Search NASA⌕ Search

SEARCH · Search NASA

Results for “Parallel processors”

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 55 records · Page 3

Software Synthesis for High Productivity Exascale Computing

Over the three years of our project, we accomplished three key milestones: We demonstrated how ideas from generative programming and software synthesis can help support the development of bulk-synchronous distributed memory kernels. These ideas are realized in a new language called MSL, a C-like language that combines synthesis features with high level notations for array manipulation and bulk-synchronous parallelism to simplify the semantic analysis required for synthesis. We also demonstrated that these high level notations map easily to low level C code and show that the performance of this generated code matches that of handwritten Fortran. Second, we introduced the idea of solver-aided domain-specific languages (SDSLs), which are an emerging class of computer-aided programming systems. SDSLs ease the construction of programs by automating tasks such as verification, debugging, synthesis, and non-deterministic execution. SDSLs are implemented by translating the DSL program into logical constraints. Next, we developed a symbolic virtual machine called Rosette, which simplifies the construction of such SDSLs and their compilers. We have used Rosette to build SynthCL, a subset of OpenCL that supports synthesis. Third, we developed novel numeric algorithms that move as little data as possible, either between levels of a memory hierarchy or between parallel processors over a network. We achieved progress in three aspects of this problem. First we determined lower bounds on communication. Second, we compared these lower bounds to widely used versions of these algorithms, and noted that these widely used algorithms usually communicate asymptotically more than is necessary. Third, we identified or invented new algorithms for most linear algebra problems that do attain these lower bounds, and demonstrated large speed-ups in theory and practice.

97 MATHEMATICS AND COMPUTING↗

EMAPS: An Efficient Multiscale Approach to Plasma Systems with Non-MHD Scale Effects

Global interactions of energetic ions with magnetoplasmas and neutral gases lie at the core of many space and laboratory plasma phenomena ranging from solar wind entry into and transport within planetary magnetospheres and exospheres to fast-ion driven instabilities in fusion devices to astrophysics-in-lab experiments. The ability of computational models to properly account for physical effects that underlie such interactions, namely ion kinetic, ion cyclotron, Hall, collisional and ionization processes is important for the success and planning of experimental research in plasma physics. Understanding the physics of energetic ions, in particular their nonlinear resonance interactions with Alfvén waves, is central to improving the heating performance of magnetically confined plasmas for future energy generation. Fluid models are not adequate for high-beta plasmas as they cannot fully capture ion kinetic and cyclotron physics (e.g., ion behavior in the presence of magnetic nulls, shock structures, plasma interpenetration, etc.). Recent results from global reconnection simulations show that even in a MHD-like regime there may be significant differences between kinetic and MHD simulations. Therefore, kinetic modeling becomes essential for meeting modern day challenges in plasma physics. The hybrid approximation is an intermediate approximation between the fluid and fully kinetic approximations. It eliminates light waves, removes the electron inertial temporal and spatial scales from the problem and enables full-orbit ion kinetics. As a result, hybrid codes have become effective tools for exploring ion-scale driven phenomena associated with ion beams, shocks, reconnection and turbulence that control the large-scale behavior of laboratory and space magnetoplasmas. A number of numerical issues, however, make three-dimensional (3D) large-scale hybrid simulations of inhomogeneous magnetized plasmas prohibitively expensive or even impossible. To resolve these difficulties we have developed a novel Event-driven Multiscale Asynchronous Parallel Simulation (EMAPS) technology that replaces time stepping with self-adaptive update events. Local calculations are carried out only on an “as needed basis”. EMAPS (i) guarantees accurate and stable processing of physical variables in time accurate simulations, and (ii) eliminates unnecessary computation. Applying EMAPS to the hybrid model has resulted in the development of a unique parallel code, dimension-independent (compile-time-configurable) HYPERS (Hybrid Parallel Event-Resolved Simulator) that scales to hundreds of thousands of parallel processors. HYPERS advances electromagnetic fields and particles asynchronously on time scales determined by local physical laws and mesh properties. To achieve high computational accuracy in complex device geometries, HYPERS employs high-fidelity Cartesian grids with masked conductive cells. The HYPERS model includes multiple ion species, energy and momentum conserving ion-ion collisions, and provides a number of approximations for plasma resistivity and vacuum regions. Both local and periodic boundary conditions are allowed. The HYPERS solver preserves zero divergence of magnetic field. The project has demonstrated HYPERS capabilities on a number of applications of interest to fusion and astrophysical plasma physics applications listed below. 1. Theta-pinch formation of FRCs The formation, spontaneous spin-up, and stability of theta-pinch formed field-reversed configurations have been studied self-consistently in 3D. The end-to-end hybrid simulations reveal poloidal profiles of implosion-driven fast toroidal plasma rotation and demonstrate three discharge regimes as a function of experimental parameters: the decaying stable configuration, the tilt unstable configuration, and the nonlinear evolution of a fast growing tearing mode. 2. FRC collisions with magnetic mirrors Interactions of fast plasma streams and objects with magnetic obstacles (dipoles, mirrors, etc) lie at the core of many space and laboratory plasma phenomena ranging from magnetoshells and solar wind interactions with planetary magnetospheres to compact fusion plasmas. HYPERS simulations are compared with data from the MSX experiment (LANL) that focuses on the physics of magnetized collisionless shocks through the acceleration and subsequent stagnation of FRC plasmoids against a strong magnetic mirrors and flux-conserving boundaries. 3. Exploding magnetoplasmas Results from hybrid simulations of two experiments at the LAPD and Nevada Terawatt Facility are discussed where short-pulse lasers are used to ablate solid targets to produce plasmas that expand across external magnetic fields. The first simulation recreates flutelike density striations observed at the leading edge of a carbon plasma and predicts an early destruction of the magnetic cavity in agreement with experimental evidence. In the second simulation a polyethylene target is ablated into a mixture of protons and carbon ions. A mechanism is demonstrated that allows protons to penetrate the magnetic field in the form of a collimated flow. The results are compared to experimental data and single-fluid MHD simulations. The EMAPS framework has the potential for wide application in many other engineering and scientific fields, such as climate models, biological systems, electronic devices, seismic events, oil reservation simulators that all involve advancing solutions of partial differential equations in time where the rate of activity can be adapted widely over the spatial domain depending on locally space/time phenomena (“events”).

