the methods were tested and then compared against the benchmark results. The results indicated that although the specific acoustic sensor did not enable as rich a feature set extraction, the proposed methods with the limited feature set performed just as well. This provides solid justification for both the methods and the use of the proposed acoustic sensor.

46 - INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AN↗

Conformal killing tensors and covariant Hamiltonian dynamics

A covariant algorithm for deriving the conserved quantities for natural Hamiltonian systems is combined with the non-relativistic framework of Eisenhart, and of Duval, in which the classical trajectories arise as geodesics in a higher dimensional space-time, realized by Brinkmann manifolds. Conserved quantities which are polynomial in the momenta can be built using time-dependent conformal Killing tensors with flux. The latter are associated with terms proportional to the Hamiltonian in the lower dimensional theory and with spectrum generating algebras for higher dimensional quantities of order 1 and 2 in the momenta. Illustrations of the general theory include the Runge-Lenz vector for planetary motion with a time-dependent gravitational constant G(t), motion in a time-dependent electromagnetic field of a certain form, quantum dots, the Hénon-Heiles and Holt systems, respectively, providing us with Killing tensors of rank that ranges from one to six.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Determination of ground and excited state dipole moments via electronic Stark spectroscopy: 5-methoxyindole

The dipole moments of the ground and lowest electronically excited singlet state of 5-methoxyindole have been determined by means of optical Stark spectroscopy in a molecular beam. The resulting spectra arise from a superposition of different field configurations, one with the static electric field almost parallel to the polarization of the exciting laser radiation, the other nearly perpendicular. Each field configuration leads to different intensities in the rovibronic spectrum. With an automated evolutionary algorithm approach, the spectra can be fit and the ratio of both field configurations can be determined. A simultaneous fit of two spectra with both field configurations improved the precision of the dipole moment determination by a factor of two. We find a reduction of the absolute dipole moment from 1.59(3) D to 1.14(6) D upon electronic excitation to the lowest electronically excited singlet state. At the same time, the dipole moment orientation rotates by 54{sup ∘} showing the importance of the determination of the dipole moment components. The dipole moment in the electronic ground state can approximately be obtained from a vector addition of the indole and the methoxy group dipole moments. However, in the electronically excited state, vector addition completely fails to describe the observed dipole moment. Several reasons for this behavior are discussed.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Accelerating high-order continuum kinetic plasma simulations using multiple GPUs

Kinetic plasma simulations solve the Vlasov-Poisson or Vlasov-Maxwell equations to evolve scalar-variable distribution functions in position-velocity phase space and vector-variable electromagnetic fields in configuration space. The immense computational cost of evolving high-dimensional variables, and their large number of degrees of freedom, often limits the utility of continuum kinetic simulations and presents a challenge when it comes to accurately simulating real-world physical phenomena. To address this challenge, we present techniques that accelerate and minimize the computational work required for a scalable Vlasov-Poisson solver. We show theoretical hardware compute and communication bounds for solving a fourth-order finite-volume Vlasov-Poisson system. These bounds are then used to inform and evaluate the design of performance portable algorithms for a multiple graphics processing unit (GPU) accelerated version of the Vlasov-Poisson solver VCK-CPU [1]. We demonstrate that the multi-GPU Vlasov solver implementation, VCK-GPU, simultaneously minimizes required inter-process data transfer while also being bounded by the machine network performance limits. This results in an overall strong scaling speedup per timestep of up to 40x in three-dimensional phase space (one position, two velocity coordinates) and 54x in four dimensional phase space (two position, two velocity coordinates) and a 341x increase in simulation throughput of the GPU accelerated code over the existing CPU code. The GPU code is also able to weak scale up to 256 compute nodes and 1024 GPUs. In conclusion, we demonstrate that the improved compute performance enables exploring configurations which were previously computationally infeasible, including resolving fine-scale distribution function filamentation and multi-species dynamics with realistic electron-proton mass ratios.

Continuum kinetics↗

Optimized pulsed write schemes improve linearity and write speed for low-power organic neuromorphic devices

Neuromorphic devices are becoming increasingly appealing as efficient emulators of neural networks used to model real world problems. However, no hardware to date has demonstrated the necessary high accuracy and energy efficiency gain over CMOS in both training via backpropagation and in read via vector matrix multiplication. Such shortcomings are due to device non-idealities, particularly asymmetric conductance tuning in response to uniform voltage pulse inputs. Here, by formulating a general circuit model for capacitive ion-exchange neuromorphic devices, we show that asymmetric nonlinearity in organic electrochemical neuromorphic devices (ENODes) can be suppressed by an appropriately chosen write scheme. Simulations based upon our model suggest that a nonlinear write-selector could reduce the switching voltage and energy, enabling analog tuning via a continuous set of resistance states (100 states) with extremely low switching energy (~170 fJ • µm –2 ). Furthermore, this work clarifies the pathway to neural algorithm accelerators capable of parallelism during both read and write operations.

97 MATHEMATICS AND COMPUTING↗