Search NASA⌕ Search

SEARCH · Search NASA

Results for “Parallel computers.”

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 631 records · Page 35

Parallel checksumming of data chunks of a shared data object using a log-structured file system

Checksum values are generated and used to verify the data integrity. A client executing in a parallel computing system stores a data chunk to a shared data object on a storage node in the parallel computing system. The client determines a checksum value for the data chunk; and provides the checksum value with the data chunk to the storage node that stores the shared object. The data chunk can be stored on the storage node with the corresponding checksum value as part of the shared object. The storage node may be part of a Parallel Log-Structured File System (PLFS), and the client may comprise, for example, a Log-Structured File System client on a compute node or burst buffer. The checksum value can be evaluated when the data chunk is read from the storage node to verify the integrity of the data that is read.

97 MATHEMATICS AND COMPUTING↗

Computational homogenization at extreme scales

Multi-scale simulations at extreme scales in terms of both physical length scales and computational resources are presented. In this letter, we introduce a hierarchically parallel computational homogenization solver that employs hundreds of thousands of computing cores and resolves in material length scales (from to ). Simulations of this kind are important in understanding the multi-scale essence of many natural and synthetically made materials. Thus, we present a simulation consisting of 53.8 Billion finite elements with 28.1 Billion nonlinear equations that is solved on 393,216 computing cores (786,432 threads). The excellent parallel performance of the computational homogenization solver is demonstrated by a strong scaling test from 4,096 to 262,144 cores. A fully coupled multi-scale damage simulation shows a complex crack profile at the micro-scale and the macroscopic crack tunneling phenomenon. Such large and predictive simulations are an important step towards Virtual Materials Testing and can aid in development of new material formulations with extreme properties. Furthermore, the high computational efficiency of our computational homogenization solver holds great promise for utilizing the next generation of exascale parallel computing platforms that are expected to accelerate computations through orders of magnitude increase in parallelism rather than speed of each processor.

Engineering↗

Sierra/SolidMechanics 5.6 User's Manual

Sierra/SolidMechanics (Sierra/SM) is a three-dimensional solid mechanics code with a versatile element library, nonlinear material models, large deformation capabilities, and contact. It is built on the SIERRA Framework. SIERRA provides a data management framework in a parallel computing environment that allows the addition of capabilities in a modular fashion. Contact capabilities are parallel and scalable. This document provides information about the functionality in Sierra/SM and the command structure required to access this functionality in a user input file. This document is divided into chapters based primarily on functionality. For example, the command structure related to the use of various element types is grouped in one chapter; descriptions of material models are grouped in another chapter. Sierra/SM provides both explicit transient dynamics and implicit quasistatics and dynamics capabilities. Both the explicit and implicit modules are highly scalable in a parallel computing environment. In the past, the explicit and implicit capabilities were provided by two separate codes, known as Presto and Adagio, respectively. These capabilities have been consolidated into a single code. The executable is named Adagio, but it provides the full suite of solid mechanics capabilities, for both implicit and explicit. The Presto executable has been disabled as a consequence of this consolidation.

97 MATHEMATICS AND COMPUTING↗

A time-parallel method for scalable heat transfer simulations of additive manufacturing

Here, a major challenge in simulating the thermal behavior in additive manufacturing processes is the disparate length and time scales between transport phenomena occurring in the melt pool and the component. A common simulation approach relies on spatial decomposition for parallel computing, but due to the nature of heat transfer in AM, where most of the computational expenditure is localized near the melt pool, the computational speedup from spatial parallelization saturates quickly. Therefore, additional parallelism by means of time-domain decomposition is needed to fully take advantage of high-performance computing (HPC) resources. This work introduces a time-parallel method to improve the computational scalability of additive manufacturing simulations on HPC systems, while maintaining high temporal resolution of heat transfer near the melt pool. The method, inspired by the nonlinear paraexp formalism, performs an iterative superposition of nonlinear solutions to the initial value problem, integrating the heat equation across overlapping time-parallel intervals. For a single layer of the NIST AMB2018–01 L7 benchmark problem, the method achieves a 38.51x speedup in wall-clock time with a maximum error in the global temperature solution of 0.99%. This reduces the total solution time from 196.72 min to 5.11 min on 128 nodes of the ORNL Frontier supercomputer. The tradeoff between accuracy and total wall-clock time is investigated and recommendations for time-parallel deployment for AM problems are made.