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Final Report for "Implimentation and Evaluation of Multigrid Linear Solvers into Extended Magnetohydrodynamic Codes for Petascale Computing"

Extended magnetohydrodynamic (MHD) codes are used to model the large, slow-growing instabilities that are projected to limit the performance of International Thermonuclear Experimental Reactor (ITER). The multiscale nature of the extended MHD equations requires an implicit approach. The current linear solvers needed for the implicit algorithm scale poorly because the resultant matrices are so ill-conditioned. A new solver is needed, especially one that scales to the petascale. The most successful scalable parallel processor solvers to date are multigrid solvers. Applying multigrid techniques to a set of equations whose fundamental modes are dispersive waves is a promising solution to CEMM problems. For the Phase 1, we implemented multigrid preconditioners from the HYPRE project of the Center for Applied Scientific Computing at LLNL via PETSc of the DOE SciDAC TOPS for the real matrix systems of the extended MHD code NIMROD which is a one of the primary modeling codes of the OFES-funded Center for Extended Magnetohydrodynamic Modeling (CEMM) SciDAC. We implemented the multigrid solvers on the fusion test problem that allows for real matrix systems with success, and in the process learned about the details of NIMROD data structures and the difficulties of inverting NIMROD operators. The further success of this project will allow for efficient usage of future petascale computers at the National Leadership Facilities: Oak Ridge National Laboratory, Argonne National Laboratory, and National Energy Research Scientific Computing Center. The project will be a collaborative effort between computational plasma physicists and applied mathematicians at Tech-X Corporation, applied mathematicians Front Range Scientific Computations, Inc. (who are collaborators on the HYPRE project), and other computational plasma physicists involved with the CEMM project.

43 PARTICLE ACCELERATORS↗

Multicore: Fallout From a Computing Evolution (LBNL Summer Lecture Series)

Summer Lecture Series 2008: Parallel computing used to be reserved for big science and engineering projects, but in two years that's all changed. Even laptops and hand-helds use parallel processors. Unfortunately, the software hasn't kept pace. Kathy Yelick, Director of the National Energy Research Scientific Computing Center at Berkeley Lab, describes the resulting chaos and the computing community's efforts to develop exciting applications that take advantage of tens or hundreds of processors on a single chip.

Yelick, Kathy↗

Randomized selection on the GPU

We implement here a fast and memory-sparing probabilistic top N selection algorithm on the GPU. To our knowledge, this is the first direct selection in the literature for the GPU. The algorithm proceeds via a probabilistic-guess-and-chcck process searching for the Nth element. It always gives a correct result and always terminates. The use of randomization reduces the amount of data that needs heavy processing, and so reduces the average time required for the algorithm. Probabilistic Las Vegas algorithms of this kind are a form of stochastic optimization and can be well suited to more general parallel processors with limited amounts of fast memory.

