Search NASA⌕ Search

SEARCH · Search NASA

Results for “Computer Algebra”

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 541 records · Page 30
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

CP decomposition for tensors via alternating least squares with QR decomposition

The CP tensor decomposition is used in applications such as machine learning and signal processing to discover latent low-rank structure in multidimensional data. Computing a CP decomposition via an alternating least squares (ALS) method reduces the problem to several linear least squares problems. The standard way to solve these linear least squares subproblems is to use the normal equations, which inherit special tensor structure that can be exploited for computational efficiency. However, the normal equations are sensitive to numerical ill-conditioning, which can compromise the results of the decomposition. In this paper, we develop versions of the CP-ALS algorithm using the QR decomposition and the singular value decomposition, which are more numerically stable than the normal equations, to solve the linear least squares problems. Our algorithms utilize the tensor structure of the CP-ALS subproblems efficiently, have the same complexity as the standard CP-ALS algorithm when the input is dense and the rank is small, and are shown via examples to produce more stable results when ill-conditioning is present. Our MATLAB implementation achieves the same running time as the standard algorithm for small ranks, and we show that the new methods can obtain lower approximation error.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Hierarchical Model Reduction Driven by Machine Learning for Parametric Advection-Diffusion-Reaction Problems in the Presence of Noisy Data

Abstract We propose a new approach to generate a reliable reduced model for a parametric elliptic problem, in the presence of noisy data. The reference model reduction procedure is the directional HiPOD method, which combines Hierarchical Model reduction with a standard Proper Orthogonal Decomposition, according to an offline/online paradigm. In this paper we show that directional HiPOD looses in terms of accuracy when problem data are affected by noise. This is due to the interpolation driving the online phase, since it replicates, by definition, the noise trend. To overcome this limit, we replace interpolation with Machine Learning fitting models which better discriminate relevant physical features in the data from irrelevant unstructured noise. The numerical assessment, although preliminary, confirms the potentialities of the new approach.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

Interatomic Coulombic decay widths of helium trimer: Ab initio calculations

We report on an extensive study of interatomic Coulombic decay (ICD) widths in helium trimer computed using a fully ab initio method based on the Fano theory of resonances. Algebraic diagrammatic construction for one-particle Green’s function is utilized for the solution of the many-electron problem. An advanced and universal approach to partitioning of the configuration space into discrete states and continuum subspaces is described and employed. Total decay widths are presented for all ICD-active states of the trimer characterized by one-site ionization and additional excitation of an electron into the second shell. Selected partial decay widths are analyzed in detail, showing how three-body effects can qualitatively change the character of certain relaxation transitions. Previously unreported type of three-electron decay processes is identified in one class of the metastable states.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2015

Batched matrix computations on hardware accelerators based on GPUs

Scientific applications require solvers that work on many small size problems that are independent from each other. At the same time, the high-end hardware evolves rapidly and becomes ever more throughput-oriented and thus there is an increasing need for an effective approach to develop energy-efficient, high-performance codes for these small matrix problems that we call batched factorizations. The many applications that need this functionality could especially benefit from the use of GPUs, which currently are four to five times more energy efficient than multicore CPUs on important scientific workloads. This study, consequently, describes the development of the most common, one-sided factorizations, Cholesky, LU, and QR, for a set of small dense matrices. The algorithms we present together with their implementations are, by design, inherently parallel. In particular, our approach is based on representing the process as a sequence of batched BLAS routines that are executed entirely on a GPU. Importantly, this is unlike the LAPACK and the hybrid MAGMA factorization algorithms that work under drastically different assumptions of hardware design and efficiency of execution of the various computational kernels involved in the implementation. Thus, our approach is more efficient than what works for a combination of multicore CPUs and GPUs for the problems sizes of interest of the application use cases. The paradigm where upon a single chip (a GPU or a CPU) factorizes a single problem at a time is not at all efficient in our applications’ context. We illustrate all of these claims through a detailed performance analysis. With the help of profiling and tracing tools, we guide our development of batched factorizations to achieve up to two-fold speedup and three-fold better energy efficiency as compared against our highly optimized batched CPU implementations based on MKL library. Finally, the tested system featured two sockets of Intel Sandy Bridge CPUs and we compared with a batched LU factorizations featured in the CUBLAS library for GPUs, we achieve as high as 2.5× speedup on the NVIDIA K40 GPU.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

