Search NASA⌕ Search

SEARCH · Search NASA

Results for “parallel computing algorithms”

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 1,783 records · Page 99

mystic - a brief introduction

The mystic framework enables the solving of difficult optimization problems through three major innovations: (1) plug-and-play component-based optimization algorithms, (2) the ability to dynamically construct and apply generalzed trnasforms, K , where x' = K(x) , and (3) workflows providing asynchronous execution and statefulness. These features, in conjunction with supporting packages pathos , pyina , and klepto , can provide massively-parallel scalable workflows for quickly solving optimization problems in complex nonlinear spaces. mystic can be used to simplify non-convex optimization problems by transforming away nonlinearities through user-built kernel transforms. This tutorial demonstrates using mystic for machine learning with automated dimensional-reduction, using embarrasingly parallel solver ensembles to find an accurate interpolated surrogate for a nonlinear surface, and the determination of worst-case bounds on expectation value of an objective function under uncertainty.

97 MATHEMATICS AND COMPUTING↗

Energy-efficient mapping of large-scale workflows under deadline constraints in big data computing systems

Large-scale workflows for big data analytics have become a main consumer of energy in data centers where moldable parallel computing models such as MapReduce are widely applied to meet high computational demands with time-varying computing resources. The granularity of task partitioning in each moldable job of such big data workflows has a significant impact on energy efficiency, which remains largely unexplored. In this paper, we analyze the properties of moldable jobs and formulate a workflow mapping problem to minimize the dynamic energy consumption of a given workflow request under a deadline constraint in big data systems. Since this problem is strongly NP-hard, we design a fully polynomial-time approximation scheme (FPTAS) for a special case with a pipeline-structured workflow on a homogeneous cluster and a heuristic for the generalized problem with an arbitrary workflow on a heterogeneous cluster. The performance superiority of the proposed solution in terms of dynamic energy saving and deadline missing rate is illustrated by extensive simulation results in comparison with existing algorithms, and further validated by real-life workflow implementation and experimental results in Hadoop/YARN systems.

97 MATHEMATICS AND COMPUTING↗

3-regular three-XORSAT planted solutions benchmark of classical and quantum heuristic optimizers

With current semiconductor technology reaching its physical limits, special-purpose hardware has emerged as an option to tackle specific computing-intensive challenges. Optimization in the form of solving quadratic unconstrained binary optimization problems, or equivalently Ising spin glasses, has been the focus of several new dedicated hardware platforms. These platforms come in many different flavors, from highly-efficient hardware implementations on digital-logic of established algorithms to proposals of analog hardware implementing new algorithms. In this work, we use a mapping of a specific class of linear equations whose solutions can be found efficiently, to a hard constraint satisfaction problem (three-regular three-XORSAT, or an Ising spin glass) with a 'golf-course' shaped energy landscape, to benchmark several of these different approaches. We perform a scaling and prefactor analysis of the performance of Fujitsu's digital annealer unit (DAU), the D-Wave advantage quantum annealer, a virtual MemComputing machine, Toshiba's simulated bifurcation machine (SBM), the SATonGPU algorithm from Bernaschi et al, and our implementation of parallel tempering. We identify the SATonGPU and DAU as currently having the smallest scaling exponent for this benchmark, with SATonGPU having a small scaling advantage and in addition having by far the smallest prefactor thanks to its use of massive parallelism. Furthermore, our work provides an objective assessment and a snapshot of the promise and limitations of dedicated optimization hardware relative to a particular class of optimization problems.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

PMU-Based Decoupled State Estimation for Unsymmetrical Power Systems

Modal decomposition of measurement equations has already been shown to simplify the formulation and resulting computational complexity of three-phase state estimation of systems where all the transmission lines are three-phase and fully transposed. When there are non-transposed and/or mixed-phase lines, modal decomposition can no longer fully decouple the threephase measurement equations. Here, this paper addresses the above shortcoming by proposing a simple yet practical solution based on the commonly used numerical compensation techniques. Thus, it enables application of the powerful decoupling approach to any type of three-phase networks which may contain non-transposed or mixed-phase lines and are fully observable by PMUs. The proposed procedure modifies the measurement set by deriving additive terms that compensate for the neglected unsymmetrical effects. It will be shown that unbalanced systems including nontransposed and mixed-phase elements, can still be transformed into three decoupled subsystems and solved in parallel by the proposed approach. Performance of the proposed algorithm is validated against several IEEE test cases.

