SEARCH · Search NASA
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.
Some of the parallel computational strategies are reviewed, and their adequacy assessed for a given architecture and a given problem. The performance of both extreme parallel hardware technologies on real-life structural problems is also reported.
The principal objective of this research is to develop, test, and implement coarse-grained, parallel-processing strategies for nonlinear dynamic simulations of practical structural problems. There are contributions to four main areas: finite element modeling and analysis of rotational dynamics, numerical algorithms for parallel nonlinear solutions, automatic partitioning techniques to effect load-balancing among processors, and an integrated parallel analysis system.
SPADES (Solver for PArallel Discrete Event Simulation) is an open-source parallel discrete event simulation (PDES) package built on the AMReX library. Targeted at solving discrete event systems in parallel, this software package aims to be performance portable and scalable on heterogeneous computing architectures, e.g., graphic processing units (GPU). SPADES implements optimistic synchronization with rollback through an implementation of the Time Warp algorithm. An alternative conservative synchronization approach is also implemented using the Lower Bound on Incoming Time Stamp. In our implementation, logical processes are represented as cells in a grid and event messages are represented as particles. SPADES supports various parallel decomposition strategies, including the use of the Message Passing Interface (MPI) and OpenMP threading. All major GPU architectures (e.g., Intel, AMD, NVIDIA) are supported through the use of performance portability functionalities implemented in AMReX. The SPADES software is released in NREL Software Record SWR-24-99 “SPADES (Scalable Parallel Discrete Events Simulation)”.
An analysis is presented of several factors influencing the performance of a parallel implementation of the UCLA atmospheric general circulation model (AGCM) on massively parallel computer systems. Several modificaitons to the original parallel AGCM code aimed at improving its numerical efficiency, interprocessor communication cost, load-balance and issues affecting single-node code performance are discussed.
This two-part paper presents the results of a benchmarked analytical-numerical investigation into the operational characteristics of a unified parallel processing strategy for implicit fluid mechanics formulations. This hierarchical poly tree (HPT) strategy is based on multilevel substructural decomposition. The Tree morphology is chosen to minimize memory, communications and computational effort. The methodology is general enough to apply to existing finite difference (FD), finite element (FEM), finite volume (FV) or spectral element (SE) based computer programs without an extensive rewrite of code. In addition to finding large reductions in memory, communications, and computational effort associated with a parallel computing environment, substantial reductions are generated in the sequential mode of application. Such improvements grow with increasing problem size. Along with a theoretical development of general 2-D and 3-D HPT, several techniques for expanding the problem size that the current generation of computers are capable of solving, are presented and discussed. Among these techniques are several interpolative reduction methods. It was found that by combining several of these techniques that a relatively small interpolative reduction resulted in substantial performance gains. Several other unique features/benefits are discussed in this paper. Along with Part 1's theoretical development, Part 2 presents a numerical approach to the HPT along with four prototype CFD applications. These demonstrate the potential of the HPT strategy.
An investigation is conducted to find good parallel algorithms for solving systems of nonlinear equations using probability-one homotopy methods. Particular attention is paid to algorithms for the hypercube. Methods for one of the most computationally expensive steps of the homotopy approach, the computation of the kernel of the Jacobian matrix of the homotopy map, are studied. General nonlinear systems of equations with small and dense Jacobian matrices are considered, however, polynomial systems are not, since their structure leads to different strategies for parallelism. The mathematics behind the homotopy algorithm is summarized and the use of orthogonal factorizations is discussed. Parallel algorithms for orthogonal factorizations and triangular system solving are described. Computational results are presented and discussed.
Pipelining is a common strategy for extracting parallelism from a collection of independent computational tasks, each of which is spread among a number of processors and has an implied data dependence. When implemented on MIMD parallel computers with finite process interrupt times, pipeline algorithms suffer from slowdown--in addition to the expected pipeline fill time--due to a wave-like propagation of delays. This phenomenon, which has been observed experimentally using the performance monitoring system AIMS, is investigated analytically, and an optimal correction is derived to eliminate the wave. Efficiency increase through the correction is verified experimentally.