97 MATHEMATICS AND COMPUTING↗

Comparing current cluster, massively parallel, and accelerated systems

Currently there is large architectural diversity in high perfonnance computing systems. They include 'commodity' cluster systems that optimize per-node performance for small jobs, massively parallel processors (MPPs) that optimize aggregate perfonnance for large jobs, and accelerated systems that optimize both per-node and aggregate performance but only for applications custom-designed to take advantage of such systems. Because of these dissimilarities, meaningful comparisons of achievable performance are not straightforward. In this work we utilize a methodology that combines both empirical analysis and performance modeling to compare clusters (represented by a 4,352-core IB cluster), MPPs (represented by a 147,456-core BG/P), and accelerated systems (represented by the 129,600-core Roadrunner) across a workload of four applications. Strengths of our approach include the ability to compare architectures - as opposed to specific implementations of an architecture - attribute each application's performance bottlenecks to characteristics unique to each system, and to explore performance scenarios in advance of their availability for measurement. Our analysis illustrates that application performance is essentially unrelated to relative peak performance but that application performance can be both predicted and explained using modeling.

97 MATHEMATICS AND COMPUTING↗

Singular value decomposition utilizing parallel algorithms on graphical processors

One of the current challenges in underwater acoustic array signal processing is the detection of quiet targets in the presence of noise. In order to enable robust detection, one of the key processing steps requires data and replica whitening. This, in turn, involves the eigen-decomposition of the sample spectral matrix, Cx = 1/K xKX(k)XH(k) where X(k) denotes a single frequency snapshot with an element for each element of the array. By employing the singular value decomposition (SVD) method, the eigenvectors and eigenvalues can be determined directly from the data without computing the sample covariance matrix, reducing the computational requirements for a given level of accuracy (van Trees, Optimum Array Processing). (Recall that the SVD of a complex matrix A involves determining V, , and U such that A = U VH where U and V are orthonormal and is a positive, real, diagonal matrix containing the singular values of A. U and V are the eigenvectors of AAH and AHA, respectively, while the singular values are the square roots of the eigenvalues of AAH.) Because it is desirable to be able to compute these quantities in real time, an efficient technique for computing the SVD is vital. In addition, emerging multicore processors like graphical processing units (GPUs) are bringing parallel processing capabilities to an ever increasing number of users. Since the computational tasks involved in array signal processing are well suited for parallelization, it is expected that these computations will be implemented using GPUs as soon as users have the necessary computational tools available to them. Thus, it is important to have an SVD algorithm that is suitable for these processors. This work explores the effectiveness of two different parallel SVD implementations on an NVIDIA Tesla C2050 GPU (14 multiprocessors, 32 cores per multiprocessor, 1.15 GHz clock - peed). The first algorithm is based on a two-step algorithm which bidiagonalizes the matrix using Householder transformations, and then diagonalizes the intermediate bidiagonal matrix through implicit QR shifts. This is similar to that implemented for real matrices by Lahabar and Narayanan ("Singular Value Decomposition on GPU using CUDA", IEEE International Parallel Distributed Processing Symposium 2009). The implementation is done in a hybrid manner, with the bidiagonalization stage done using the GPU while the diagonalization stage is done using the CPU, with the GPU used to update the U and V matrices. The second algorithm is based on a one-sided Jacobi scheme utilizing a sequence of pair-wise column orthogonalizations such that A is replaced by AV until the resulting matrix is sufficiently orthogonal (that is, equal to U ). V is obtained from the sequence of orthogonalizations, while can be found from the square root of the diagonal elements of AH A and, once is known, U can be found from column scaling the resulting matrix. These implementations utilize CUDA Fortran and NVIDIA's CUB LAS library. The primary goal of this study is to quantify the comparative performance of these two techniques against themselves and other standard implementations (for example, MATLAB). Considering that there is significant overhead associated with transferring data to the GPU and with synchronization between the GPU and the host CPU, it is also important to understand when it is worthwhile to use the GPU in terms of the matrix size and number of concurrent SVDs to be calculated.

