Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “task decomposition”

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 109 records · Page 6
All sourcesNASA NTRSDOE OSTI
DOE OSTI2014

RADAR: Runtime Asymmetric Data-access Driven Scientific Data Replication

Efficient I/O on large-scale spatiotemporal scientific data requires scrutiny of both the logical layout of the data (e.g., row-major vs. column-major) and the physical layout (e.g., distribution on parallel filesystems). For increasingly complex datasets, hand optimization is a difficult matter prone to error and not scalable to the increasing heterogeneity of analysis workloads. Given these factors, we present a partial data replication system called RADAR. We capture datatype- and collective-aware I/O access patterns (indicating logical access) via MPI-IO tracing and use a combination of coarse-grained and fine-grained performance modeling to evaluate and select optimized physical data distributions for the task at hand. Unlike conventional methods, we store all replica data and metadata, along with the original untouched data, under a single file container using the object abstraction in parallel filesystems. Our system results in manyfold improvements in some commonly used subvolume decomposition access patterns.Moreover, the modeling approach can determine whether such optimizations should be undertaken in the first place.

access pattern↗
DOE OSTI2026

BoBa

BoBa is a C++ software library for working with large matrices, tensors, and tensor decompositions. The library provides tools for dense matrix and tensor operations, tensor decompositions, and tensor decomposition methods that support modern CPU and GPU architectures. It includes portable abstractions for linear algebra, tensor algebra, and multidimensional computation. BoBa is intended for scientific computing applications that involve large multidimensional data sets or high dimensional mathematical models. Its capabilities support tasks such as data compression, linear algebra, efficient numerical computation, and the development of scalable algorithms for heterogeneous hardware. Tutorials, tests, and example applications are included to help users learn and apply the library.

Yao, Jin [Lawrence Livermore National Laboratory (↗
DOE OSTI2019

From Escherichia coli mutant 13 C labeling data to a core kinetic model: A kinetic model parameterization pipeline

Kinetic models of metabolic networks offer the promise of quantitative phenotype prediction. The mechanistic characterization of enzyme catalyzed reactions allows for tracing the effect of perturbations in metabolite concentrations and reaction fluxes in response to genetic and environmental perturbation that are beyond the scope of stoichiometric models. In this study, we develop a two-step computational pipeline for the rapid parameterization of kinetic models of metabolic networks using a curated metabolic model and available 13 C-labeling distributions under multiple genetic and environmental perturbations. The first step involves the elucidation of all intracellular fluxes in a core model of E. coli containing 74 reactions and 61 metabolites using 13 C-Metabolic Flux Analysis ( 13 C-MFA). Here, fluxes corresponding to the mid-exponential growth phase are elucidated for seven single gene deletion mutants from upper glycolysis, pentose phosphate pathway and the Entner-Doudoroff pathway. The computed flux ranges are then used to parameterize the same (i.e., k-ecoli74) core kinetic model for E. coli with 55 substrate-level regulations using the newly developed K-FIT parameterization algorithm. The K-FIT algorithm employs a combination of equation decomposition and iterative solution techniques to evaluate steady-state fluxes in response to genetic perturbations. k-ecoli74 predicted 86% of flux values for strains used during fitting within a single standard deviation of 13 C-MFA estimated values. By performing both tasks using the same network, errors associated with lack of congruity between the two networks are avoided, allowing for seamless integration of data with model building. Product yield predictions and comparison with previously developed kinetic models indicate shifts in flux ranges and the presence or absence of mutant strains delivering flux towards pathways of interest from training data significantly impact predictive capabilities. Using this workflow, the impact of completeness of fluxomic datasets and the importance of specific genetic perturbations on uncertainties in kinetic parameter estimation are evaluated.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2013

Image segmentation by hierarchial agglomeration of polygons using ecological statistics

