Search NASA⌕ Search

SEARCH · Search NASA

Results for “parallelization”

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 37 records · Page 2

Streamline Integration Using MPI-Hybrid Parallelism on a Large Multicore Architecture

Streamline computation in a very large vector field data set represents a significant challenge due to the nonlocal and data-dependent nature of streamline integration. In this paper, we conduct a study of the performance characteristics of hybrid parallel programming and execution as applied to streamline integration on a large, multicore platform. With multicore processors now prevalent in clusters and supercomputers, there is a need to understand the impact of these hybrid systems in order to make the best implementation choice. We use two MPI-based distribution approaches based on established parallelization paradigms, parallelize over seeds and parallelize over blocks, and present a novel MPI-hybrid algorithm for each approach to compute streamlines. Our findings indicate that the work sharing between cores in the proposed MPI-hybrid parallel implementation results in much improved performance and consumes less communication and I/O bandwidth than a traditional, nonhybrid distributed implementation.

97 MATHEMATICS AND COMPUTING↗

Parallel heat transport in integrable and chaotic magnetic fields

The study of transport in magnetized plasmas is a problem of fundamental interest in controlled fusion, space plasmas, and astrophysics research. Three issues make this problem particularly challenging: (i) The extreme anisotropy between the parallel (i.e., along the magnetic field), {chi}{sub ||} , and the perpendicular, {chi}{sub Up-Tack }, conductivities ({chi}{sub ||} /{chi}{sub Up-Tack} may exceed 10{sup 10} in fusion plasmas); (ii) Nonlocal parallel transport in the limit of small collisionality; and (iii) Magnetic field lines chaos which in general complicates (and may preclude) the construction of magnetic field line coordinates. Motivated by these issues, we present a Lagrangian Green's function method to solve the local and non-local parallel transport equation applicable to integrable and chaotic magnetic fields in arbitrary geometry. The method avoids by construction the numerical pollution issues of grid-based algorithms. The potential of the approach is demonstrated with nontrivial applications to integrable (magnetic island), weakly chaotic (Devil's staircase), and fully chaotic magnetic field configurations. For the latter, numerical solutions of the parallel heat transport equation show that the effective radial transport, with local and non-local parallel closures, is non-diffusive, thus casting doubts on the applicability of quasilinear diffusion descriptions. General conditions for the existence of non-diffusive, multivalued flux-gradient relations in the temperature evolution are derived.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Parallel Reconstruction of Three Dimensional Magnetohydrodynamic Equilibria in Plasma Confinement Devices

Fast, accurate three dimensional reconstructions of plasma equilibria, crucial for physics interpretation of fusion data generated within confinement devices like stellarators/ tokamaks, are computationally very expensive and routinely require days, even weeks, to complete using serial approaches. Here, we present a parallel implementation of the three dimensional plasma reconstruction code, V3FIT. A formal analysis to identify the performance bottlenecks and scalability limits of this new parallel implementation, which combines both task and data parallelism, is presented. The theoretical findings are supported by empirical performance results on several thousands of processor cores of a Cray XC30 supercomputer. Parallel V3FIT is shown to deliver over 40X speedup, enabling fusion scientists to carry out three dimensional plasma equilibrium reconstructions at unprecedented scales in only a few hours (instead of in days/weeks) for the first time.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

PANDORA: A Parallel Dendrogram Construction Algorithm for Single Linkage Clustering on GPU

This paper introduces Pandora, a parallel algorithm for computing dendrograms, the hierarchical cluster trees for single linkage clustering (SLC). Current parallel approaches construct dendrograms by partitioning a minimum spanning tree and removing edges. However, they struggle with skewed, hard-to-parallelize real-world dendrograms. Consequently, computing dendrograms is the sequential bottleneck in HDBSCAN*[21], a popular SLC variant. Pandora uses recursive tree contraction to address this limitation. Pandora contracts nodes to construct progressively smaller trees. It computes the smallest contracted dendrogram and expands it by inserting contracted edges. This recursive strategy is highly parallel, skew-independent, work-optimal, and well-suited for GPUs and multicores. We develop a performance portable implementation of Pandora in Kokkos[31] and evaluate its performance on multicore CPUs and multi-vendor GPUs (e.g., Nvidia, AMD) for dendrogram construction in HDBSCAN*. Multithreaded Pandora is 2.2x faster than the current best-multithreaded implementation. Our GPU version achieves 6-20x speedup on AMD GPUs and 10-37x on NVIDIA GPUs over multithreaded Pandora. Pandora removes HDBSCAN*’s sequential bottleneck, greatly boosting efficiency, particularly with GPUs.