42 ENGINEERING↗

A Work-Efficient Parallel Sparse Matrix-Sparse Vector Multiplication Algorithm

We design and develop a work-efficient multithreaded algorithm for sparse matrix-sparse vector multiplication (SpMSpV) where the matrix, the input vector, and the output vector are all sparse. SpMSpV is an important primitive in the emerging GraphBLAS standard and is the workhorse of many graph algorithms including breadth-first search, bipartite graph matching, and maximal independent set. As thread counts increase, existing multithreaded SpMSpV algorithms can spend more time accessing the sparse matrix data structure than doing arithmetic. Our shared-memory parallel SpMSpV algorithm is work efficient in the sense that its total work is proportional to the number of arithmetic operations required. The key insight is to avoid each thread individually scan the list of matrix columns. Our algorithm is simple to implement and operates on existing column-based sparse matrix formats. It performs well on diverse matrices and vectors with heterogeneous sparsity patterns. A high-performance implementation of the algorithm attains up to 15x speedup on a 24-core Intel Ivy Bridge processor and up to 49x speedup on a 64-core Intel KNL manycore processor. In contrast to implementations of existing algorithms, the performance of our algorithm is sustained on a variety of different input types include matrices representing scale-free and high-diameter graphs.

97 MATHEMATICS AND COMPUTING↗

Elastic Solutions to 2D Plane Strain Problems: Nonlinear Contact and Settlement Analysis for Shallow Foundations

The classical Neumann boundary value problem of an isotropic, homogeneous elastic half-plane under plane strain conditions is readdressed as the limiting case of the fully three-dimensional problem. Analytical solutions of the stress and strain tensors are obtained by taking the limit from known three-dimensional solutions. It is shown that the displacement fields for the plane strain problem are not well defined. A small number of simple expressions are developed, which provide a general solution for linearly-varying traction over arbitrary regions on the boundary. A simple, efficient, and rapidly convergent algorithm is developed which uses these solutions as analytic elements and provides a solution approach to the general boundary value problem. The method is verified against known solutions for Hertzian contact between parallel cylinders. Two numerical examples are presented for the analysis of shallow foundation systems. In the first, the boundary conditions are informed by analytical elastoplastic calculations and a strain influence analysis is performed and compared with the Schmertmann method. Subsequently, empirical laboratory contact traction distributions measured by Bauer et al., in both the normal and tangential directions are employed as boundary conditions for an analysis of the underlying stress field.

42 ENGINEERING↗

A GENERALIZED STOCHASTIC COLLOCATION APPROACH TO CONSTRAINED OPTIMIZATION FOR RANDOM DATA IDENTIFICATION PROBLEMS

We present a scalable, parallel mechanism for stochastic identification/control for problems constrained by partial differential equations with random input data. Several identification objectives will be discussed that either minimize the expectation of a tracking cost functional or minimize the difference of desired statistical quantities in the appropriate $L^p$ norm, and the distributed parameters/control can both deterministic or stochastic. Given an objective we prove the existence of an optimal solution, establish the validity of the Lagrange multiplier rule and obtain a stochastic optimality system of equations. The modeling process may describe the solution in terms of high dimensional spaces, particularly in the case when the input data (coefficients, forcing terms, boundary conditions, geometry, etc) are affected by a large amount of uncertainty. For higher accuracy, the computer simulation must increase the number of random variables (dimensions), and expend more effort approximating the quantity of interest in each individual dimension. Hence, we introduce a novel stochastic parameter identification algorithm that integrates an adjoint-based deterministic algorithm with the sparse grid stochastic collocation FEM approach. This allows for decoupled, moderately high dimensional, parameterized computations of the stochastic optimality system, where at each collocation point, deterministic analysis and techniques can be utilized. The advantage of our approach is that it allows for the optimal identification of statistical moments (mean value, variance, covariance, etc.) or even the whole probability distribution of the input random fields, given the probability distribution of some responses of the system (quantities of physical interest). Our rigorously derived error estimates, for the fully discrete problems, will be described and used to compare the efficiency of the method with several other techniques. Numerical examples illustrate the theoretical results and demonstrate the distinctions between the various stochastic identification objectives.

Webster, Clayton G↗

Quantifying the debonding of inclusions through tomography and computational homology.