A method for rapid hierarchical image segmentation based on perceptually driven contour completion and scene statistics is disclosed. The method begins with an initial fine-scale segmentation of an image, such as obtained by perceptual completion of partial contours into polygonal regions using region-contour correspondences established by Delaunay triangulation of edge pixels as implemented in VISTA. The resulting polygons are analyzed with respect to their size and color/intensity distributions and the structural properties of their boundaries. Statistical estimates of granularity of size, similarity of color, texture, and saliency of intervening boundaries are computed and formulated into logical (Boolean) predicates. The combined satisfiability of these Boolean predicates by a pair of adjacent polygons at a given segmentation level qualifies them for merging into a larger polygon representing a coarser, larger-scale feature of the pixel image and collectively obtains the next level of polygonal segments in a hierarchy of fine-to-coarse segmentations. The iterative application of this process precipitates textured regions as polygons with highly convolved boundaries and helps distinguish them from objects which typically have more regular boundaries. The method yields a multiscale decomposition of an image into constituent features that enjoy a hierarchical relationship with features at finer and coarser scales. This provides a traversable graph structure from which feature content and context in terms of other features can be derived, aiding in automated image understanding tasks. The method disclosed is highly efficient and can be used to decompose and analyze large images.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Data processing pipeline for Tianlai experiment

The Tianlai project is a 21cm intensity mapping experiment for detecting dark energy by measuring the baryon acoustic oscillation (BAO) features in the large scale structure power spectrum. This experiment provides an opportunity to test the data processing methods for cosmological 21cm signal extraction, which is still a great challenge in current radio astronomy research. The 21cm signal is much weaker than the foregrounds and easily aected by the imperfections in the instrumental responses. Furthermore, processing the large volumes of interferometer data poses a practical challenge. We have developed a data processing pipeline called tlpipe to process the drift scan survey data from the Tianlai experiment. It performs oine data processing tasks such as radio frequency interference (RFI) agging, array calibration, binning, and map-making, etc. It also includes utility functions needed for the data analysis, such as data selection, transformation, visualization and others. A number of new algorithms are implemented, for example the eigenvector decomposition method for array calibration and the Tikhnov regularization for m-mode analysis. In this paper we describe the design and implementation of the pipeline and illustrate its functions with some analysis of real data. Finally, we outline directions for future development of this publicly code.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2014

Inverse transport calculations in optical imaging with subspace optimization algorithms

Inverse boundary value problems for the radiative transport equation play an important role in optics-based medical imaging techniques such as diffuse optical tomography (DOT) and fluorescence optical tomography (FOT). Despite the rapid progress in the mathematical theory and numerical computation of these inverse problems in recent years, developing robust and efficient reconstruction algorithms remains a challenging task and an active research topic. We propose here a robust reconstruction method that is based on subspace minimization techniques. The method splits the unknown transport solution (or a functional of it) into low-frequency and high-frequency components, and uses singular value decomposition to analytically recover part of low-frequency information. Minimization is then applied to recover part of the high-frequency components of the unknowns. We present some numerical simulations with synthetic data to demonstrate the performance of the proposed algorithm.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2016

Image quality comparison between single energy and dual energy CT protocols for hepatic imaging

