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 1,081 records · Page 60

Parallel performance of algebraic multigrid domain decomposition

Algebraic multigrid (AMG) is a widely used scalable solver and preconditioner for large-scale linear systems resulting from the discretization of a wide class of elliptic PDEs. While AMG has optimal computational complexity, the cost of communication has become a significant bottleneck that limits its scalability as processor counts continue to grow on modern machines. This article examines the design, implementation, and parallel performance of a novel algorithm, algebraic multigrid domain decomposition (AMG-DD), designed specifically to limit communication. The goal of AMG-DD is to provide a low-communication alternative to standard AMG V-cycles by trading some additional computational overhead for a significant reduction in communication cost. Numerical results show that AMG-DD achieves superior accuracy per communication cost compared with AMG, and speedup over AMG is demonstrated on a large GPU cluster.

97 MATHEMATICS AND COMPUTING↗

Modeling a Million-Node Slim Fly Network Using Parallel Discrete-Event Simulation

As supercomputers close in on exascale performance, the increased number of processors and processing power translates to an increased demand on the underlying network interconnect. The Slim Fly network topology, a new lowdiameter and low-latency interconnection network, is gaining interest as one possible solution for next-generation supercomputing interconnect systems. In this paper, we present a high-fidelity Slim Fly it-level model leveraging the Rensselaer Optimistic Simulation System (ROSS) and Co-Design of Exascale Storage (CODES) frameworks. We validate our Slim Fly model with the Kathareios et al. Slim Fly model results provided at moderately sized network scales. We further scale the model size up to n unprecedented 1 million compute nodes; and through visualization of network simulation metrics such as link bandwidth, packet latency, and port occupancy, we get an insight into the network behavior at the million-node scale. We also show linear strong scaling of the Slim Fly model on an Intel cluster achieving a peak event rate of 36 million events per second using 128 MPI tasks to process 7 billion events. Detailed analysis of the underlying discrete-event simulation performance shows that a million-node Slim Fly model simulation can execute in 198 seconds on the Intel cluster.

interconnection networks↗

Approximate algorithms for partitioning and assignment problems

The problem of optimally assigning the modules of a parallel/pipelined program over the processors of a multiple computer system under certain restrictions on the interconnection structure of the program as well as the multiple computer system was considered. For a variety of such programs it is possible to find linear time if a partition of the program exists in which the load on any processor is within a certain bound. This method, when combined with a binary search over a finite range, provides an approximate solution to the partitioning problem. The specific problems considered were: a chain structured parallel program over a chain-like computer system, multiple chain-like programs over a host-satellite system, and a tree structured parallel program over a host-satellite system. For a problem with m modules and n processors, the complexity of the algorithm is no worse than O(mnlog(W sub T/epsilon)), where W sub T is the cost of assigning all modules to one processor and epsilon the desired accuracy.

Iqbal, M. A.↗

Parallel Processing Systems for Passive Ranging During Helicopter Flight

The complexity of rotorcraft missions involving operations close to the ground result in high pilot workload. In order to allow a pilot time to perform mission-oriented tasks, sensor-aiding and automation of some of the guidance and control functions are highly desirable. Images from an electro-optical sensor provide a covert way of detecting objects in the flight path of a low-flying helicopter. Passive ranging consists of processing a sequence of images using techniques based on optical low computation and recursive estimation. The passive ranging algorithm has to extract obstacle information from imagery at rates varying from five to thirty or more frames per second depending on the helicopter speed. We have implemented and tested the passive ranging algorithm off-line using helicopter-collected images. However, the real-time data and computation requirements of the algorithm are beyond the capability of any off-the-shelf microprocessor or digital signal processor. This paper describes the computational requirements of the algorithm and uses parallel processing technology to meet these requirements. Various issues in the selection of a parallel processing architecture are discussed and four different computer architectures are evaluated regarding their suitability to process the algorithm in real-time. Based on this evaluation, we conclude that real-time passive ranging is a realistic goal and can be achieved with a short time.