An Informal Overview of the Unitary Group Approach

The Unitary Groups Approach (UGA) is an elegant and conceptually unified approach to quantum structure calculations. It has been widely used in molecular structure calculations, and holds the promise of a single computational approach to structure calculations in a variety of different fields. We explore the possibility of extending the UGA to computations in atomic and nuclear structure as a simpler alternative to traditional Racah algebra-based approaches. We provide a simple introduction to the basic UGA and consider some of the issues in using the UGA with spin-dependent, multi-body Hamiltonians requiring multi-shell bases adapted to additional symmetries. While the UGA is perfectly capable of dealing with such problems, it is seen that the complexity rises dramatically, and the UGA is not at this time, a simpler alternative to Racah algebra-based approaches.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2018

Simplifications for hydronic system models in modelica

Building systems and their heating, ventilation and air conditioning flow networks, are becoming increasingly complex. Some building energy simulation tools simulate these flow networks using pressure drop equations. These flow network models typically generate coupled algebraic nonlinear systems of equations, which become increasingly more difficult to solve as their sizes increase. This leads to longer computation times and can cause the solver to fail. These problems also arise when using the equation-based modelling language Modelica and Annex 60-based libraries. This may limit the applicability of the library to relatively small problems unless problems are restructured. This paper discusses two algebraic loop types and presents an approach that decouples algebraic loops into smaller parts, or removes them completely. The approach is applied to a case study model where an algebraic loop of 86 iteration variables is decoupled into smaller parts with a maximum of five iteration variables.

42 ENGINEERING↗
DOE OSTI2018

On efficient Hessian computation using the edge pushing algorithm in Julia

Evaluating the Hessian matrix of second-order derivatives at a sequence of points can be costly when applying second-order methods for nonlinear optimization. In this work, we discuss our experiences implementing the recently proposed Edge Pushing (EP) method in Julia as an experimental replacement for the current colouring-based methods used by JuMP, an open-source algebraic modelling language. We propose an alternative data structure for sparse Hessians to avoid the use of hash tables and analyse the space and time complexity of EP method. In our benchmarks, we find that EP is very competitive in terms of both preprocessing time and Hessian evaluation time. As a result, we identify cases where EP closes the performance gap between JuMP's previous implementation and the implementation in AMPL, a commercial software package with similar functionality.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

An efficient tensor transpose algorithm for multicore CPU, Intel Xeon Phi, and NVidia Tesla GPU

An efficient parallel tensor transpose algorithm is suggested for shared-memory computing units, namely, multicore CPU, Intel Xeon Phi, and NVidia GPU. The algorithm operates on dense tensors (multidimensional arrays) and is based on the optimization of cache utilization on x86 CPU and the use of shared memory on NVidia GPU. From the applied side, the ultimate goal is to minimize the overhead encountered in the transformation of tensor contractions into matrix multiplications in computer implementations of advanced methods of quantum many-body theory (e.g., in electronic structure theory and nuclear physics). A particular accent is made on higher-dimensional tensors that typically appear in the so-called multireference correlated methods of electronic structure theory. Depending on tensor dimensionality, the presented optimized algorithms can achieve an order of magnitude speedup on x86 CPUs and 2-3 times speedup on NVidia Tesla K20X GPU with respect to the na ve scattering algorithm (no memory access optimization). Furthermore, the tensor transpose routines developed in this work have been incorporated into a general-purpose tensor algebra library (TAL-SH).

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Multistate hybrid time-dependent density functional theory with surface hopping accurately captures ultrafast thymine photodeactivation

We report an efficient analytical implementation of first-order nonadiabatic derivative couplings between arbitrary Born–Oppenheimer states in the hybrid time-dependent density functional theory (TDDFT) framework using atom-centered basis functions. Our scheme is based on quadratic response theory and includes orbital relaxation terms neglected in previous approaches. Simultaneous computation of multiple derivative couplings and energy gradients enables efficient multistate nonadiabatic molecular dynamics simulations in conjunction with Tully's fewest switches surface hopping (SH) method. We benchmark the thus obtained multistate TDDFT-SH scheme by simulating ultrafast decay of UV-photoexcited thymine, for which accurate gas-phase data from ultrafast spectroscopy experiments are available. The calculations predict a fast 153 fs decay from the bright S 2 to the dark S 1 excited state, followed by a much slower 14 ps S 1 deactivation to the ground state; statistical uncertainties were estimated using bootstrap sampling. These results agree well with the experimentally observed time constants of 100–200 fs and 5–7 ps, respectively, unlike previous multiconfigurational self-consistent field and second-order algebraic diagrammatic construction calculations. Furthermore, our results support the S 1 -trapping hypothesis. For thymine, the computational cost of a single TDDFT-SH time-step including the lowest 3 states, all couplings and gradients, is ~5 times larger than the cost of a single Born–Oppenheimer dynamics time step for the ground state in our implementation. Thus, ps nonadiabatic dynamics simulations using multistate hybrid TDDFT-SH for systems with up to ~100 atoms are possible without drastic approximations on single workstation nodes. Our implementation will be made available through Turbomole.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