A geometric multigrid algorithm is introduced for solving nonsymmetric linear systems resulting from the discretization of the variable density Navier–Stokes equations on nonuniform structured rectilinear grids and high-Reynolds number flows. The restriction operation is defined such that the resulting system on the coarser grids is symmetric, thereby allowing for the use of efficient smoother algorithms. To achieve an optimal rate of convergence, the sequence of interpolation and restriction operations are determined through a dynamic procedure. A parallel partitioning strategy is introduced to minimize communication while maintaining the load balance between all processors. To test the proposed algorithm, we consider two cases: 1) homogeneous isotropic turbulence discretized on uniform grids and 2) turbulent duct flow discretized on stretched grids. Testing the algorithm on systems with up to a billion unknowns shows that the cost varies linearly with the number of unknowns. This O (N) behavior confirms the robustness of the proposed multigrid method regarding ill-conditioning of large systems characteristic of multiscale high-Reynolds number turbulent flows. The robustness of our method to density variations is established by considering cases where density varies sharply in space by a factor of up to 104, showing its applicability to two-phase flow problems. Strong and weak scalability studies are carried out, employing up to 30,000 processors, to examine the parallel performance of our implementation. Excellent scalability of our solver is shown for a granularity as low as 104 to 105 unknowns per processor. At its tested peak throughput, it solves approximately 4 billion unknowns per second employing over 16,000 processors with a parallel efficiency higher than 50%.
In High Energy Physics (HEP), experimentalists generate large volumes of data that, when analyzed, helps us better understand the fundamental particles and their interactions. This data is often captured in many files of small size, creating a data management challenge for scientists. In order to better facilitate data management, transfer, and analysis on large scale platforms, it is advantageous to aggregate data further into a smaller number of larger files. However, this translation process can consume significant time and resources, and if performed incorrectly the resulting aggregated files can be inefficient for highly parallel access during analysis on large scale platforms. In this paper, we present our case study on parallel I/O strategies and HDF5 features for reducing data aggregation time, making effective use of compression, and ensuring efficient access to the resulting data during analysis at scale. We focus on NOvA detector data in this case study, a large-scale HEP experiment generating many terabytes of data. Here, the lessons learned from our case study inform the handling of similar datasets, thus expanding community knowledge related to this common data management task.
Practical solution of stochastic programming problems generally requires the use of parallel computing resources. Here, we describe the open source package mpi-sppy, in which efficient and scalable parallelization is a central feature. We report computational experiments that demonstrate the ability to solve very large stochastic programming problems - including mixed-integer variants - in minutes of wall clock time, efficiently leveraging significant parallel computing resources. We report results for the largest publicly available instances of stochastic mixed-integer unit commitment problems, solving to provably tight optimality gaps. In addition, we introduce a novel software architecture that facilitates combinations of methods for accelerating convergence that can be combined in plug-and-play manner. Finally, the mpi-sppy package is written in Python, leverages the widely used Pyomo (http://www.pyomo.org) library for modeling mathematical programs, builds on existing MPI implementations to ensure efficiency and scalability, and is available via http://github.com/Pyomo/mpi-sppy.
Multicomponent polymer mixtures are ubiquitous in biological self-organization but are notoriously difficult to study computationally. Plagued by both slow single molecule relaxation times and slow equilibration within dense mixtures, molecular dynamics simulations are typically infeasible at the spatial scales required to study the stability of mesophase structure. Polymer field theories offer an attractive alternative, but analytical calculations are only tractable for mean-field theories and nearby perturbations, constraints that become especially problematic for fluctuation-induced effects such as coacervation. Here, we show that a recently developed technique for obtaining numerical solutions to partial differential equations based on operator learning, neural operators, lends itself to a highly scalable training strategy by parallelizing per-species operator maps. We illustrate the efficacy of our approach on six-component mixtures with randomly selected compositions and that it significantly outperforms the state-of-the-art pseudospectral integrators for field-theoretic simulations, especially as polymer lengths become long.