Sridhar, Bavavar↗

Finding Tropical Cyclones on a Cloud Computing Cluster: Using Parallel Virtualization for Large-Scale Climate Simulation Analysis

Extensive computing power has been used to tackle issues such as climate changes, fusion energy, and other pressing scientific challenges. These computations produce a tremendous amount of data; however, many of the data analysis programs currently only run a single processor. In this work, we explore the possibility of using the emerging cloud computing platform to parallelize such sequential data analysis tasks. As a proof of concept, we wrap a program for analyzing trends of tropical cyclones in a set of virtual machines (VMs). This approach allows the user to keep their familiar data analysis environment in the VMs, while we provide the coordination and data transfer services to ensure the necessary input and output are directed to the desired locations. This work extensively exercises the networking capability of the cloud computing systems and has revealed a number of weaknesses in the current cloud system software. In our tests, we are able to scale the parallel data analysis job to a modest number of VMs and achieve a speedup that is comparable to running the same analysis task using MPI. However, compared to MPI based parallelization, the cloud-based approach has a number of advantages. The cloud-based approach is more flexible because the VMs can capture arbitrary software dependencies without requiring the user to rewrite their programs. The cloud-based approach is also more resilient to failure; as long as a single VM is running, it can make progress while as soon as one MPI node fails the whole analysis job fails. In short, this initial work demonstrates that a cloud computing system is a viable platform for distributed scientific data analyses traditionally conducted on dedicated supercomputing systems.

97 MATHEMATICS AND COMPUTING↗

Time parallelization of plasma simulations using the parareal algorithm

Simulation of fusion plasmas involve a broad range of timescales. In magnetically confined plasmas, such as in ITER, the timescale associated with the microturbulence responsible for transport and confinement timescales vary by an order of 10^6 10^9. Simulating this entire range of timescales is currently impossible, even on the most powerful supercomputers available. Space parallelization has so far been the most common approach to solve partial differential equations. Space parallelization alone has led to computational saturation for fluid codes, which means that the walltime for computaion does not linearly decrease with the increasing number of processors used. The application of the parareal algorithm to simulations of fusion plasmas ushers in a new avenue of parallelization, namely temporal parallelization. The algorithm has been successfully applied to plasma turbulence simulations, prior to which it has been applied to other relatively simpler problems. This work explores the extension of the applicability of the parareal algorithm to ITER relevant problems, starting with a diffusion-convection model.

Samaddar, D.↗

Integrated Data and Task Management for Scientific Applications

Several emerging application areas require intelligent management of distributed data and tasks that encapsulate execution units for collection of processors or processor groups. This paper describes an integration of data and task parallelism to address the needs of such applications in context of the Global Array (GA) programming model. GA provides programming interfaces for managing shared arrays based on non-partitioned global address space programming model concepts. Compatibility with MPI enables the scientific programmer to benefit from performance and productivity advantages of these high level programming abstractions using standard programming languages and compilers.

97 MATHEMATICS AND COMPUTING↗

SQDFT: Spectral Quadrature method for large-scale parallel O ( N ) Kohn–Sham calculations at high temperature

We present SQDFT: a large-scale parallel implementation of the Spectral Quadrature (SQ) method for $\mathscr{O}(N)$ Kohn–Sham Density Functional Theory (DFT) calculations at high temperature. Specifically, we develop an efficient and scalable finite-difference implementation of the infinite-cell Clenshaw–Curtis SQ approach, in which results for the infinite crystal are obtained by expressing quantities of interest as bilinear forms or sums of bilinear forms, that are then approximated by spatially localized Clenshaw–Curtis quadrature rules. We demonstrate the accuracy of SQDFT by showing systematic convergence of energies and atomic forces with respect to SQ parameters to reference diagonalization results, and convergence with discretization to established planewave results, for both metallic and insulating systems. Here, we further demonstrate that SQDFT achieves excellent strong and weak parallel scaling on computer systems consisting of tens of thousands of processors, with near perfect $\mathscr{O}(N)$ scaling with system size and wall times as low as a few seconds per self-consistent field iteration. Finally, we verify the accuracy of SQDFT in large-scale quantum molecular dynamics simulations of aluminum at high temperature.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

A nonhydrostatic formulation for MPAS-Ocean

The Model for Prediction Across Scales-Ocean (MPAS-Ocean) is an open-source, global ocean model and is one component of a family of climate models within the MPAS framework, including atmosphere, sea-ice, and land-ice models. Here, in this work, a new formulation for the ocean model is presented that solves the nonhydrostatic, incompressible Boussinesq equations on an unstructured, staggered, z-level grid. The introduction of this nonhydrostatic capability is necessary for the resolution of internal wave dynamics and large eddy simulations. Compared to the standard, hydrostatic formulation, a nonhydrostatic pressure solver and a vertical momentum equation are added, where the PETSc (Portable Extensible Toolkit for Scientific Computation) library is used for the inversion of a large sparse system for the nonhydrostatic pressure. Numerical results on a stratified seiche, internal solitary wave, overflow and lock-exchange test cases are presented, and the parallel efficiency of the code is evaluated using up to 1024 processors.

