Search NASA⌕ Search

SEARCH · Search NASA

Results for “Approximate Computing”

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 1,171 records · Page 65
All sourcesNASA NTRSDOE OSTI
DOE OSTI2018

Neural Networks as Surrogates of Nonlinear High-Dimensional Parameter-to-Prediction Maps

We present a preliminary investigation of the use of Multi-Layer Perceptrons (MLP) and Recurrent Neural Networks (RNNs) as surrogates of parameter-to-prediction maps of computational expensive dynamical models. In particular, we target the approximation of Quantities of Interest (QoIs) derived from the solution of a Partial Differential Equations (PDEs) at different time instants. In order to limit the scope of our study while targeting a relevant application, we focus on the problem of computing variations in the ice sheets mass (our QoI), which is a proxy for global mean sea-level changes. We present a number of neural network formulations and compare their performance with that of Polynomial Chaos Expansions (PCE) constructed on the same data.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

Towards time-dependent current-density-functional theory in the non-linear regime

Time-Dependent Density-Functional Theory (TDDFT) is a well-established theoretical approach to describe and understand irradiation processes in clusters and molecules. However, within the so-called adiabatic local density approximation (ALDA) to the exchange-correlation (xc) potential, TDDFT can show insufficiencies, particularly in violently dynamical processes. This is because within ALDA the xc potential is instantaneous and is a local functional of the density, which means that this approximation neglects memory effects and long-range effects. A way to go beyond ALDA is to use Time-Dependent Current-Density-Functional Theory (TDCDFT), in which the basic quantity is the current density rather than the density as in TDDFT. This has been shown to offer an adequate account of dissipation in the linear domain when the Vignale-Kohn (VK) functional is used. Here, we go beyond the linear regime and we explore this formulation in the time domain. In this case, the equations become very involved making the computation out of reach; we hence propose an approximation to the VK functional which allows us to calculate the dynamics in real time and at the same time to keep most of the physics described by the VK functional. We apply this formulation to the calculation of the time-dependent dipole moment of Ca, Mg and Na{sub 2}. Our results show trends similar to what was previously observed in model systems or within linear response. In the non-linear domain, our results show that relaxation times do not decrease with increasing deposited excitation energy, which sets some limitations to the practical use of TDCDFT in such a domain of excitations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2025

Direct numerical simulations for hybrid rocket boundary layers: Performance modeling and scaling

This paper presents a comprehensive performance and scaling analysis of direct numerical simulations for reacting boundary layers, focusing on slab burner configurations. Using a PETSc-based finite volume CFD framework, the study evaluates the scalability and computational cost of flow, chemistry, and radiation evaluations across 2D and 3D simulations. Polymethyl methacrylate (PMMA) is the fuel with pure O 2 as the oxidizer, modeled using a detailed chemical kinetics mechanism with 113 species and 660 reactions. A ray-tracing-based radiation solver, designed for distributed memory applications, is implemented to model radiation heat transfer. Parallel scalability is analyzed for the coupled flow, chemistry, and radiation heat transfer processes. Weak and strong scaling studies are conducted on up to 15,000 computational ranks, revealing robust performance when flow cells exceed 200 per rank. Chemistry evaluations dominate the computational cost in large 3D simulations, accounting for approximately 40% of the total runtime, while flow processes contribute around 35%, and radiation solver contributions remain below 10% due to reduced evaluation frequencies. GPU accelerated chemistry evaluation, implemented with Zero-RK, demonstrates significant promise, achieving up to a 4x speedup for workloads exceeding 30,000 cells per GPU. However, diminishing returns are observed for smaller workloads due to CPU-GPU communication overhead. This study identifies key challenges, including memory bottlenecks and the effects of domain partitioning on flow scalability, while highlighting the potential of GPU-accelerated chemistry to reduce computational costs. In conclusion, these findings provide realizable run configurations for 2D, 3D, and GPU-accelerated cases, offering insights for optimizing reactive flow solvers.