Purpose: Multi-detector computed tomography (MDCT) enables volumetric scans in a single breath hold and is clinically useful for hepatic imaging. For simple tasks, conventional single energy (SE) computed tomography (CT) images acquired at the optimal tube potential are known to have better quality than dual energy (DE) blended images. However, liver imaging is complex and often requires imaging of both structures containing iodinated contrast media, where atomic number differences are the primary contrast mechanism, and other structures, where density differences are the primary contrast mechanism. Hence it is conceivable that the broad spectrum used in a dual energy acquisition may be an advantage. In this work we are interested in comparing these two imaging strategies at equal-dose and more complex settings. Methods: We developed numerical anthropomorphic phantoms to mimic realistic clinical CT scans for medium size and large size patients. MDCT images based on the defined phantoms were simulated using various SE and DE protocols at pre- and post-contrast stages. For SE CT, images from 60 kVp through 140 with 10 kVp steps were considered; for DE CT, both 80/140 and 100/140 kVp scans were simulated and linearly blended at the optimal weights. To make a fair comparison, the mAs of each scan was adjusted to match the reference radiation dose (120 kVp, 200 mAs for medium size patients and 140 kVp, 400 mAs for large size patients). Contrast-to-noise ratio (CNR) of liver against other soft tissues was used to evaluate and compare the SE and DE protocols, and multiple pre- and post-contrasted liver-tissue pairs were used to define a composite CNR. To help validate the simulation results, we conducted a small clinical study. Eighty-five 120 kVp images and 81 blended 80/140 kVp images were collected and compared through both quantitative image quality analysis and an observer study. Results: In the simulation study, we found that the CNR of pre-contrast SE image mostly increased with increasing kVp while for post-contrast imaging 90 kVp or lower yielded higher CNR images, depending on the differential iodine concentration of each tissue. Similar trends were seen in DE blended CNR and those from SE protocols. In the presence of differential iodine concentration (i.e., post-contrast), the CNR curves maximize at lower kVps (80–120), with the peak shifted rightward for larger patients. The combined pre- and post-contrast composite CNR study demonstrated that an optimal SE protocol has better performance than blended DE images, and the optimal tube potential for SE scan is around 90 kVp for a medium size patients and between 90 and 120 kVp for large size patients (although low kVp imaging requires high x-ray tube power to avoid photon starvation). Also, a tin filter added to the high kVp beam is not only beneficial for material decomposition but it improves the CNR of the DE blended images as well. The dose adjusted CNR of the clinical images also showed the same trend and radiologists favored the SE scans over blended DE images. Conclusions: Our simulation showed that an optimized SE protocol produces up to 5% higher CNR for a range of clinical tasks. The clinical study also suggested 120 kVp SE scans have better image quality than blended DE images. Hence, blended DE images do not have a fundamental CNR advantage over optimized SE images.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

Quantum advantage for differential equation analysis

Quantum algorithms for differential equation solving, data processing, and machine learning potentially offer an exponential speedup over all known classical algorithms. However, there also exist obstacles to obtaining this potential speedup in useful problem instances. The essential obstacle for quantum differential equation solving is that outputting useful information may require difficult postprocessing, and the essential obstacle for quantum data processing and machine learning is that inputting the data is a difficult task just by itself. In this study, we demonstrate that, when combined, these difficulties solve one another. We show how the output of quantum differential equation solving can serve as the input for quantum data processing and machine learning, allowing dynamical analysis in terms of principal components, power spectra, and wavelet decompositions. To illustrate this, we consider continuous-time Markov processes on epidemiological and social networks. These quantum algorithms provide an exponential advantage over existing classical Monte Carlo methods.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2012

Event-Based Parareal: A data-flow based implementation of parareal

Parareal is an iterative algorithm that, in effect, achieves temporal decomposition for a time-dependent system of differential or partial differential equations. A solution is obtained in a shorter wall-clock time, but at the expense of increased compute cycles. The algorithm combines a fine solver that solves the system to acceptable accuracy with an approximate coarse solver. The critical task for the successful implementation of parareal on any system is the development of a coarse solver that leads to convergence in a small number of iterations compared to the number of time slices in the full time interval, and is, at the same time, much faster than the fine solver. Fast coarse solvers may not lead to sufficiently rapid convergence, and slow coarse solvers may not lead to significant gains even if the number of iterations to convergence is satisfactory. We find that the difficulty of meeting these conflicting demands can be substantially eased by using a data-driven, event-based implementation of parareal instead of the conventional algorithm where solver tasks are executed sequentially. For given convergence properties, the event-based approach relaxes the speed requirements on the coarse solver by a factor of , where is the number of iterations required for a converged solution. This may, for many problems, lead to an efficient parareal implementation that would otherwise not be possible or would require substantial coarse solver development.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2016

An Efficient Multicore Implementation of a Novel HSS-Structured Multifrontal Solver Using Randomized Sampling

Here, we present a sparse linear system solver that is based on a multifrontal variant of Gaussian elimination and exploits low-rank approximation of the resulting dense frontal matrices. We use hierarchically semiseparable (HSS) matrices, which have low-rank off-diagonal blocks, to approximate the frontal matrices. For HSS matrix construction, a randomized sampling algorithm is used together with interpolative decompositions. The combination of the randomized compression with a fast ULV HSS factoriz ation leads to a solver with lower computational complexity than the standard multifrontal method for many applications, resulting in speedups up to 7 fold for problems in our test suite. The implementation targets many-core systems by using task parallelism with dynamic runtime scheduling. Numerical experiments show performance improvements over state-of-the-art sparse direct solvers. The implementation achieves high performance and good scalability on a range of modern shared memory parallel systems, including the Intel Xeon Phi (MIC). The code is part of a software package called STRUMPACK - STRUctured Matrices PACKage, which also has a distributed memory component for dense rank-structured matrices.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2012