Additive manufacturing↗

Implementation and Characterization of Three-Dimensional Particle-in-Cell Codes on Multiple-Instruction-Multiple-Data Massively Parallel Supercomputers

A three-dimensional electrostatic particle-in-cell (PIC) plasma simulation code has been developed on coarse-grain distributed-memory massively parallel computers with message passing communications. Our implementation is the generalization to three-dimensions of the general concurrent particle-in-cell (GCPIC) algorithm. In the GCPIC algorithm, the particle computation is divided among the processors using a domain decomposition of the simulation domain. In a three-dimensional simulation, the domain can be partitioned into one-, two-, or three-dimensional subdomains ("slabs," "rods," or "cubes") and we investigate the efficiency of the parallel implementation of the push for all three choices. The present implementation runs on the Intel Touchstone Delta machine at Caltech; a multiple-instruction-multiple-data (MIMD) parallel computer with 512 nodes. We find that the parallel efficiency of the push is very high, with the ratio of communication to computation time in the range 0.3%-10.0%. The highest efficiency (> 99%) occurs for a large, scaled problem with 64(sup 3) particles per processing node (approximately 134 million particles of 512 nodes) which has a push time of about 250 ns per particle per time step. We have also developed expressions for the timing of the code which are a function of both code parameters (number of grid points, particles, etc.) and machine-dependent parameters (effective FLOP rate, and the effective interprocessor bandwidths for the communication of particles and grid points). These expressions can be used to estimate the performance of scaled problems--including those with inhomogeneous plasmas--to other parallel machines once the machine-dependent parameters are known.

Lyster, P. M.↗

Improving computational efficiency of Monte Carlo simulations with variance reduction

CCFE perform Monte-Carlo transport simulations on large and complex tokamak models such as ITER. Such simulations are challenging since streaming and deep penetration effects are equally important. In order to make such simulations tractable, both variance reduction (VR) techniques and parallel computing are used. It has been found that the application of VR techniques in such models significantly reduces the efficiency of parallel computation due to 'long histories'. VR in MCNP can be accomplished using energy-dependent weight windows. The weight window represents an 'average behaviour' of particles, and large deviations in the arriving weight of a particle give rise to extreme amounts of splitting being performed and a long history. When running on parallel clusters, a long history can have a detrimental effect on the parallel efficiency - if one process is computing the long history, the other CPUs complete their batch of histories and wait idle. Furthermore some long histories have been found to be effectively intractable. To combat this effect, CCFE has developed an adaptation of MCNP which dynamically adjusts the WW where a large weight deviation is encountered. The method effectively 'de-optimises' the WW, reducing the VR performance but this is offset by a significant increase in parallel efficiency. Testing with a simple geometry has shown the method does not bias the result. This 'long history method' has enabled CCFE to significantly improve the performance of MCNP calculations for ITER on parallel clusters, and will be beneficial for any geometry combining streaming and deep penetration effects. (authors)

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Efficient Execution of Recursive Programs on Commodity Vector Hardware

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.

resursion↗

Research on Fast-Doppler-Broadening of neutron cross sections

A Fast-Doppler-Broadening method is developed in this work to broaden Continuous Energy neutron cross-sections for Monte Carlo calculations. Gauss integration algorithm and parallel computing are implemented in this method, which is unprecedented in the history of cross section processing. Compared to the traditional code (NJOY, SIGMA1, etc.), the new Fast-Doppler-Broadening method shows a remarkable speedup with keeping accuracy. The purpose of using Gauss integration is to avoid complex derivation of traditional broadening formula and heavy load of computing complementary error function that slows down the Doppler broadening process. The OpenMP environment is utilized in parallel computing which can take full advantage of modern multi-processor computers. Combination of the two can reduce processing time of main actinides (such as {sup 238}U, {sup 235}U) to an order of magnitude of 1{approx}2 seconds. This new method is fast enough to be applied to Online Doppler broadening. It can be combined or coupled with Monte Carlo transport code to solve temperature dependent problems and neutronics-thermal hydraulics coupled scheme which is a big challenge for the conventional NJOY-MCNP system. Examples are shown to determine the efficiency and relative errors compared with the NJOY results. A Godiva Benchmark is also used in order to test the ACE libraries produced by the new method. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Automatic Differentiation of C++ Codes on Emerging Manycore Architectures with Sacado