CFD Scalability↗
DOE OSTI2014

Magneto-optical absorption in semiconducting spherical quantum dots: Influence of the dot-size, confining potential, and magnetic field

Semiconducting quantum dots – more fancifully dubbed artificial atoms – are quasi-zero dimensional, tiny, man-made systems with charge carriers completely confined in all three dimensions. The scientific quest behind the synthesis of quantum dots is to create and control future electronic and optical nanostructures engineered through tailoring size, shape, and composition. The complete confinement – or the lack of any degree of freedom for the electrons (and/or holes) – in quantum dots limits the exploration of spatially localized elementary excitations such as plasmons to direct rather than reciprocal space. Here we embark on a thorough investigation of the magneto-optical absorption in semiconducting spherical quantum dots characterized by a confining harmonic potential and an applied magnetic field in the symmetric gauge. This is done within the framework of Bohm-Pines’ random-phase approximation that enables us to derive and discuss the full Dyson equation that takes proper account of the Coulomb interactions. As an application of our theoretical strategy, we compute various single-particle and many-particle phenomena such as the Fock-Darwin spectrum; Fermi energy; magneto-optical transitions; probability distribution; and the magneto-optical absorption in the quantum dots. It is observed that the role of an applied magnetic field on the absorption spectrum is comparable to that of a confining potential. Increasing (decreasing) the strength of the magnetic field or the confining potential is found to be analogous to shrinking (expanding) the size of the quantum dots: resulting into a blue (red) shift in the absorption spectrum. The Fermi energy diminishes with both increasing magnetic-field and dot-size; and exhibits saw-tooth-like oscillations at large values of field or dot-size. Unlike laterally confined quantum dots, both (upper and lower) magneto-optical transitions survive even in the extreme instances. However, the intra-Landau level transitions are seen to be forbidden. The spherical quantum dots have an edge over the strictly two-dimensional quantum dots in that the additional (magnetic) quantum number makes the physics richer (but complex). A deeper grasp of the Coulomb blockade, quantum coherence, and entanglement can lead to a better insight into promising applications involving lasers, detectors, storage devices, and quantum computing.

36 MATERIALS SCIENCE↗
DOE OSTI2013

THE BEHAVIOR OF TRANSVERSE WAVES IN NONUNIFORM SOLAR FLUX TUBES. I. COMPARISON OF IDEAL AND RESISTIVE RESULTS

Magnetohydrodynamic (MHD) waves are ubiquitously observed in the solar atmosphere. Kink waves are a type of transverse MHD waves in magnetic flux tubes that are damped due to resonant absorption. The theoretical study of kink MHD waves in solar flux tubes is usually based on the simplification that the transverse variation of density is confined to a nonuniform layer much thinner than the radius of the tube, i.e., the so-called thin boundary approximation. Here, we develop a general analytic method to compute the dispersion relation and the eigenfunctions of ideal MHD waves in pressureless flux tubes with transversely nonuniform layers of arbitrary thickness. Results for kink waves are produced and compared with fully numerical resistive MHD eigenvalue computations in the limit of small resistivity. We find that the frequency and resonant damping rate are the same in both ideal and resistive cases. The actual results for thick nonuniform layers deviate from the behavior predicted in the thin boundary approximation and strongly depend on the shape of the nonuniform layer. The eigenfunctions in ideal MHD are very different from those in resistive MHD. The ideal eigenfunctions display a global character regardless of the thickness of the nonuniform layer, while the resistive eigenfunctions are localized around the resonance and are indistinguishable from those of ordinary resistive Alfvén modes. Consequently, the spatial distribution of wave energy in the ideal and resistive cases is dramatically different. This poses a fundamental theoretical problem with clear observational consequences.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2015

An adhesive contact mechanics formulation based on atomistically induced surface traction