This report describes a Laboratory Directed Research and Development (LDRD) project to use of synchrotron-radiation computed tomography (SRCT) data to determine the conditions and mechanisms that lead to void nucleation in rolled alloys. The Advanced Light Source (ALS) at Lawrence Berkeley National Laboratory (LBNL) has provided SRCT data of a few specimens of 7075-T7351 aluminum plate (widely used for aerospace applications) stretched to failure, loaded in directions perpendicular and parallel to the rolling direction. The resolution of SRCT data is 900nm, which allows elucidation of the mechanisms governing void growth and coalescence. This resolution is not fine enough, however, for nucleation. We propose the use statistics and image processing techniques to obtain sub-resolution scale information from these data, and thus determine where in the specimen and when during the loading program nucleation occurs and the mechanisms that lead to it. Quantitative analysis of the tomography data, however, leads to the conclusion that the reconstruction process compromises the information obtained from the scans. Alternate, more powerful reconstruction algorithms are needed to address this problem, but those fall beyond the scope of this project.

43 PARTICLE ACCELERATORS↗

Optimizing High-Throughput Inference on Graph Neural Networks at Shared Computing Facilities with the NVIDIA Triton Inference Server

Abstract With machine learning applications now spanning a variety of computational tasks, multi-user shared computing facilities are devoting a rapidly increasing proportion of their resources to such algorithms. Graph neural networks (GNNs), for example, have provided astounding improvements in extracting complex signatures from data and are now widely used in a variety of applications, such as particle jet classification in high energy physics (HEP). However, GNNs also come with an enormous computational penalty that requires the use of GPUs to maintain reasonable throughput. At shared computing facilities, such as those used by physicists at Fermi National Accelerator Laboratory (Fermilab), methodical resource allocation and high throughput at the many-user scale are key to ensuring that resources are being used as efficiently as possible. These facilities, however, primarily provide CPU-only nodes, which proves detrimental to time-to-insight and computational throughput for workflows that include machine learning inference. In this work, we describe how a shared computing facility can use the NVIDIA Triton Inference Server to optimize its resource allocation and computing structure, recovering high throughput while scaling out to multiple users by massively parallelizing their machine learning inference. To demonstrate the effectiveness of this system in a realistic multi-user environment, we use the Fermilab Elastic Analysis Facility augmented with the Triton Inference Server to provide scalable and high-throughput access to a HEP-specific GNN and report on the outcome.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Local search to improve coordinate-based task mapping

We present a local search strategy to improve the coordinate-based mapping of a parallel job’s tasks to the MPI ranks of its parallel allocation in order to reduce network congestion and the job’s communication time. The goal is to reduce the number of network hops between communicating pairs of ranks. Our target is applications with a nearest-neighbor stencil communication pattern running on mesh systems with non-contiguous processor allocation, such as Cray XE and XK Systems. Utilizing the miniGhost mini-app, which models the shock physics application CTH, we demonstrate that our strategy reduces application running time while also reducing the runtime variability. Furthermore, we further show that mapping quality can vary based on the selected allocation algorithm, even between allocation algorithms of similar apparent quality.

97 MATHEMATICS AND COMPUTING↗

Layer-Parallel Training of Deep Residual Neural Networks

Residual neural networks (ResNets) are a promising class of deep neural networks that have shown excellent performance for a number of learning tasks, e.g., image classification and recognition. Mathematically, ResNet architectures can be interpreted as forward Euler discretizations of a nonlinear initial value problem whose time-dependent control variables represent the weights of the neural network. Hence, training a ResNet can be cast as an optimal control problem of the associated dynamical system. For similar time-dependent optimal control problems arising in engineering applications, parallel-in-time methods have shown notable improvements in scalability. This paper demonstrates the use of those techniques for efficient and effective training of ResNets. The proposed algorithms replace the classical (sequential) forward and backward propagation through the network layers with a parallel nonlinear multigrid iteration applied to the layer domain. This adds a new dimension of parallelism across layers that is attractive when training very deep networks. From this basic idea, we derive multiple layer-parallel methods. The most efficient version employs a simultaneous optimization approach where updates to the network parameters are based on inexact gradient information in order to speed up the training process. Finally, using numerical examples from supervised classification, we demonstrate that the new approach achieves a training performance similar to that of traditional methods, but enables layer-parallelism and thus provides speedup over layer-serial methods through greater concurrency.