AXI4MLIR: User-Driven Automatic Host Code Generation for Custom AXI-Based Accelerators

Tensor algebra operations represent an important class of algorithms used across many applications, including machine learning, scientific computing, and data analytics. As a result, the efficient generation of custom accelerators for tensor operations has received increased attention. Previous efforts have produced automated tools enabling users to prototype and explore optimized accelerators. However, little effort has been focused on the host-accelerator interaction in these tools. Efficient use of hardware accelerators requires knowledge about the accelerator's capabilities (operations, data formats, and opcode support), the host CPU microarchitecture (e.g., memory hierarchy), the host-accelerator interface, and the application's features (which code regions should be mapped onto an accelerator). Manually rewriting the original applications to facilitate improved custom accelerator mapping is an error-prone and time-consuming endeavor. To cope with this, we propose AXI4MLIR, a new framework to automatically generate and optimize the communication between the host CPU and arbitrary accelerators that implement linear algebra algorithms. AXI4MLIR extends the MLIR compiler framework to automatically generate efficient host-accelerator driver code for accelerators with AXI-based interfaces. Our compiler extensions enable automatic driver code generation while carefully considering the host's memory hierarchy and target accelerator features. To demonstrate the flexibility and utility of AXI4MLIR, we test it with diverse use cases that include different types of accelerators, tiling scenarios, and dataflow schemes. We compare our experimental results to manual implementations of host-accelerator driver code and find that our approach can reduce CPU cache references by 56% and deliver up to a 1.65x speedup.

Bohm Agostini, Nicolas↗
DOE OSTI2019

A configurational force for adaptive re-meshing of gradient-enhanced poromechanics problems with history-dependent variables

Here, we introduce a mesh-adaption framework that employs a multi-physical configurational force and Lie algebra to capture multiphysical responses of fluid-infiltrating geological materials while maintaining the efficiency of the computational models. To resolve sharp gradients of both displacement and pore pressure, we introduce an energy-estimate-free re-meshing criterion by extending the configurational force theory to consider the energy dissipation due to the fluid diffusion and the gradient-dependent plastic flow. To establish new equilibria after remeshing, the local tensorial history-dependent variables at the integration points are first decomposed into spectral forms. Then, the principal values and directions are projected onto smooth fields interpolated by the basis function of the finite element space via the Lie-algebra mapping. Our numerical results indicate that this Lie algebra operator in general leads to a new trial state closer to the equilibrium than the ones obtained from the tensor component mapping approach. A new configurational force for dissipative fluid-infiltrating porous materials that exhibit gradient-dependent plastic flow is introduced such that the remeshing may accommodate the need to resolve the sharp pressure gradient as well as the strain localization. The predicted responses are found to be not influenced by the mesh size due to the micromorphic regularization, while the adaptive meshing enables us to capture the width of deformation bands without the necessity of employing fine mesh everywhere in the domain.

42 ENGINEERING↗
DOE OSTI2021

CSRI Summer Proceedings 2021

The Computer Science Research Institute (CSRI) brings university faculty and students to Sandia National Laboratories for focused collaborative research on Department of Energy (DOE) computer and computational science problems. The institute provides an opportunity for university researches to learn about problems in computer and computational science at DOE laboratories, and help transfer results of their research to programs at the labs. Some specific CSRI research interest areas are: scalable solvers, optimization, algebraic preconditioners, graph-based, discrete, and combinatorial algorithms, uncertainty estimation, validation and verification methods, mesh generation, dynamic load-balancing, virus and other malicious-code defense, visualization, scalable cluster computers, beyond Moore’s Law computing, exascale computing tools and application design, reduced order and multiscale modeling, parallel input/output, and theoretical computer science. The CSRI Summer Program is organized by CSRI and includes a weekly seminar series and the publication of a summer proceedings.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

CSRI Summer Proceedings 2021

The Computer Science Research Institute (CSRI) brings university faculty and students to Sandia National Laboratories for focused collaborative research on Department of Energy (DOE) computer and computational science problems. The institute provides an opportunity for university researches to learn about problems in computer and computational science at DOE laboratories, and help transfer results of their research to programs at the labs. Some specific CSRI research interest areas are: scalable solvers, optimization, algebraic preconditioners, graph-based, discrete, and combinatorial algorithms, uncertainty estimation, validation and verification methods, mesh generation, dynamic load-balancing, virus and other malicious-code defense, visualization, scalable cluster computers, beyond Moore’s Law computing, exascale computing tools and application design, reduced order and multiscale modeling, parallel input/output, and theoretical computer science. The CSRI Summer Program is organized by CSRI and includes a weekly seminar series and the publication of a summer proceedings.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