In this work, we have developed a novel multiscale computational contact formulation based on the generalized Derjuguin approximation for continua that are characterized by atomistically enriched constitutive relations in order to study macroscopic interaction between arbitrarily shaped deformable continua. The proposed adhesive contact formulation makes use of the microscopic interaction forces between individual particles in the interacting bodies. In particular, the double-layer volume integral describing the contact interaction (energy, force vector, matrix) is converted into a double-layer surface integral through a mathematically consistent approach that employs the divergence theorem and a special partitioning technique. The proposed contact model is formulated in the nonlinear continuum mechanics framework and implemented using the standard finite element method. With no large penalty constant, the stiffness matrix of the system will in general be well-conditioned, which is of great significance for quasi-static analysis. Three numerical examples are presented to illustrate the capability of the proposed method. Results indicate that with the same mesh configuration, the finite element computation based on the surface integral approach is faster and more accurate than the volume integral based approach. In addition, the proposed approach is energy preserving even in a very long dynamic simulation.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2023

Evidence for the utility of quantum computing before fault tolerance

Quantum computing promises to offer substantial speed-ups over its classical counterpart for certain problems. However, the greatest impediment to realizing its full potential is noise that is inherent to these systems. The widely accepted solution to this challenge is the implementation of fault-tolerant quantum circuits, which is out of reach for current processors. Here we report experiments on a noisy 127-qubit processor and demonstrate the measurement of accurate expectation values for circuit volumes at a scale beyond brute-force classical computation. We argue that this represents evidence for the utility of quantum computing in a pre-fault-tolerant era. These experimental results are enabled by advances in the coherence and calibration of a superconducting processor at this scale and the ability to characterize and controllably manipulate noise across such a large device. We establish the accuracy of the measured expectation values by comparing them with the output of exactly verifiable circuits. In the regime of strong entanglement, the quantum computer provides correct results for which leading classical approximations such as pure-state-based 1D (matrix product states, MPS) and 2D (isometric tensor network states, isoTNS) tensor network methods break down. These experiments demonstrate a foundational tool for the realization of near-term quantum applications.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Metaplectic geometrical optics for modeling caustics in uniform and non-uniform media

As an approximate theory that is highly regarded for its computational efficiency, geometrical optics (GO) is widely used for modeling waves in various areas of physics. However, GO fails at caustics, which significantly limits its applicability. A new framework, called metaplectic geometrical optics (MGO), has recently been developed that allows caustics of certain types to be modeled accurately within the GO framework. Here, we extend MGO to the most general case. Further, to illustrate our new theory, we also apply it to several sample problems, including calculations of two-dimensional wavefields near fold and cusp caustics. In contrast with traditional GO solutions, the corresponding MGO solutions are finite everywhere and approximate the true wavefield well near these caustics.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

An adaptive Hessian approximated stochastic gradient MCMC method

Bayesian approaches have been successfully integrated into training deep neural networks. One popular family is stochastic gradient Markov chain Monte Carlo methods (SG-MCMC), which have gained increasing interest due to their ability to handle large datasets and the potential to avoid overfitting. Although standard SG-MCMC methods have shown great performance in a variety of problems, they may be inefficient when the random variables in the target posterior densities have scale differences or are highly correlated. Here, we present an adaptive Hessian approximated stochastic gradient MCMC method to incorporate local geometric information while sampling from the posterior. The idea is to apply stochastic approximation (SA) to sequentially update a preconditioning matrix at each iteration. The preconditioner possesses second-order information and can guide the random walk of a sampler efficiently. Instead of computing and saving the full Hessian of the log posterior, we use limited memory of the samples and their stochastic gradients to approximate the inverse Hessian-vector multiplication in the updating formula. Moreover, by smoothly optimizing the preconditioning matrix via SA, our proposed algorithm can asymptotically converge to the target distribution with a controllable bias under mild conditions. To reduce the training and testing computational burden, we adopt a magnitude-based weight pruning method to enforce the sparsity of the network. Our method is user-friendly and demonstrates better learning results compared to standard SG-MCMC updating rules. The approximation of inverse Hessian alleviates storage and computational complexities for large dimensional models. Numerical experiments are performed on several problems, including sampling from 2D correlated distribution, synthetic regression problems, and learning the numerical solutions of heterogeneous elliptic PDE. The numerical results demonstrate great improvement in both the convergence rate and accuracy.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2013