97 MATHEMATICS AND COMPUTING↗

Large-scale Nanostructure Simulations from X-ray Scattering Data On Graphics Processor Clusters

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.

36 MATERIALS SCIENCE↗

Accuracy of the explicit energy-conserving particle-in-cell method for under-resolved simulations of capacitively coupled plasma discharges

The traditional explicit electrostatic momentum-conserving particle-in-cell algorithm requires strict resolution of the electron Debye length to deliver numerical stability and accuracy. The explicit electrostatic energy-conserving particle-in-cell algorithm alleviates this constraint with minimal modification to the traditional algorithm, retaining its simplicity, ease of parallelization, and acceleration on modern supercomputing architectures. In this article, we apply the algorithm to model a one-dimensional radio frequency capacitively coupled plasma discharge relevant to industrial applications. The energy-conserving approach closely matches the results from the momentum-conserving algorithm and retains accuracy even for cell sizes up to 8 times the electron Debye length. For even larger cells, the algorithm loses accuracy due to poor resolution of steep gradients within the radio frequency sheath. Accuracy can be recovered by adopting a non-uniform grid, which resolves the sheath and allows for cell sizes up to 32 times the electron Debye length in the quasi-neutral bulk of the discharge. The effect is an up to 8 times reduction in the number of required simulation cells, an improvement that can compound in higher-dimensional simulations. We therefore consider the explicit energy-conserving algorithm as a promising approach to significantly reduce the computational cost of full-scale device simulations and a pathway to delivering kinetic simulation capabilities of use to industry.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Exascale Hardware Architectures Working Group

The ASC Exascale Hardware Architecture working group is challenged to provide input on the following areas impacting the future use and usability of potential exascale computer systems: processor, memory, and interconnect architectures, as well as the power and resilience of these systems. Going forward, there are many challenging issues that will need to be addressed. First, power constraints in processor technologies will lead to steady increases in parallelism within a socket. Additionally, all cores may not be fully independent nor fully general purpose. Second, there is a clear trend toward less balanced machines, in terms of compute capability compared to memory and interconnect performance. In order to mitigate the memory issues, memory technologies will introduce 3D stacking, eventually moving on-socket and likely on-die, providing greatly increased bandwidth but unfortunately also likely providing smaller memory capacity per core. Off-socket memory, possibly in the form of non-volatile memory, will create a complex memory hierarchy. Third, communication energy will dominate the energy required to compute, such that interconnect power and bandwidth will have a significant impact. All of the above changes are driven by the need for greatly increased energy efficiency, as current technology will prove unsuitable for exascale, due to unsustainable power requirements of such a system. These changes will have the most significant impact on programming models and algorithms, but they will be felt across all layers of the machine. There is clear need to engage all ASC working groups in planning for how to deal with technological changes of this magnitude. The primary function of the Hardware Architecture Working Group is to facilitate codesign with hardware vendors to ensure future exascale platforms are capable of efficiently supporting the ASC applications, which in turn need to meet the mission needs of the NNSA Stockpile Stewardship Program. This issue is relatively immediate, as there is only a small window of opportunity to influence hardware design for 2018 machines. Given the short timeline a firm co-design methodology with vendors is of prime importance.

97 MATHEMATICS AND COMPUTING↗

Algorithmic Patterns for $$\mathcal {H}$$-Matrices on Many-Core Processors

In this work, we consider the reformulation of hierarchical ( \(\mathcal {H}\) ) matrix algorithms for many-core processors with a model implementation on graphics processing units (GPUs). \(\mathcal {H}\) matrices approximate specific dense matrices, e.g., from discretized integral equations or kernel ridge regression, leading to log-linear time complexity in dense matrix–vector products. The parallelization of \(\mathcal {H}\) matrix operations on many-core processors is difficult due to the complex nature of the underlying algorithms. While previous algorithmic advances for many-core hardware focused on accelerating existing \(\mathcal {H}\) matrix CPU implementations by many-core processors, we here aim at totally relying on that processor type. As main contribution, we introduce the necessary parallel algorithmic patterns allowing to map the full \(\mathcal {H}\) matrix construction and the fast matrix–vector product to many-core hardware. In this work, crucial ingredients are space filling curves, parallel tree traversal and batching of linear algebra operations. The resulting model GPU implementation hmglib is the, to the best of the authors knowledge, first entirely GPU-based Open Source \(\mathcal {H}\) matrix library of this kind. We investigate application examples as present in kernel ridge regression, Gaussian Process Regression and kernel-based interpolation. In this context, an in-depth performance analysis and a comparative performance study against a standard multi-core CPU \(\mathcal {H}\) matrix library highlights profound speedups of our many-core parallel approach.