A Scalable Interior‐Point Gauss–Newton Method for PDE‐Constrained Optimization With Bound Constraints

Here, we present a scalable approach to solve a class of partial differential equation (PDE)‐constrained optimization problems with bound constraints. This approach utilizes a robust full‐space interior‐point (IP)‐Gauss–Newton optimization method. To cope with the poorly‐conditioned IP‐Gauss–Newton saddle‐point linear systems that need to be solved approximately, once per optimization step, we propose two spectrally related preconditioners. These preconditioners leverage the limited informativeness of data in regularized PDE‐constrained optimization problems. A block Gauss–Seidel preconditioner is proposed for the GMRES‐based solution of the IP‐Gauss–Newton linear systems. It is shown, for a large‐class of PDE‐ and bound‐constrained optimization problems, that the spectrum of the block Gauss–Seidel preconditioned IP‐Gauss–Newton matrix is asymptotically independent of discretization and is not impacted by the ill‐conditioning that notoriously plagues interior‐point methods. We exploit symmetry of the IP‐Gauss–Newton linear systems and propose a regularization and log‐barrier Hessian preconditioner for the preconditioned conjugate gradient (PCG)‐based solution of the equivalent IP‐Gauss–Newton–Schur complement linear systems. The eigenvalues of the block Gauss–Seidel preconditioned IP‐Gauss–Newton matrix, that are not equal to one, are identical to the eigenvalues of the regularization and log‐barrier Hessian preconditioned Schur complement matrix. The scalability of the approach is demonstrated on two example problems. The numerical solution of these optimization problems is shown to require a discretization independent number of IP‐Gauss–Newton linear solves. Furthermore, the linear systems are solved in a discretization and IP ill‐conditioning independent number of preconditioned Krylov subspace iterations. The parallel scalability of the preconditioner, achieved via algebraic multigrid component solvers when applicable, and the aforementioned algorithmic scalability permits a parallel scalable means to compute solutions of a large class of PDE‐ and bound‐constrained problems.

PDE-constrained optimization↗
DOE OSTI2022

Local Noether theorem for quantum lattice systems and topological invariants of gapped states

Here, we study generalizations of the Berry phase for quantum lattice systems in arbitrary dimensions. For a smooth family of gapped ground states in d dimensions, we define a closed d + 2-form on the parameter space, which generalizes the curvature of the Berry connection. Its cohomology class is a topological invariant of the family. When the family is equivariant under the action of a compact Lie group G, topological invariants take values in the equivariant cohomology of the parameter space. These invariants unify and generalize the Hall conductance and the Thouless pump. A key role in these constructions is played by a certain differential graded Fréchet–Lie algebra attached to any quantum lattice system. As a by-product, we describe ambiguities in charge densities and conserved currents for arbitrary lattice systems with rapidly decaying interactions.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Unitary circuit synthesis for tomography of generalized coherent states

We present a method that outputs a sequence of simple unitary operations to prepare a given quantum state that is a generalized coherent state. Our method takes as inputs the expectation values of some relevant observables. Such expectations can be estimated by performing projective measurements on $\mathcal{O}$(M 3 log(M/δ)/ε 2 ) copies of the state, where M is the dimension of an associated Lie algebra, ε is a precision parameter, and 1 - δ is the required confidence level. The method can be implemented on a classical computer and runs in time $\mathcal{O}$(M 4 log(M/ε)). It provides $\mathcal{O}$(Mlog(M/ε)) simple unitaries that form the sequence. The overall complexity is then polynomial in M, being very efficient in cases where M is significantly smaller than the Hilbert space dimension, as for some fermion algebras. When the algebra of relevant observables is given by certain Pauli matrices, each simple unitary may be easily decomposed into two-qubit gates. We discuss applications to efficient quantum state tomography and classical simulations of quantum circuits. $\mathcal{O}$

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2013

Annual Report: Carbon Capture Simulation Initiative (CCSI) (30 September 2013)