Large-scale simulations on nonuniform particle distributions that evolve over time are widely used in cosmology, molecular dynamics, and engineering. Such data are often saved in an unstructured format that neither preserves spatial locality nor provides metadata for accelerating spatial or attribute subset queries, leading to poor performance of visualization tasks. Furthermore, the parallel I/O strategy used typically writes a file per process or a single shared file, neither of which is portable or scalable across different HPC systems. We present a portable technique for scalable, spatially aware adaptive aggregation that preserves spatial locality in the output. We evaluate our approach on two supercomputers, Stampede2 and Summit, and demonstrate that it outperforms prior approaches at scale, achieving up to 2.5× faster writes and reads for nonuniform distributions. Furthermore, the layout written by our method is directly suitable for visual analytics, supporting low-latency reads and attribute-based filtering with little overhead.
Purpose: Degradation of image quality caused by cardiac and respiratory motions hampers the diagnostic quality of cardiac PET. It has been shown that improved diagnostic accuracy of myocardial defect can be achieved by tagged MR (tMR) based PET motion correction using simultaneous PET-MR. However, one major hurdle for the adoption of tMR-based PET motion correction in the PET-MR routine is the long acquisition time needed for the collection of fully sampled tMR data. In this work, the authors propose an accelerated tMR acquisition strategy using parallel imaging and/or compressed sensing and assess the impact on the tMR-based motion corrected PET using phantom and patient data. Methods: Fully sampled tMR data were acquired simultaneously with PET list-mode data on two simultaneous PET-MR scanners for a cardiac phantom and a patient. Parallel imaging and compressed sensing were retrospectively performed by GRAPPA and kt-FOCUSS algorithms with various acceleration factors. Motion fields were estimated using nonrigid B-spline image registration from both the accelerated and fully sampled tMR images. The motion fields were incorporated into a motion corrected ordered subset expectation maximization reconstruction algorithm with motion-dependent attenuation correction. Results: Although tMR acceleration introduced image artifacts into the tMR images for both phantom and patient data, motion corrected PET images yielded similar image quality as those obtained using the fully sampled tMR images for low to moderate acceleration factors (<4). Quantitative analysis of myocardial defect contrast over ten independent noise realizations showed similar results. It was further observed that although the image quality of the motion corrected PET images deteriorates for high acceleration factors, the images were still superior to the images reconstructed without motion correction. Conclusions: Accelerated tMR images obtained with more than 4 times acceleration can still provide relatively accurate motion fields and yield tMR-based motion corrected PET images with similar image quality as those reconstructed using fully sampled tMR data. The reduction of tMR acquisition time makes it more compatible with routine clinical cardiac PET-MR studies.
X-ray scattering is a valuable tool for measuring the structural properties of materialsused in the design and fabrication of energy-relevant nanodevices (e.g., photovoltaic, energy storage, battery, fuel, and carbon capture andsequestration devices) that are key to the reduction of carbon emissions. Although today's ultra-fast X-ray scattering detectors can provide tremendousinformation on the structural properties of materials, a primary challenge remains in the analyses of the resulting data. We are developing novelhigh-performance computing algorithms, codes, and software tools for the analyses of X-ray scattering data. In this paper we describe two such HPCalgorithm advances. Firstly, we have implemented a flexible and highly efficient Grazing Incidence Small Angle Scattering (GISAXS) simulation code based on theDistorted Wave Born Approximation (DWBA) theory with C++/CUDA/MPI on a cluster of GPUs. Our code can compute the scattered light intensity from any givensample in all directions of space; thus allowing full construction of the GISAXS pattern. Preliminary tests on a single GPU show speedups over 125x compared tothe sequential code, and almost linear speedup when executing across a GPU cluster with 42 nodes, resulting in an additional 40x speedup compared to usingone GPU node. Secondly, for the structural fitting problems in inverse modeling, we have implemented a Reverse Monte Carlo simulation algorithm with C++/CUDAusing one GPU. Since there are large numbers of parameters for fitting in the in X-ray scattering simulation model, the earlier single CPU code required weeks ofruntime. Deploying the AccelerEyes Jacket/Matlab wrapper to use GPU gave around 100x speedup over the pure CPU code. Our further C++/CUDA optimization deliveredan additional 9x speedup.