Sao, Piyush↗

Extracting SIMD Parallelism from Recursive Task-Parallel Programs

The pursuit of computational efficiency has led to the proliferation of throughput-oriented hardware, from GPUs to increasingly wide vector units on commodity processors and accelerators. This hardware is designed to efficiently execute data-parallel computations in a vectorized manner. However, many algorithms are more naturally expressed as divide-and-conquer, recursive, task-parallel computations. In the absence of data parallelism, it seems that such algorithms are not well suited to throughput-oriented architectures. This paper presents a set of novel code transformations that expose the data parallelism latent in recursive, task- parallel programs. These transformations facilitate straightforward vectorization of task-parallel programs on commodity hardware. We also present scheduling policies that maintain high utilization of vector resources while limiting space usage. Across several task-parallel benchmarks, we demonstrate both efficient vector resource utilization and substantial speedup on chips using Intel’s SSE4.2 vector units, as well as accelerators using Intel’s AVX512 units. We then show through rigorous sampling that, in practice, our vectorization techniques are effective for a much larger class of programs.

Recursive Programs, task parallelism, vectorizatio↗

Utilizing High Performance Computing for Chemistry: Parallel Computational Chemistry

Parallel hardware has become readily available to the computational chemistry research community. This perspective will review the current state of parallel computational chemistry software utilizing high-performance parallel computing platforms.Hardware and software trends and their effect on quantum chemistry methodologies, algorithms, and software development will also be discussed.

99 GENERAL AND MISCELLANEOUS↗

Basker: Parallel sparse LU factorization utilizing hierarchical parallelism and data layouts

Transient simulation in circuit simulation tools, such as SPICE and Xyce, depend on scalable and robust sparse LU factorizations for efficient numerical simulation of circuits and power grids. As the need for simulations of very large circuits grow, the prevalence of multicore architectures enable us to use shared memory parallel algorithms for such simulations. A parallel factorization is a critical component of such shared memory parallel simulations. We develop a parallel sparse factorization algorithm that can solve problems from circuit simulations efficiently, and map well to architectural features. This new factorization algorithm exposes hierarchical parallelism to accommodate irregular structure that arise in our target problems. It also uses a hierarchical two-dimensional data layout which reduces synchronization costs and maps to memory hierarchy found in multicore processors. We present an OpenMP based implementation of the parallel algorithm in a new multithreaded solver called Basker in the Trilinos framework. Here, we present performance evaluations of Basker on the Intel SandyBridge and Xeon Phi platforms using circuit and power grid matrices taken from the University of Florida sparse matrix collection and from Xyce circuit simulation. Basker achieves a geometric mean speedup of 5.91× on CPU (16 cores) and 7.4× on Xeon Phi (32 cores) relative to state-of-the-art solver KLU. Basker outperforms Intel MKL Pardiso solver (PMKL) by as much as 30× on CPU (16 cores) and 7.5× on Xeon Phi (32 cores) for low fill-in circuit matrices. Furthermore, Basker provides 5.4× speedup on a challenging matrix sequence taken from an actual Xyce simulation.

97 MATHEMATICS AND COMPUTING↗

The effect of exit beam phase aberrations on parallel beam coherent x-ray reconstructions

Diffraction artifacts from imperfect x-ray windows near the sample are an important consideration in the design of coherent x-ray diffraction measurements. In this study, we used simulated and experimental diffraction patterns in two and three dimensions to explore the effect of phase imperfections in a beryllium window (such as a void or inclusion) on the convergence behavior of phasing algorithms and on the ultimate reconstruction. A predictive relationship between beam wavelength, sample size, and window position was derived to explain the dependence of reconstruction quality on beryllium defect size. Defects corresponding to this prediction cause the most damage to the sample exit wave and induce signature error oscillations during phasing that can be used as a fingerprint of experimental x-ray window artifacts. The relationship between x-ray window imperfection size and coherent x-ray diffractive imaging reconstruction quality explored in this work can play an important role in designing high-resolution in situ coherent imaging instrumentation and will help interpret the phasing behavior of coherent diffraction measured in these in situ environments.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Transverse dynamics of water across the melting: a parallel Neutron and X Ray Inelastic Scattering study