Automatic differentiation (AD) is a well-known technique for evaluating analytic derivatives of calculations implemented on a computer, with numerous software tools available for incorporating AD technology into complex applications. However, a growing challenge for AD is the efficient differentiation of parallel computations implemented on emerging manycore computing architectures such as multicore CPUs, GPUs, and accelerators as these devices become more pervasive. In this work, we explore forward mode, operator overloading-based differentiation of C++ codes on these architectures using the widely available Sacado AD software package. In particular, we leverage Kokkos, a C++ tool providing APIs for implementing parallel computations that is portable to a wide variety of emerging architectures. Here we describe the challenges that arise when differentiating code for these architectures using Kokkos, and two approaches for overcoming them that ensure optimal memory access patterns as well as expose additional dimensions of fine-grained parallelism in the derivative calculation. We describe the results of several computational experiments that demonstrate the performance of the approach on a few contemporary CPU and GPU architectures. We then conclude with applications of these techniques to the simulation of discretized systems of partial differential equations.

97 MATHEMATICS AND COMPUTING↗

Box schemes and their implementation on the iPSC/860

Research on algoriths for efficiently solving fluid flow problems on massively parallel computers is continued in the present paper. Attention is given to the implementation of a box scheme on the iPSC/860, a massively parallel computer with a peak speed of 10 Gflops and a memory of 128 Mwords. A domain decomposition approach to parallelism is used.

Chattot, J. J.↗

More About Software for No-Loss Computing

A document presents some additional information on the subject matter of "Integrated Hardware and Software for No- Loss Computing" (NPO-42554), which appears elsewhere in this issue of NASA Tech Briefs. To recapitulate: The hardware and software designs of a developmental parallel computing system are integrated to effectuate a concept of no-loss computing (NLC). The system is designed to reconfigure an application program such that it can be monitored in real time and further reconfigured to continue a computation in the event of failure of one of the computers. The design provides for (1) a distributed class of NLC computation agents, denoted introspection agents, that effects hierarchical detection of anomalies; (2) enhancement of the compiler of the parallel computing system to cause generation of state vectors that can be used to continue a computation in the event of a failure; and (3) activation of a recovery component when an anomaly is detected.

Edmonds, Iarina↗

Experience in using SIMD and MIMD parallelism for computational fluid dynamics

One of the key objectives of the Applied Research Branch in the Numerical Aerodynamic Simulation (NAS) Systems Division at NASA Ames Research Center is the accelerated introduction of highly parallel machines into a fully operational environment. In this report we summarize some of the experiences with the parallel testbed machines at the NAS Applied Research Branch. We discuss the performance results obtained from the implementation of two computational fluid dynamics (CFD) applications, an unstructured grid solver and a particle simulation, on the Connection Machine CM-2 and the Intel iPSC/860.

Simon, Horst D.↗

Calculation of Higher Eigenpairs of the Transport Equation Using IRAM Based on Domain Decomposition

The higher eigenpairs of the neutron transport (or diffusion) equation are useful in core flux distribution reconstruction, transient analysis, stability analysis, and loading pattern optimization. There are various methods for the calculation of higher eigenpairs, such as the power iteration method with filtering technique, the Generalized Modified Power Method (GMPM), the power iteration method with innovative boundary conditions for symmetric cores, and Krylov subspace methods. In regard to Krylov subspace methods, the Implicitly Restarted Arnoldi Method (IRAM) and ORTHOMIN were implemented to calculate higher eigenpairs for both transport and diffusion equations. In Warsa's work, however, there was a significant drawback that problems with energy upscatter could not be treated efficiently. In Modak's work, the mode to which his method was likely to converge depended on the shape of initial guess flux, and only one mode could be computed each time, making it impractical for real use. Furthermore, the computational burden for higher eigenpairs is generally much larger than for only the first eigenpair, especially for neutron transport equations. Therefore, the implementation of parallel computation for higher eigenpairs is essential to improve the efficiency, and no such work has been reported in literature so far. In this summary, we propose an innovative method to calculate higher eigenpairs of the neutron transport equation based on high performance computation. In this method, 2D Matrix Method of Characteristic (MMOC) is used to discretize the neutron transport equation; IRAM is used to evaluate higher eigenpairs; spatial domain decomposition is used to realize parallel computation. The highlight of this method is using the MGMD PGMRES (Multi-Group and Multi-Domain coupled Parallel Generalized Minimal Residual Method) algorithm to simultaneously accelerate the upscatter iterations and the inner interface angular fluxes iterations, avoiding the drawback in Warsa's method and the efficiency problem of conventional Jacobi iteration. To the best of our knowledge, our work represents the first application of IRAM based on domain decomposition for higher eigenpairs.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Resource Selection Using Execution and Queue Wait Time Predictions