99 GENERAL AND MISCELLANEOUS↗

GVT Algorithms and Discrete Event Dynamics on 128K+ Processor Cores

Parallel discrete event simulation (PDES) represents a class of codes that are challenging to scale to large number of processors due to tight global timestamp-ordering and fine-grained event execution. One of the critical factors in scaling PDES is the efficiency of the underlying global virtual time (GVT) algorithm needed for correctness of parallel execution and speed of progress. Although many GVT algorithms have been proposed previously, few have been proposed for scalable asynchronous execution and none customized to exploit one-sided communication. Moreover, the detailed performance effects of actual GVT algorithm implementations on large platforms are unknown. Here, three major GVT algorithms intended for scalable execution on high-performance systems are studied: (1) a synchronous GVT algorithm that affords ease of implementation, (2) an asynchronous GVT algorithm that is more complex to implement but can relieve blocking latencies, and (3) a variant of the asynchronous GVT algorithm, proposed and studied for the first time here, to exploit one-sided communication in extant supercomputing platforms. Performance results are presented of implementations of these algorithms on over 64,000 cores of a Cray XT5 system, exercised on a range of parameters: optimistic and conservative synchronization, fine- to medium-grained event computation, synthetic and non-synthetic applications, and different lookahead values. Performance of tens of billions of events executed per second are registered, exceeding the speeds of any known PDES engine, and showing asynchronous GVT algorithms to outperform state-of-the-art synchronous GVT algorithms. Detailed PDES-specific runtime metrics are presented to further the understanding of tightly-coupled discrete event execution dynamics on massively parallel platforms.

99 GENERAL AND MISCELLANEOUS↗

Multithreaded parallelization of the energy and analytic gradient in the fragment molecular orbital method

The fragment molecular orbital method in GAMESS is parallelized in a multithreaded OpenMP implementation combined with the MPI version of the two-level generalized distributed data interface. The energy and analytic gradient in gas phase and the polarizable continuum model of solvation are parallelized in this hybrid three-level scheme, achieving a large memory footprint reduction and a high parallel efficiency on Intel Xeon Phi processors. The parallel efficiency is demonstrated on the Stampede2 and Theta supercomputers using up to 2048 nodes (262 144 threads).

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Effective Vectorization with OpenMP 4.5

This paper describes how the Single Instruction Multiple Data (SIMD) model and its extensions in OpenMP work, and how these are implemented in different compilers. Modern processors are highly parallel computational machines which often include multiple processors capable of executing several instructions in parallel. Understanding SIMD and executing instructions in parallel allows the processor to achieve higher performance without increasing the power required to run it. SIMD instructions can significantly reduce the runtime of code by executing a single operation on large groups of data. The SIMD model is so integral to the processor s potential performance that, if SIMD is not utilized, less than half of the processor is ever actually used. Unfortunately, using SIMD instructions is a challenge in higher level languages because most programming languages do not have a way to describe them. Most compilers are capable of vectorizing code by using the SIMD instructions, but there are many code features important for SIMD vectorization that the compiler cannot determine at compile time. OpenMP attempts to solve this by extending the C++/C and Fortran programming languages with compiler directives that express SIMD parallelism. OpenMP is used to pass hints to the compiler about the code to be executed in SIMD. This is a key resource for making optimized code, but it does not change whether or not the code can use SIMD operations. However, in many cases critical functions are limited by a poor understanding of how SIMD instructions are actually implemented, as SIMD can be implemented through vector instructions or simultaneous multi-threading (SMT). We have found that it is often the case that code cannot be vectorized, or is vectorized poorly, because the programmer does not have sufficient knowledge of how SIMD instructions work.

97 MATHEMATICS AND COMPUTING↗

A parallel additive Schwarz preconditioned Jacobi-Davidson algorithm for polynomial eigenvalue problems in quantum dot simulation

We develop a parallel Jacobi-Davidson approach for finding a partial set of eigenpairs of large sparse polynomial eigenvalue problems with application in quantum dot simulation. A Jacobi-Davidson eigenvalue solver is implemented based on the Portable, Extensible Toolkit for Scientific Computation (PETSc). The eigensolver thus inherits PETSc's efficient and various parallel operations, linear solvers, preconditioning schemes, and easy usages. The parallel eigenvalue solver is then used to solve higher degree polynomial eigenvalue problems arising in numerical simulations of three dimensional quantum dots governed by Schroedinger's equations. We find that the parallel restricted additive Schwarz preconditioner in conjunction with a parallel Krylov subspace method (e.g. GMRES) can solve the correction equations, the most costly step in the Jacobi-Davidson algorithm, very efficiently in parallel. Besides, the overall performance is quite satisfactory. We have observed near perfect superlinear speedup by using up to 320 processors. The parallel eigensolver can find all target interior eigenpairs of a quintic polynomial eigenvalue problem with more than 32 million variables within 12 minutes by using 272 Intel 3.0 GHz processors.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Performance and Feature Improvements in Parareal-based Power System Dynamic Simulation

In recent years, a novel Parareal-based approach has been developed for fast transient simulations of large power system interconnections. Parareal belongs to the class of Parallel-in-time algorithms for solution of systems of differential-algebraic equations in parallel over an interval of time. The selection of a reasonably fast and accurate coarse solution is crucial to improve the performance of Parareal algorithm. Semi-analytical solution methods are one promising approach to achieve this goal. They have been investigated, and some preliminary results are presented here. In addition, Parareal-based simulator has been expanded to enable co-simulation with OpenDSS, a widely used open-source distribution system simulator. Preserving the parallel nature of the Parareal approach and taking advantage of the parallel capabilities of the latest versions of OpenDSS, each distribution system can be solved in their entirety on different processors in parallel within the main Parareal simulator. This paper also presents the structure of the transmission and distribution co-simulation and some results with different dynamic models of inverter-based resources in the distribution systems.