Computing the partition function, ensemble averages, and density of states for lattice spin systems by sampling the mean

An algorithm to approximately calculate the partition function (and subsequently ensemble averages) and density of states of lattice spin systems through non-Monte-Carlo random sampling is developed. This algorithm (called the sampling-the-mean algorithm) can be applied to models where the up or down spins at lattice nodes interact to change the spin states of other lattice nodes, especially non-Ising-like models with long-range interactions such as the biological model considered here. Because it is based on the Central Limit Theorem of probability, the sampling-the-mean algorithm also gives estimates of the error in the partition function, ensemble averages, and density of states. Easily implemented parallelization strategies and error minimizing sampling strategies are discussed. The sampling-the-mean method works especially well for relatively small systems, systems with a density of energy states that contains sharp spikes or oscillations, or systems with little a priori knowledge of the density of states.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Globally Optimizing QAOA Circuit Depth for Constrained Optimization Problems

We develop a global variable substitution method that reduces n-variable monomials in combinatorial optimization problems to equivalent instances with monomials in fewer variables. We apply this technique to 3-SAT and analyze the optimal quantum unitary circuit depth needed to solve the reduced problem using the quantum approximate optimization algorithm. For benchmark 3-SAT problems, we find that the upper bound of the unitary circuit depth is smaller when the problem is formulated as a product and uses the substitution method to decompose gates than when the problem is written in the linear formulation, which requires no decomposition.

3-SAT↗
DOE OSTI2023

Hierarchical off-diagonal low-rank approximation of Hessians in inverse problems, with application to ice sheet model initialization

Obtaining lightweight and accurate approximations of discretized objective functional Hessians in inverse problems governed by partial differential equations (PDEs) is essential to make both deterministic and Bayesian statistical large-scale inverse problems computationally tractable. The cubic computational complexity of dense linear algebraic tasks, such as Cholesky factorization, that provide a means to sample Gaussian distributions and determine solutions of Newton linear systems is a computational bottleneck at large-scale. These tasks can be reduced to log-linear complexity by utilizing hierarchical off-diagonal low-rank (HODLR) matrix approximations. In this work, we show that a class of Hessians that arise from inverse problems governed by PDEs are well approximated by the HODLR matrix format. In particular, we study inverse problems governed by PDEs that model the instantaneous viscous flow of ice sheets. In these problems, we seek a spatially distributed basal sliding parameter field such that the flow predicted by the ice sheet model is consistent with ice sheet surface velocity observations. Here, we demonstrate the use of HODLR Hessian approximation to efficiently sample the Laplace approximation of the posterior distribution with covariance further approximated by HODLR matrix compression. Computational studies are performed which illustrate ice sheet problem regimes for which the Gauss–Newton data-misfit Hessian is more efficiently approximated by the HODLR matrix format than the low-rank (LR) format. We then demonstrate that HODLR approximations can be favorable, when compared to global LR approximations, for large-scale problems by studying the data-misfit Hessian associated with inverse problems governed by the first-order Stokes flow model on the Humboldt glacier and Greenland ice sheet.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2014

Variational principles with Padé approximants for tearing mode analysis

Tearing modes occur in several distinct physical regimes, and it is often important to compute the inner layer response for these modes with various effects. There is a need for an approximate and efficient method of solving the inner layer equations in all these regimes. In this paper, we introduce a method of solving the inner layer equations based on using a variational principle with Padé approximants. For all the regimes considered, the main layer equations to be solved are inhomogeneous, and Padé approximants give a convenient and efficient method of satisfying the correct asymptotic behavior at the edge of the layer. Results using this variational principle—Padé approximant method in three of these regimes is presented. These regimes are the constant-ψ resistive-inertial (RI) regime, the constant-ψ viscoresistive regime, and the non-constant-ψ inviscid tearing regime. The last regime includes the constant-ψ RI regime and the inertial regime. The results show that reasonable accuracy can be obtained very efficiently with Padé approximants having a small number of parameters.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2020