3D Poisson equation↗

A nonhydrostatic formulation for MPAS-Ocean

The Model for Prediction Across Scales-Ocean (MPAS-Ocean) is an open-source, global ocean model and is one component of a family of climate models within the MPAS framework, including atmosphere, sea-ice, and land-ice models. Here, in this work, a new formulation for the ocean model is presented that solves the nonhydrostatic, incompressible Boussinesq equations on an unstructured, staggered, z-level grid. The introduction of this nonhydrostatic capability is necessary for the resolution of internal wave dynamics and large eddy simulations. Compared to the standard, hydrostatic formulation, a nonhydrostatic pressure solver and a vertical momentum equation are added, where the PETSc (Portable Extensible Toolkit for Scientific Computation) library is used for the inversion of a large sparse system for the nonhydrostatic pressure. Numerical results on a stratified seiche, internal solitary wave, overflow and lock-exchange test cases are presented, and the parallel efficiency of the code is evaluated using up to 1024 processors.

3D Poisson equation↗

An optimal, parallel, fully implicit Newton-Krylov solver for three-dimensional viscoresistive magnetohydrodynamics

The conceptual development and implementation of a scalable nonlinear solver for the time-dependent three-dimensional (3D) compressible resistive magnetohydrodynamics model (MHD) is discussed. The approach is based on Jacobian-free Newton-Krylov technology, preconditioned with multigrid methods for algorithmic scalability. The key to the approach is the reformulation of the hyperbolic MHD system into a parabolic one, which is amenable to multigrid techniques. Such reformulation (parabolization) aims to render the modified system block diagonally dominant (unlike the original MHD system, which is diagonally submissive for implicit time steps {delta}t larger than the explicit Courant-Friedrichs-Lewy time step). The algorithm has been tested on a variety of 2D and 3D configurations which demonstrate its excellent algorithmic scalability properties. In particular, it is shown that, serially, the CPU time for a given simulation scales linearly with the number of unknowns, and that large CPU gains ({approx}30) are attainable with respect to an explicit approach even for moderate grids (256x256). In parallel, the algorithm features excellent scalability properties up to thousands of processors (4096) and millions of unknowns ({approx}1.3x10{sup 8})

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Scalable parallel solution coupling for multi-physics reactor simulation.

Reactor simulation depends on the coupled solution of various physics types, including neutronics, thermal/hydraulics, and structural mechanics. This paper describes the formulation and implementation of a parallel solution coupling capability being developed for reactor simulation. The coupling process consists of mesh and coupler initialization, point location, field interpolation, and field normalization. We report here our test of this capability on an example problem, namely, a reflector assembly from an advanced burner test reactor. Performance of this coupler in parallel is reasonable for the chosen problem size and range of processor counts. The runtime is dominated by startup costs, which amortize over the entire coupled simulation. Future efforts will include adding more sophisticated interpolation and normalization methods, to accommodate different numerical solvers used in various physics modules and to obtain better conservation properties for certain field types.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Accurate and Scalable O(N) Algorithm for First-Principles Molecular-Dynamics Computations on Large Parallel Computers

We present the first truly scalable first-principles molecular dynamics algorithm with O(N) complexity and controllable accuracy, capable of simulating systems with finite band gaps of sizes that were previously impossible with this degree of accuracy. By avoiding global communications, we provide a practical computational scheme capable of extreme scalability. Accuracy is controlled by the mesh spacing of the finite difference discretization, the size of the localization regions in which the electronic wave functions are confined, and a cutoff beyond which the components of the overlap matrix can be omitted when computing selected elements of its inverse. We demonstrate the algorithm's excellent parallel scaling for up to 101 952 atoms on 23 328 processors, with a wall-clock time of the order of 1 min per molecular dynamics time step and numerical error on the forces of less than 7x10 -4 Ha/Bohr.

36 MATERIALS SCIENCE↗

Solving large sparse eigenvalue problems on supercomputers

An important problem in scientific computing consists in finding a few eigenvalues and corresponding eigenvectors of a very large and sparse matrix. The most popular methods to solve these problems are based on projection techniques on appropriate subspaces. The main attraction of these methods is that they only require the use of the matrix in the form of matrix by vector multiplications. The implementations on supercomputers of two such methods for symmetric matrices, namely Lanczos' method and Davidson's method are compared. Since one of the most important operations in these two methods is the multiplication of vectors by the sparse matrix, methods of performing this operation efficiently are discussed. The advantages and the disadvantages of each method are compared and implementation aspects are discussed. Numerical experiments on a one processor CRAY 2 and CRAY X-MP are reported. Possible parallel implementations are also discussed.