97 MATHEMATICS AND COMPUTING↗

Characteristics of an envelope model for laser-plasma accelerator simulation

Simulation of laser-plasma accelerator (LPA) experiments is computationally intensive due to the disparate length scales involved. Current experiments extend hundreds of laser wavelengths transversely and many thousands in the propagation direction, making explicit PIC simulations enormously expensive and requiring massively parallel execution in 3D. Simulating the next generation of LPA experiments is expected to increase the computational requirements yet further, by a factor of 1000. We can substantially improve the performance of LPA simulations by modeling the envelope evolution of the laser field rather than the field itself. This allows for much coarser grids, since we need only resolve the plasma wavelength and not the laser wavelength, and therefore larger timesteps can be used. Thus an envelope model can result in savings of several orders of magnitude in computational resources. By propagating the laser envelope in a Galilean frame moving at the speed of light, dispersive errors can be avoided and simulations over long distances become possible. The primary limitation to this envelope model is when the laser pulse develops large frequency shifts, and thus the slowly-varying envelope assumption is no longer valid. Here we describe the model and its implementation, and show rigorous benchmarks for the algorithm, establishing second-order convergence and correct laser group velocity. We also demonstrate simulations of LPA phenomena such as self-focusing and meter-scale acceleration stages using the model.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Computing resonant modes of accelerator cavities by solving nonlinear eigenvalue problems via rational approximation

Here, we present an efficient and reliable algorithm for solving a class of nonlinear eigenvalue problems arising from the modeling of particle accelerator cavities. The eigenvalue nonlinearity in these problems results from the use of waveguides to couple external power sources or to allow certain excited electromagnetic modes to exit the cavity. We use a rational approximation to reduce the nonlinear eigenvalue problem first to a rational eigenvalue problem. We then apply a special linearization procedure to turn the rational eigenvalue problem into a larger linear eigenvalue problem with the same eigenvalues, which can be solved by existing iterative methods. By using a compact scheme to represent both the linearized operator and the eigenvectors to be computed, we obtain a numerical method that only involves solving linear systems of equations of the same dimension as the original nonlinear eigenvalue problem. We refer to this method as a compact rational Krylov (CORK) method. We implemented the CORK method in the Omega3P module of the Advanced Computational Electromagnetic 3D Parallel (ACE3P) simulation suite and validated it by comparing the computed cavity resonant frequencies and damping Q factors of a small model problem to those obtained from a fitting procedure that uses frequency responses computed by another ACE3P module called S3P. We also used the CORK method to compute trapped modes damped in an ideal eight 9-cell SRF cavity cryomodule. This was the first time it was possible to compute these modes directly. The damping Q factors of the computed modes match well with those measured in experiments and the difference in resonant frequencies is within the range introduced by cavity imperfection. Therefore, the CORK method is an extremely valuable tool for computational cavity design.

43 PARTICLE ACCELERATORS↗

Development of INSPCT-S for inspection of spent fuel pool

In this paper, we discuss an accurate and fast software tool (INSPCT-S, Inspection of Nuclear Spent fuel-Pool Calculation Tool, version Spreadsheet) developed for calculation of the response of fission chambers placed in a spent fuel pool, such as Atucha-I. INSPCT-S is developed for identification of suspicious regions of the pool that may have missing or substitute assemblies. INSPCT-S uses a hybrid algorithm based on the adjoint function methodology. The neutron source is comprised of spontaneous fission, ({alpha}, n) interactions, and subcritical multiplication. The former is evaluated using the ORIGEN-ARP code, and the latter is obtained with the fission matrix (FM) formulation. The FM coefficients are determined using the MCNP Monte Carlo code, and the importance function is determined using the PENTRAN 3-D parallel Sn code. Three databases for the neutron source, FM elements, and adjoint flux are prepared as functions of different parameters including burnup, cooling time, enrichment, and pool lattice size. INSPCT-S uses the aforementioned databases and systems of equations to calculate detector responses, which are subsequently compared with normalized experimental data. If this comparison is not satisfied, INSPCT-S utilizes color coding to identify the suspicious regions of a spent fuel pool. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