Dynamic load balancing algorithm for molecular dynamics based on Voronoi cells domain decompositions

We present a new algorithm for automatic parallel load balancing in classical molecular dynamics. It assumes a spatial domain decomposition of particles into Voronoi cells. It is a gradient method which attempts to minimize a cost function by displacing Voronoi sites associated with each processor/sub-domain along steepest descent directions. Excellent load balance has been obtained for quasi-2D and 3D practical applications, with up to 440·10 6 particles on 65,536 MPI tasks.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2019

Efficient Solution Methods for Large-scale Optimization Problems Constrained by Time-dependent Partial Differential Equations (Final Report)

The objectives of this project are to devise and analyze efficient solvers for several classes of PDE- constrained optimization problems. The project was initially focused on, but not restricted to time- dependent PDE-constraints such as parabolic equations, time-dependent fluid flows (Navier-Stokes equations), hyperbolic equations (Burgers equation, shallow water equations). Other constraints considered are PDEs with stochastic coefficients. Additional elements of interest are the handling of problems with control and/or state inequality constraints or non-differentiable regularization terms, and boundary controls. A direction of work not initially listed among the goals, which was triggered by conversations with DOE scientists from the Sandia National Labs, is to develop efficient solvers for optimization- based domain-decomposition PDE-solvers. Also targeted in the project are a number of technical issues necessary for facilitating the integration of the methods resulted from this project with existing high-performance packages developed at DOE national labs; this integration is expected to further enable the scientific community to use the novel methods. An example of such a task is the need to develop algebraic multigrid (AMG) versions of the algorithms. A list of accomplishments in several areas is detailed as well as a list of publications.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Bundle Networks: Fiber Bundles, Local Trivializations, and a Generative Approach to Exploring Many-to-one Maps

Many-to-one maps are ubiquitous in machine learning, from the image recognition model that assigns a multitude of distinct images to the concept of ``cat'' to the time series forecasting model which assigns a range of distinct time-series to a single scalar regression value. While the primary use of such models is naturally to associate correct output to each input, in many problems it is also useful to be able to explore, understand, and sample from a model's fibers, which are the set of input values x such that f(x) = y, for fixed y in the output space. In this paper we show that popular generative architectures are ill-suited to such tasks. Motivated by this we introduce a novel generative architecture, a Bundle Network, based on the concept of a fiber bundle from (differential) topology. BundleNets exploit the idea of a local trivialization wherein a space can be locally decomposed into a product space that cleanly encodes the many-to-one nature of the map. By enforcing this decomposition in BundleNets and by utilizing state-of-the-art invertible components, investigating a network's fibers becomes natural.

Courts, Nicolas C.↗
DOE OSTI2015

Solving a Large Scale Thermal Radiation Problem Using an Interoperable Executive Library Framework on Petascale Supercomputers

We present a novel methodology to compute the transient thermal condi- tion of a set of objects in an open space environment. The governing energy equa- tion and the convective energy transfer are solved by the sparse iterative solvers. The average radiating energy on a set of surfaces is represented by a linear system of the radiosity equations, which is factorized by an out-of-core parallel Cholesky decomposition solver. The coupling and interplay of the direct radiosity solver us- ing GPUs and the CPU-based sparse solver are handled by a light weight software integrator called Interoperable Executive Library (IEL). IEL manages the distribu- tion of data and memory, coordinates communication among parallel processes, and also directs execution of the set of loosely coupled physics tasks as warranted by the thermal condition of the simulated object and its surrounding environment.