Philippe, Bernard↗

Robot acting on moving bodies (RAMBO): Preliminary results

A robot system called RAMBO is being developed. It is equipped with a camera, which, given a sequence of simple tasks, can perform these tasks on a moving object. RAMBO is given a complete geometric model of the object. A low level vision module extracts and groups characteristic features in images of the object. The positions of the object are determined in a sequence of images, and a motion estimate of the object is obtained. This motion estimate is used to plan trajectories of the robot tool to relative locations nearby the object sufficient for achieving the tasks. More specifically, low level vision uses parallel algorithms for image enchancement by symmetric nearest neighbor filtering, edge detection by local gradient operators, and corner extraction by sector filtering. The object pose estimation is a Hough transform method accumulating position hypotheses obtained by matching triples of image features (corners) to triples of model features. To maximize computing speed, the estimate of the position in space of a triple of features is obtained by decomposing its perspective view into a product of rotations and a scaled orthographic projection. This allows the use of 2-D lookup tables at each stage of the decomposition. The position hypotheses for each possible match of model feature triples and image feature triples are calculated in parallel. Trajectory planning combines heuristic and dynamic programming techniques. Then trajectories are created using parametric cubic splines between initial and goal trajectories. All the parallel algorithms run on a Connection Machine CM-2 with 16K processors.

Davis, Larry S.↗

Parallel Architecture For Robotics Computation

Universal Real-Time Robotic Controller and Simulator (URRCS) is highly parallel computing architecture for control and simulation of robot motion. Result of extensive algorithmic study of different kinematic and dynamic computational problems arising in control and simulation of robot motion. Study led to development of class of efficient parallel algorithms for these problems. Represents algorithmically specialized architecture, in sense capable of exploiting common properties of this class of parallel algorithms. System with both MIMD and SIMD capabilities. Regarded as processor attached to bus of external host processor, as part of bus memory.

Fijany, Amir↗

Robot Acting on Moving Bodies (RAMBO): Interaction with tumbling objects

Interaction with tumbling objects will become more common as human activities in space expand. Attempting to interact with a large complex object translating and rotating in space, a human operator using only his visual and mental capacities may not be able to estimate the object motion, plan actions or control those actions. A robot system (RAMBO) equipped with a camera, which, given a sequence of simple tasks, can perform these tasks on a tumbling object, is being developed. RAMBO is given a complete geometric model of the object. A low level vision module extracts and groups characteristic features in images of the object. The positions of the object are determined in a sequence of images, and a motion estimate of the object is obtained. This motion estimate is used to plan trajectories of the robot tool to relative locations rearby the object sufficient for achieving the tasks. More specifically, low level vision uses parallel algorithms for image enhancement by symmetric nearest neighbor filtering, edge detection by local gradient operators, and corner extraction by sector filtering. The object pose estimation is a Hough transform method accumulating position hypotheses obtained by matching triples of image features (corners) to triples of model features. To maximize computing speed, the estimate of the position in space of a triple of features is obtained by decomposing its perspective view into a product of rotations and a scaled orthographic projection. This allows use of 2-D lookup tables at each stage of the decomposition. The position hypotheses for each possible match of model feature triples and image feature triples are calculated in parallel. Trajectory planning combines heuristic and dynamic programming techniques. Then trajectories are created using dynamic interpolations between initial and goal trajectories. All the parallel algorithms run on a Connection Machine CM-2 with 16K processors.

Davis, Larry S.↗

Efficient computation of aerodynamic influence coefficients for aeroelastic analysis on a transputer network

Aeroelastic analysis is multi-disciplinary and computationally expensive. Hence, it can greatly benefit from parallel processing. As part of an effort to develop an aeroelastic capability on a distributed memory transputer network, a parallel algorithm for the computation of aerodynamic influence coefficients is implemented on a network of 32 transputers. The aerodynamic influence coefficients are calculated using a 3-D unsteady aerodynamic model and a parallel discretization. Efficiencies up to 85 percent were demonstrated using 32 processors. The effect of subtask ordering, problem size, and network topology are presented. A comparison to results on a shared memory computer indicates that higher speedup is achieved on the distributed memory system.

Janetzke, David C.↗