The Carbon Capture Simulation Initiative (CCSI) is a partnership among national laboratories, industry and academic institutions that is developing and deploying state-of-the-art computational modeling and simulation tools to accelerate the commercialization of carbon capture technologies from discovery to development, demonstration, and ultimately the widespread deployment to hundreds of power plants. The CCSI Toolset will provide end users in industry with a comprehensive, integrated suite of scientifically validated models, with uncertainty quantification (UQ), optimization, risk analysis and decision making capabilities. The CCSI Toolset incorporates commercial and open-source software currently in use by industry and is also developing new software tools as necessary to fill technology gaps identified during execution of the project. Ultimately, the CCSI Toolset will (1) enable promising concepts to be more quickly identified through rapid computational screening of devices and processes; (2) reduce the time to design and troubleshoot new devices and processes; (3) quantify the technical risk in taking technology from laboratory-scale to commercial-scale; and (4) stabilize deployment costs more quickly by replacing some of the physical operational tests with virtual power plant simulations. CCSI is led by the National Energy Technology Laboratory (NETL) and leverages the Department of Energy (DOE) national laboratories’ core strengths in modeling and simulation, bringing together the best capabilities at NETL, Los Alamos National Laboratory (LANL), Lawrence Berkeley National Laboratory (LBNL), Lawrence Livermore National Laboratory (LLNL), and Pacific Northwest National Laboratory (PNNL). The CCSI’s industrial partners provide representation from the power generation industry, equipment manufacturers, technology providers and engineering and construction firms. The CCSI’s academic participants (Carnegie Mellon University, Princeton University, West Virginia University, Boston University and the University of Texas at Austin) bring unparalleled expertise in multiphase flow reactors, combustion, process synthesis and optimization, planning and scheduling, and process control techniques for energy processes. During Fiscal Year (FY) 13, CCSI announced the initial release of its first set of computational tools and models during the October 2012 meeting of its Industry Advisory Board. This initial release led to five companies licensing the CCSI Toolset under a Test and Evaluation Agreement this year. By the end of FY13, the CCSI Technical Team had completed development of an updated suite of computational tools and models. The list below summarizes the new and enhanced toolset components that were released following comprehensive testing during October 2013. 1. FOQUS. Framework for Optimization and Quantification of Uncertainty and Sensitivity. Package includes: FOQUS Graphic User Interface (GUI), simulation-based optimization engine, Turbine Client, and heat integration capabilities. There is also an updated simulation interface and new configuration GUI for connecting Aspen Plus or Aspen Custom Modeler (ACM) simulations to FOQUS and the Turbine Science Gateway. 2. A new MFIX-based Computational Fluid Dynamics (CFD) model to predict particle attrition. 3. A new dynamic reduced model (RM) builder, which generates computationally efficient RMs of the behavior of a dynamic system. 4. A completely re-written version of the algebraic surrogate model builder for optimization (ALAMO). The new version is several orders of magnitude faster than the initial release and eliminates the MATLAB dependency. 5. A new suite of high resolution filtered models for the hydrodynamics associated with horizontal cylindrical objects in a flow path. 6. The new Turbine Science Gateway (Cluster), which supports FOQUS for running multiple simulations for optimization or UQ using a local computer or cluster. 7. A new statistical tool (BSS-ANOVA-UQ) for calibration and validation of CFD models. 8. A new basic data submodel in Aspen Plus format for a representative high viscosity capture solvent, 2-MPZ system. 9. An updated RM tool for CFD (REVEAL) that can create a RM from MFIX. A new lightweight, stand-alone version will be available in late 2013. 10. An updated RM integration tool to convert the RM from REVEAL into a CAPE-OPEN or ACM model for use in a process simulator. 11. An updated suite of unified steady-state and dynamic process models for solid sorbent carbon capture included bubbling fluidized bed and moving bed reactors. 12. An updated and unified set of compressor models including steady-state design point model and dynamic model with surge detection. 13. A new framework for the synthesis and optimization of coal oxycombustion power plants using advanced optimization algorithms. This release focuses on modeling and optimization of a cryogenic air separation unit (ASU). 14. A new technical risk model in spreadsheet format. 15. An updated version of the sorbent kinetic/equilibrium model for parameter estimation for the 1st generation sorbent model. 16. An updated process synthesis superstructure model to determine optimal process configurations utilizing surrogate models from ALAMO for adsorption and regeneration in a solid sorbent process. 17. Validation models for NETL Carbon Capture Unit utilizing sorbent AX. Additional validation models will be available for sorbent 32D in 2014. 18. An updated hollow fiber membrane model and system example for carbon capture. 19. An updated reference power plant model in Thermoflex that includes additional steam extraction and reinjection points to enable heat integration module. 20. An updated financial risk model in spreadsheet format.

01 COAL, LIGNITE, AND PEAT↗