Wong, Kwai↗
DOE OSTI2024

Distance preserving machine learning for uncertainty aware accelerator capacitance predictions

Abstract Accurate uncertainty estimations are essential for producing reliable machine learning models, especially in safety-critical applications such as accelerator systems. Gaussian process models are generally regarded as the gold standard for this task; however, they can struggle with large, high-dimensional datasets. Combining deep neural networks with Gaussian process approximation techniques has shown promising results, but dimensionality reduction through standard deep neural network layers is not guaranteed to maintain the distance information necessary for Gaussian process models. We build on previous work by comparing the use of the singular value decomposition against a spectral-normalized dense layer as a feature extractor for a deep neural Gaussian process approximation model and apply it to a capacitance prediction problem for the High Voltage Converter Modulators in the Oak Ridge Spallation Neutron Source. Our model shows improved distance preservation and predicts in-distribution capacitance values with less than 1% error.

43 PARTICLE ACCELERATORS↗
DOE OSTI2017

Partial differential equations preconditioner resilient to soft and hard faults

We present a domain-decomposition-based preconditioner for the solution of partial differential equations (PDEs) that is resilient to both soft and hard faults. The algorithm reformulates the PDE as a sampling problem, followed by a solution update through data manipulation that is resilient to both soft and hard faults. This reformulation allows us to recast the problem as a set of independent tasks, and exploit data locality to reduce global communication. We discuss two different parallel implementations: (a) a single program multiple data (SPMD) version based on a one-to-one mapping between subdomain and MPI processes responsible for both state and computation; and (b) an asynchronous server–client implementation where all state information is held by the servers and clients are designed solely as computational units. We present a scalability comparison of both implementations under nominal conditions, showing efficiency within ~80% for up to 12,000 cores. We present a resilience analysis under different fault scenarios based on the server–client implementation. This framework provides resiliency to hard faults such that if a client crashes, it stops asking for work, and the servers simply distribute the work among all of the other clients alive. Erroneous subdomain solves (e.g. due to soft faults) appear as corrupted data, which is either rejected if that causes a task to fail, or is seamlessly filtered out during the regression stage through a suitable noise model. Three different types of faults are modeled: hard faults modeling nodes (or clients) crashing; soft faults occurring during the communication of the tasks between server and clients; and soft faults occurring during task execution. We demonstrate the resiliency of the approach for a 2D elliptic PDE, and explore the effect of the faults at various failure rates.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2013

How real-time cosmology can distinguish between different anisotropic models

We present a new analysis on how to distinguish between isotropic and anisotropic cosmological models based on tracking the angular displacements of a large number of distant quasars over an extended period of time, and then performing a multipole-vector decomposition of the resulting displacement maps. We find that while the GAIA mission operating at its nominal specifications does not have sufficient angular resolution to resolve anisotropic universes from isotropic ones using this method within a reasonable timespan of ten years, a next-generation GAIA-like survey with a resolution ten times better should be equal to the task. Distinguishing between different anisotropic models is however more demanding. Keeping the observational timespan to ten years, we find that the angular resolution of the survey will need to be of order 0.1 μas in order for certain rotating anisotropic models to produce a detectable signature that is also unique to models of this class. However, should such a detection become possible, it would immediately allow us to rule out large local void models.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2018

ExaTN

A challenging scientific problem that will benefit from exascale computing is predictive simulation of two- and three-dimensional quantum many-body Hamiltonians in condensed matter physics, together with the related strong entanglement problems in quantum chemistry and quantum computing. Efficient numerical simulations will require new insights into wavefunction compression algorithms that scale on extremely heterogeneous HPC architectures. To address this need, we propose to develop a scalable math library of parallel multi-linear algebraic numerical primitives for hierarchical tensor compression. Our library will subsequently be used for building an efficient, performance-portable HPC framework for simulating strongly correlated quantum systems. In particular, our will will be the first to deliver a massively parallel implementation of the multiscale entanglement renormalization ansatz (MERA), a hierarchical tensor decomposition quantum scheme capable of expressing local expectation values of strongly entangled higher-dimensional quantum systems efficiently. To address the extreme complexity and high computational cost of such schemes, we will automate their parallel implementation by delivering an integrated software framework, ExaTN (Exascale Tensor Networks), that will process hierarchical tensor representations on exascale HPC systems via an asymchronous task-based programming model. As a specific application, our framework will provide scientists with a capability of simulating strongly entangled quantum systems, thus positioning ORNL as a strategic leader in modeling of strongly correlated materials.

McCaskey, AlexanderJ↗