Computational grids provide users with many possible places to execute their applications. We wish to help users select where to run their applications by providing predictions of the execution times of applications on space shared parallel computers and predictions of when scheduling systems for such parallel computers will start applications. Our predictions are based on instance based learning techniques and simulations of scheduling algorithms. We find that our execution time prediction techniques have an average error of 37 percent of the execution times for trace data recorded from SGI Origins at NASA Ames Research Center and that this error is 67 percent lower than the error of user estimates. We also find that the error when predicting how long applications will wait in scheduling queues is 95 percent of mean queue wait times when using our execution time predictions and this is 57 percent lower than if we use user execution time estimates.

Warren, Smith↗

Resource Selection Using Execution and Queue Wait Time Predictions

Computational grids provide users with many possible places to execute their applications. We wish to help users select where to run their applications by providing predictions of the execution times of applications on space shared parallel computers and predictions of when scheduling systems for such parallel computers will start applications. Our predictions are based on instance based learning techniques and simulations of scheduling algorithms. We find that our execution time prediction techniques have an average error of 37 percent of the execution times for trace data recorded from SGI Origins at NASA Ames Research Center and that this error is 67 percent lower than the error of user estimates. We also find that the error when predicting how long applications will wait in scheduling queues is 95 percent of mean queue wait times when using our execution time predictions and this is 57 percent lower than if we use user execution time estimates.

Smith, Warren↗

Execution environment for intelligent real-time control systems

Modern telerobot control technology requires the integration of symbolic and non-symbolic programming techniques, different models of parallel computations, and various programming paradigms. The Multigraph Architecture, which has been developed for the implementation of intelligent real-time control systems is described. The layered architecture includes specific computational models, integrated execution environment and various high-level tools. A special feature of the architecture is the tight coupling between the symbolic and non-symbolic computations. It supports not only a data interface, but also the integration of the control structures in a parallel computing environment.

Sztipanovits, Janos↗

Implementation of CMFD Acceleration Scheme in PROTEUS-MOC

The PROTEUS-MOC code is a three-dimensional (3D) transport code, which is being developed at Argonne National Laboratory under the DOE-NE's Nuclear Energy Advanced Modeling and Simulation (NEAMS) program. It provides a fully consistent 3D discretization of the transport equation by combining the two-dimensional (2D) method of characteristics (MOC) for the planar direction with the discontinuous Galerkin method for the treatment of the axial variable. PROTEUS-MOC uses the unstructured finite element method (FEM) and thus allows the explicit representation of any extruded geometries. It also has a capability of massively parallel computation in energy group, angle, plane, and space using a top-level GMRES linear algebra solver. For parallel computation in space, it uses an unstructured domain decomposition by adapting METIS. The computational burden for direct heterogeneous core calculations based on unstructured geometries is excessive for practical reactor problems, and hence it is essential to implement an efficient acceleration scheme in PROTEUS-MOC. Recently, the coarse mesh finite difference (CMFD) method, which was originally applied to the nodal diffusion methods, has been successfully applied to the MOC transport solvers, resulting in a remarkable reduction in the number of fission source iterations. In addition, a modified CMFD method has been developed to overcome the limitation of the conventional CMFD that is applicable only to neutronics solvers with a conservative discretization. This CMFD method was derived to be consistent with the FEM, which is based on a nonconservative discretization. In this method, the neutron balance in each coarse mesh is enforced by introducing a pseudo absorption cross section (PAXS), and the well-established alternating solution process of CMFD and transport calculations is employed to accelerate source convergence. This modified CMFD acceleration scheme was adopted in PROTEUS-MOC to accelerate the fission source iteration. Pin-cell and assembly based CMFD schemes have been implemented by defining individual pin cells or assemblies as a coarse mesh. For this, an algorithm to map the finite elements to coarse meshes was developed, and the routines were developed to determine the coupling coefficients and PAXS of the CMFD equations from the transport solution. Preliminary performance tests were carried out using the C5G7 benchmark problem. This paper presents the main features of this CMFD acceleration scheme and the preliminary test results.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