Weighted greedy-optimal design of computer experiments for kernel-based and Gaussian process model emulation and calibration

This article is concerned with the approximation of high-dimensional functions by kernel-based methods. Motivated by uncertainty quantification, which often necessitates the construction of approximations that are accurate with respect to a probability density function of random variables, we aim at minimizing the approximation error with respect to a weighted $L^p$-norm. We present a greedy procedure for designing computer experiments based upon a weighted modification of the pivoted Cholesky factorization. The method successively generates nested samples with the goal of minimizing error in regions of high probability. Numerical experiments validate that this new importance sampling strategy is superior to other sampling approaches, especially when used with non-product probability density functions. We also show how to use the proposed algorithm to efficiently generate surrogates for inferring unknown model parameters from data.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Data-driven modeling and control of dynamical systems using Koopman and Perron-Frobenius operators

This dissertation studies the data-driven modeling and control problem of nonlinear systems by exploiting the linear operator theoretic framework involving Koopman and Perro-Frobenius operator. A systematic linear-operator based controller design procedure has been established, which can be used to solve a variety of nonlinear control problems, including feedback stabilization using control Lyapunov functions, optimal quadratic regulation using Koopman eigenfunctions and convex optimization formulation of optimal control problem using P-F and Koopman operator approximation. As the core of data-driven modeling, we first propose a new algorithm for the finite-dimensional approximation of the linear transfer Koopman and Perron-Frobenius operator from time-series data. We argue that the existing approach for the finite-dimensional approximation of these transfer operators such as Dynamic Mode Decomposition (DMD) and Extended Dynamic Mode Decomposition (EDMD) do not capture two important properties of these operators, namely positivity and Markov property. The algorithm we propose preserves these two properties. We call the proposed algorithm as naturally structured DMD (NSDMD) since it retains the inherent properties of these operators. Naturally structured DMD algorithm leads to a better approximation of the steady-state dynamics of the system regarding computing Koopman and Perron- Frobenius operator eigenfunctions and eigenvalues. However, preserving positivity property is critical for capturing the real transient dynamics of the system. This positivity property of the transfer operators and it's finite-dimensional approximation play an important role for controller and estimator design of nonlinear systems. To solve the feedback stabilization problem for nonlinear control systems, we tried to take advantage of the Koopman operator framework. The Koopman operator approach provides a linear representation for a nonlinear dynamical system and a bilinear representation for a nonlinear control system. The problem of feedback stabilization of a nonlinear control system is then transformed to the stabilization of a bilinear control system. We propose a control Lyapunov function (CLF)-based approach for the design of stabilizing feedback controllers for the bilinear system. The search for finding a CLF for the bilinear control system is formulated as a convex optimization problem. This leads to a schematic procedure for designing CLF-based stabilizing feedback controllers for the bilinear system and hence the original nonlinear system. Another advantage of the proposed controller design approach outlined in this dissertation is that it does not require explicit knowledge of system dynamics. In particular, the bilinear representation of a nonlinear control system in the Koopman eigenfunction space can be obtained from time-series data. Next, we study the optimal quadratic regulation problem for nonlinear systems. The linear operator theoretic framework involving the Koopman operator is used to lift the dynamics of nonlinear control system to an infinite-dimensional bilinear system. The optimal quadratic regulation problem for nonlinear system is formulated in terms of the finite-dimensional approximation of the bilinear system. A convex optimization-based approach is proposed for solving the quadratic regulator problem for bilinear system. We applied a variety of examples and compared the simulation results between our framework and conventional LQR control using linearized model. For more general optimal control problems, we provide a density-function based convex formulation for the optimal control problem of the nonlinear system. The convex formulation relies on the duality result in the stability theory of a dynamical system involving density function and Perron-Frobenius operator. The optimal control problem is formulated as an infinite-dimensional convex optimization program. The finite-dimensional approximation of the optimization problem relies on the recent advances made in the data-driven computation of the Koopman operator, which is dual to the Perron-Frobenius operator. Simulation results are presented to demonstrate the application of the developed framework.