Joint Inelastic Neutron and X Ray Scattering (INS and IXS) measurements have been performed on heavy water across the melting point. The spectra bear clear evidence on a low and a high frequency inelastic shoulders related to a transverse and a longitudinal mode respectively. Upon increasing the momentum transfer, the spectral shape evolves from a viscoelastic regime, where the low frequency mode is clearly overdamped, towards an elastic region where its propagation becomes allowed. The crossover between the two regimes occurs whenever both the position and the width of the low frequency mode match the inverse of the structural relaxation time. Furthermore a discontinuity in the solid phase in the the shear-mode frequency, which tends to disappear upon increasing the exchanged momentum, is observed.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Parallel PIPS-SBB: multi-level parallelism for stochastic mixed-integer programs

PIPS-SBB is a distributed-memory parallel solver with a scalable data distribution paradigm. It is designed to solve MIPs with a dual-block angular structure, which is characteristic of deterministic-equivalent Stochastic Mixed-Integer Programs (SMIPs). In this paper, we present two different parallelizations of Branch & Bound (B&B), implementing both as extensions of PIPS-SBB, thus adding an additional layer of parallelism. In the first of the proposed frameworks, PIPS-PSBB, the coordination and load-balancing of the different optimization workers is done in a decentralized fashion. This new framework is designed to ensure all available cores are processing the most promising parts of the B&B tree. The second, ug[PIPS-SBB,MPI], is a parallel implementation using the Ubiquity Generator (UG), a universal framework for parallelizing B&B tree search that has been sucessfully applied to other MIP solvers. We show the effects of leveraging multiple levels of parallelism in potentially improving scaling performance beyond thousands of cores.

97 MATHEMATICS AND COMPUTING↗

First experiments probing the collision of parallel magnetic fields using laser-produced plasmas

Novel experiments to study the strongly-driven collision of parallel magnetic fields in β ∼ 10, laser-produced plasmas have been conducted using monoenergetic proton radiography. These experiments were designed to probe the process of magnetic flux pileup, which has been identified in prior laser-plasma experiments as a key physical mechanism in the reconnection of anti-parallel magnetic fields when the reconnection inflow is dominated by strong plasma flows. In the present experiments using colliding plasmas carrying parallel magnetic fields, the magnetic flux is found to be conserved and slightly compressed in the collision region. Two-dimensional (2D) particle-in-cell simulations predict a stronger flux compression and amplification of the magnetic field strength, and this discrepancy is attributed to the three-dimensional (3D) collision geometry. Future experiments may drive a stronger collision and further explore flux pileup in the context of the strongly-driven interaction of magnetic fields.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Neoclassical parallel flow calculation in the presence of external parallel momentum sources in Heliotron J

A moment approach to calculate neoclassical transport in non-axisymmetric torus plasmas composed of multiple ion species is extended to include the external parallel momentum sources due to unbalanced tangential neutral beam injections (NBIs). The momentum sources that are included in the parallel momentum balance are calculated from the collision operators of background particles with fast ions. This method is applied for the clarification of the physical mechanism of the neoclassical parallel ion flows and the multi-ion species effect on them in Heliotron J NBI plasmas. It is found that parallel ion flow can be determined by the balance between the parallel viscosity and the external momentum source in the region where the external source is much larger than the thermodynamic force driven source in the collisional plasmas. This is because the friction between C{sup 6+} and D{sup +} prevents a large difference between C{sup 6+} and D{sup +} flow velocities in such plasmas. The C{sup 6+} flow velocities, which are measured by the charge exchange recombination spectroscopy system, are numerically evaluated with this method. It is shown that the experimentally measured C{sup 6+} impurity flow velocities do not contradict clearly with the neoclassical estimations, and the dependence of parallel flow velocities on the magnetic field ripples is consistent in both results.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

A Performance Model for Allocating the Parallelism in a Multigrid-in-Time Solver

The traditional way to numerically solve time-dependent problems is to sequentially march through time, solving for one time step and then the next. The parallelism in this approach is limited to the spatial dimension, which is quickly exhausted, causing the gain from using more processors to solve a problem to diminish. One approach to overcome this barrier is to use methods that are parallel in time. These methods have the potential to achieve dramatically better performance compared to time-stepping approaches, but achieving this performance requires carefully choosing the amount of parallelism devoted to space versus the amount devoted to time. Here, we present a performance model that, for a multigrid-in-time solver, makes the decision on when to switch to parallel-in-time and on how much parallelism to devote to space vs. time. In our experiments, the model selects the best parallel configuration in most of our test cases and a configuration close to the best one in all other cases.