General technique uses a modified version of an existing technique termed the pattern search technique. New procedure called the parallel move strategy permits pattern search technique to be used with problems involving a constraint.
Within the Center for Space Construction, the SIMSTRUC project's objectives center around the development of simulation tools for the realistic analysis of large space structures. The word 'tools' is the broad sense; it designates mathematical models, finite element/finite difference formulations, computational algorithms, implementations on advanced computer architectures, and visualization capabilities. The results of our activities during the first year within the SIMSTRUC project are reported. On the modeling side, an alternative approach to fluid/thermal/structure interaction analysis that is a departure from the 'loosely coupled' and 'unified' approaches that are being currently practiced are described. The advantages of our approach both in terms of accuracy and computational efficiency were demonstrated. On the computational side, a software architecture for parallel/vector and massively parallel supercomputers that speeds up finite element and finite difference computations by several orders of magnitude is presented. As an example, the simulation of the deployment of a space structure that used to require over six hours of a workstation using a conventional finite element software, now runs on a multiprocessor using a parallel computation strategy in less than three seconds. In order to promote the physical understanding of the simulation behavior, a real-time visualization capability on the Connection Machine, which allows the analyst to watch the graphical animation of the results at the same time these are generated, was also developed. It is believed that by combining efficient analytical formulations with the state-of-the-art high performance computer implementations and superfast visualization capabilities, SIMSTRUC is moving fast towards the real-time simulation of large space structures. The designers as well as the researchers will certainly benefit from this technology.
This dissertation addresses various issues for model classes of hyperbolic conservation laws. The basic approach developed in this work employs a new family of adaptive, hp-version, finite element methods based on a special discontinuous Galerkin formulation for hyperbolic problems. The discontinuous Galerkin formulation admits high-order local approximations on domains of quite general geometry, while providing a natural framework for finite element approximations and for theoretical developments. The use of hp-versions of the finite element method makes possible exponentially convergent schemes with very high accuracies in certain cases; the use of adaptive hp-schemes allows h-refinement in regions of low regularity and p-enrichment to deliver high accuracy, while keeping problem sizes manageable and dramatically smaller than many conventional approaches. The use of discontinuous Galerkin methods is uncommon in applications, but the methods rest on a reasonable mathematical basis for low-order cases and has local approximation features that can be exploited to produce very efficient schemes, especially in a parallel, multiprocessor environment. The place of this work is to first and primarily focus on a model class of linear hyperbolic conservation laws for which concrete mathematical results, methodologies, error estimates, convergence criteria, and parallel adaptive strategies can be developed, and to then briefly explore some extensions to more general cases. Next, we provide preliminaries to the study and a review of some aspects of the theory of hyperbolic conservation laws. We also provide a review of relevant literature on this subject and on the numerical analysis of these types of problems.
A parallel multilevel strategy is developed using spectral (p) finite elements. Hierarchic bases are particularly well suited since the element matrices and vectors are nested and the multilevel projections easiliy performed. Since the basis degree is used to specify the multigrid level, an EBE strategy is natural br the multilevel technique. Results are presented for two candidate nonlinear elliptic transport problems: the augmented drift-diffusion equations of semiconductor device modeling and the stream function-vorticity equations of incompressible fluid dynamics.