Huang, Bowen↗
DOE OSTI2022

Dancoff-based Wigner-Seitz approximation for the subgroup resonance self-shielding in the VERA neutronic simulator MPACT

The MPACT neutronics module of the Virtual Environment for Reactor Analysis (VERA) has used the subgroup method for resonance self-shielding calculation, for which two-dimensional (2D) fixed-source transport calculations are performed using the method of characteristics for resonance energy groups. When considering thermal feedbacks, the subgroup calculation must be performed at each outer iteration. Therefore, the computing time for cross section processing is a significant burden for computational efficiency. The Dancoff-based Wigner-Seitz approximation (DWA) capability has been implemented into MPACT in conjunction with the subgroup method, which has been used in SCALE/XSProc since SCALE version 6.0 and has recently been called an equivalent Dancoff-factor cell (EDC) method. The issue of relatively large reactivity bias in DWA for the gadolinia rods was resolved by introducing multiple Dancoff factors. Benchmark results for the VERA pressurized and boiling water reactor benchmark suites show that the DWA capability would significantly enhance computational efficiency with comparable accuracy. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2014

Non-relativistic leptogenesis

In many phenomenologically interesting models of thermal leptogenesis the heavy neutrinos are non-relativistic when they decay and produce the baryon asymmetry of the Universe. We propose a non-relativistic approximation for the corresponding rate equations in the non-resonant case, and a systematic way for computing relativistic corrections. We determine the leading order coefficients in these equations, and the first relativistic corrections. The non-relativistic approximation works remarkably well. It appears to be consistent with results obtained using a Boltzmann equation taking into account the momentum distribution of the heavy neutrinos, while being much simpler. We also compute radiative corrections to some of the coefficients in the rate equations. Their effect is of order 1% in the regime favored by neutrino oscillation data. We obtain the correct leading order lepton number washout rate in this regime, which leads to large ( ∼ 20%) effects compared to previous computations.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2016

SU-G-IeP3-04: Effective Dose Measurements in Fast Kvp Switch Dual Energy Computed Tomography

Purpose: The objective of this study was two-fold: (a) to test a new approach to approximating organ dose by using the effective energy of the combined 80kV/140kV beam in dual-energy (DE) computed tomography (CT), and (b) to derive the effective dose (ED) in the abdomen-pelvis protocol in DECT. Methods: A commercial dual energy CT scanner was employed using a fast-kV switch abdomen/pelvis protocol alternating between 80 kV and 140 kV. MOSFET detectors were used for organ dose measurements. First, an experimental validation of the dose equivalency between MOSFET and ion chamber (as a gold standard) was performed using a CTDI phantom. Second, the ED of DECT scans was measured using MOSFET detectors and an anthropomorphic phantom. For ED calculations, an abdomen/pelvis scan was used using ICRP 103 tissue weighting factors; ED was also computed using the AAPM Dose Length Product (DLP) method and compared to the MOSFET value. Results: The effective energy was determined as 42.9 kV under the combined beam from half-value layer (HVL) measurement. ED for the dual-energy scan was calculated as 16.49 ± 0.04 mSv by the MOSFET method and 14.62 mSv by the DLP method. Conclusion: Tissue dose in the center of the CTDI body phantom was 1.71 ± 0.01 cGy (ion chamber) and 1.71 ± 0.06 (MOSFET) respectively; this validated the use of effective energy method for organ dose estimation. ED from the abdomen-pelvis scan was calculated as 16.49 ± 0.04 mSv by MOSFET and 14.62 mSv by the DLP method; this suggests that the DLP method provides a reasonable approximation to the ED.

60 APPLIED LIFE SCIENCES↗