97 MATHEMATICS AND COMPUTING↗

Accelerating Lattice QCD Multigrid on GPUs Using Fine-Grained Parallelization

The past decade has witnessed a dramatic acceleration of lattice quantum chromodynamics calculations in nuclear and particle physics. This has been due to both significant progress in accelerating the iterative linear solvers using multi-grid algorithms, and due to the throughput improvements brought by GPUs. Deploying hierarchical algorithms optimally on GPUs is non-trivial owing to the lack of parallelism on the coarse grids, and as such, these advances have not proved multiplicative. Using the QUDA library, we demonstrate that by exposing all sources of parallelism that the underlying stencil problem possesses, and through appropriate mapping of this parallelism to the GPU architecture, we can achieve high efficiency even for the coarsest of grids. Results are presented for the Wilson-Clover discretization, where we demonstrate up to 10x speedup over present state-of-the-art GPU-accelerated methods on Titan. Finally, we look to the future, and consider the software implications of our findings.

Clark, M. A.↗

PLATO (Parallel Load Assignment Tool): A Parallel Workload Partitioner for Particle-based Methods

A key first step in many large-scale high-performance scientific computing applications is the distribution of the computational domain across all participating processors of a parallel computing platform such as a shared memory machine, cluster or a tightly-coupled supercomputer. The primary goal of such a partitioning step is to ensure that the computational work performed by each process remains load-balanced throughout the execution of the parallel application. In this project, a general-purpose parallel partitioning tool, code-named PLATO, was designed and implemented. PLATO assumes that the workload is a set of points in an n-dimensional metric space. These points are abstractions of computational units that may represent point-like computational units such as celestial objects in astrophysics, atoms in materials sciences and molecules in biology/chemistry, or may represent more abstract computational units such as grid points in finite difference methods, grid elements in finite element methods or n-dimensional points in the feature space of a host of other applications. PLATO supports several key and useful capabilities, namely: (a) multiple load partitioning strategies, (b) support for arbitrary user-defined processor topology, (c) pre-fetching data from neighboring processors based on user-defined parameters, and (d) support for arbitrary number and types of attributes associated with each point in the computational domain. PLATO is implemented in the C programming language using Message Passing Interface (MPI) for inter-processor communications and its parallel performance tested on up to 48 processors of a shared memory machine. This report includes a detailed performance study of PLATO using parallel speedup and iso-efficiency as the two main metrics of parallel performance. In addition, a user manual and DOXYGEN generated code documentation have been included for completeness.

97 MATHEMATICS AND COMPUTING↗

Shift-and-invert parallel spectral transformation eigensolver: Massively parallel performance for density-functional based tight-binding

The Shift-and-invert parallel spectral transformations (SIPs), a computational approach to solve sparse eigenvalue problems, is developed for massively parallel architectures with exceptional parallel scalability and robustness. The capabilities of SIPs are demonstrated by diagonalization of density-functional based tight-binding (DFTB) Hamiltonian and overlap matrices for single-wall metallic carbon nanotubes, diamond nanowires, and bulk diamond crystals. The largest (smallest) example studied is a 128,000 (2000) atom nanotube for which ~330,000 (~5600) eigenvalues and eigenfunctions are obtained in ~190 (~5) seconds when parallelized over 266,144 (16,384) Blue Gene/Q cores. Weak scaling and strong scaling of SIPs are analyzed and the performance of SIPs is compared with other novel methods. Different matrix ordering methods are investigated to reduce the cost of the factorization step, which dominates the time-to-solution at the strong scaling limit. As a result, a parallel implementation of assembling the density matrix from the distributed eigenvectors is demonstrated.

36 MATERIALS SCIENCE↗