Park, Byungkwon↗

Scalable Parallel Execution of an Event-based Radio Signal Propagation Model for Cluttered 3D Terrains

Radio signal strength estimation is essential in many applications, including the design of military radio communications and industrial wireless installations. While classical approaches such as finite difference methods are well-known, new event-based models of radio signal propagation have been recently shown to deliver such estimates faster (via serial execution) than other methods. For scenarios with large or richly-featured geographical volumes, however, parallel processing is required to meet the memory and computation time demands. Here, we present a scalable and efficient parallel execution of a recently-developed event-based radio signal propagation model. We demonstrate its scalability to thousands of processors, with parallel speedups over 1000x. The speed and scale achieved by our parallel execution enable larger scenarios and faster execution than has ever been reported before.

99 GENERAL AND MISCELLANEOUS↗

A non-intrusive parallel-in-time adjoint solver with the XBraid library

In this paper, an adjoint solver for the multigrid-in-time software library XBraid is presented. XBraid provides a non-intrusive approach for simulating unsteady dynamics on multiple processors while parallelizing not only in space but also in the time domain (XBraid: Parallel multigrid in time, http://llnl.gov/casc/xbraid). It applies an iterative multigrid reduction in time algorithm to existing spatially parallel classical time propagators and computes the unsteady solution parallel in time. Techniques from Automatic Differentiation are used to develop a consistent discrete adjoint solver which provides sensitivity information of output quantities with respect to design parameter changes. The adjoint code runs backwards through the primal XBraid actions and accumulates gradient information parallel in time. It is highly non-intrusive as existing adjoint time propagators can easily be integrated through the adjoint interface. The adjoint code is validated on advection-dominated flow with periodic upstream boundary condition. We report it provides similar strong scaling results as the primal XBraid solver and offers great potential for speeding up the overall computational costs for sensitivity analysis using multiple processors.

97 MATHEMATICS AND COMPUTING↗

A communication-avoiding, hybrid-parallel, rank-revealing orthogonalization method.

Orthogonalization consumes much of the run time of many iterative methods for solving sparse linear systems and eigenvalue problems. Commonly used algorithms, such as variants of Gram-Schmidt or Householder QR, have performance dominated by communication. Here, 'communication' includes both data movement between the CPU and memory, and messages between processors in parallel. Our Tall Skinny QR (TSQR) family of algorithms requires asymptotically fewer messages between processors and data movement between CPU and memory than typical orthogonalization methods, yet achieves the same accuracy as Householder QR factorization. Furthermore, in block orthogonalizations, TSQR is faster and more accurate than existing approaches for orthogonalizing the vectors within each block ('normalization'). TSQR's rank-revealing capability also makes it useful for detecting deflation in block iterative methods, for which existing approaches sacrifice performance, accuracy, or both. We have implemented a version of TSQR that exploits both distributed-memory and shared-memory parallelism, and supports real and complex arithmetic. Our implementation is optimized for the case of orthogonalizing a small number (5-20) of very long vectors. The shared-memory parallel component uses Intel's Threading Building Blocks, though its modular design supports other shared-memory programming models as well, including computation on the GPU. Our implementation achieves speedups of 2 times or more over competing orthogonalizations. It is available now in the development branch of the Trilinos software package, and will be included in the 10.8 release.

99 GENERAL AND MISCELLANEOUS↗

Enhancing parallel quasi-static particle-in-cell simulations with a pipelining algorithm

A pipelining algorithm to overcome the limitation on scaling quasi-static particle-in-cell models of relativistic beams in plasmas to a very large number of processors is described. The pipelining algorithm uses multiple groups of processors and optimizes the job allocation on the processors in parallel computing. The algorithm is implemented on the quasi-static code QuickPIC and is shown to scale to over 10{sup 3} processors and increased the scale and speed by two orders of magnitude over the non-pipelined model. The new approach opens the door to performing full scale 3D simulations of future plasma wakefield accelerators or full lifetime models of beam interaction with electron clouds in circular accelerators such as the Large Hadron Collider (LHC) at CERN.

43 PARTICLE ACCELERATORS↗