Search NASA⌕ Search

SEARCH · Search NASA

Results for “parallel implementation”

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 1,243 records · Page 69
All sourcesNASA NTRSDOE OSTI
NASA NTRSDate not supplied

GPU Implementation of the OVERFLOW CFD Code

The high-performance computing (HPC) landscape is quickly changing to systems where most of the performance comes from specialized chips, specifically graphics processing units (GPUs). Such GPU systems are throughput machines, where efficient use of the GPU often requires code refactoring to expose a few orders of magnitude more fine grain parallelism than was previously used on the CPU. Recent modifications to OVERFLOW, an overset, structured grid, computational fluid dynamics flow solver, written in Fortran will be presented. These modifications include both code modernization efforts and algorithmic changes to enable OVERFLOW to efficiently utilize GPUs. Many of these algorithmic changes would likely also be applicable for other structured grid, stencil-based codes wanting to utilize GPUs. The capabilities that have been ported to run on the GPUs are presented, along with the performance gains of the GPU version relative the CPU version of OVERFLOW.

GPU Programming↗
NASA NTRSDate not supplied

GPU Implementation of the OVERFLOW CFD Code

The high-performance computing (HPC) landscape is quickly changing to systems where most of the performance comes from specialized chips, specifically graphics processing units (GPUs). Such GPU systems are throughput machines, where efficient use of the GPU often requires code refactoring to expose a few orders of magnitude more fine grain parallelism than was previously used on the CPU. Recent modifications to OVERFLOW, an overset, structured grid, computational fluid dynamics flow solver, written in Fortran will be presented. These modifications include both code modernization efforts and algorithmic changes to enable OVERFLOW to efficiently utilize GPUs. Many of these algorithmic changes would likely also be applicable for other structured grid, stencil-based codes wanting to utilize GPUs. The capabilities that have been ported to run on the GPUs are presented, along with the performance gains of the GPU version relative the CPU version of OVERFLOW.

GPU Programming↗
NASA NTRS1975

The potential of multi-port optical memories in digital computing

A high-capacity memory with a relatively high data transfer rate and multi-port simultaneous access capability may serve as the basis for new computer architectures. The implementation of a multi-port optical memory is discussed. Several computer structures are presented that might profitably use such a memory. These structures include (1) a simultaneous record access system, (2) a simultaneously shared memory computer system, and (3) a parallel digital processing structure.

Alford, C. O.↗
NASA NTRS1979

Issues in the digital implementation of control compensators

Techniques developed for the finite-precision implementation of digital filters were used, adapted, and extended for digital feedback compensators, with particular emphasis on steady state, linear-quadratic-Gaussian compensators. Topics covered include: (1) the linear-quadratic-Gaussian problem; (2) compensator structures; (3) architectural issues: serialism, parallelism, and pipelining; (4) finite wordlength effects: quantization noise, quantizing the coefficients, and limit cycles; and (5) the optimization of structures.

Moroney, P.↗
NASA NTRS1987

Electronic hardware implementations of neutral networks

This paper examines some of the present work on the development of electronic neural network hardware. In particular, the investigations currently under way at JPL on neural network hardware implementations based on custom VLSI technology, novel thin film materials, and an analog-digital hybrid architecture are reviewed. The availability of such hardware will greatly benefit and enhance the present intense research effort on the potential computational capabilities of highly parallel systems based on neural network models.

Thakoor, A. P.↗
NASA NTRS1987

Distributed control architecture for real-time telerobotic operation

The emerging field of telerobotics places new demands on control system architecture to allow both autonomous operations and natural human-machine interfacing. The feasibility of multiprocessor systems performing parallel control computations is realizable. A practical distribution of control processors is presented and the issues involved in the realization of this architecture are discussed. A prototype dual axis controller based on the NOVIX computer is described, and results of its implementation are discussed. Application of this type of control system to a replicated, redundant manipulator system is also described.

Martin, H. L.↗
NASA NTRS2016

Near-Body Grid Adaption for Overset Grids

A solution adaption capability for curvilinear near-body grids has been implemented in the OVERFLOW overset grid computational fluid dynamics code. The approach follows closely that used for the Cartesian off-body grids, but inserts refined grids in the computational space of original near-body grids. Refined curvilinear grids are generated using parametric cubic interpolation, with one-sided biasing based on curvature and stretching ratio of the original grid. Sensor functions, grid marking, and solution interpolation tasks are implemented in the same fashion as for off-body grids. A goal-oriented procedure, based on largest error first, is included for controlling growth rate and maximum size of the adapted grid system. The adaption process is almost entirely parallelized using MPI, resulting in a capability suitable for viscous, moving body simulations. Two- and three-dimensional examples are presented.

Buning, Pieter G.↗
NASA NTRS1995

A Multi-Mission Space Avionics Architecture

As the budget for the space industry is dwindling, a reusable avionics architecture applicable across multiple missions is urgently needed to reduce the development and production costs of flight projects. This paper presents a multi-mission avionics architecture which employs interface standards extensively, sot hat avionics systems can be built rapidly by assembling subsystems and instruments in a plug-and-play manner. The key feature of this architecture is a "backbone" standard parallel bus which can accommodate various standard interfaces through a repertoire of I/O modules. Hence, specific ;mission requirements can be met by selecting and integrating the appropriate subsystems and instruments into the system. This architecture can also adopt different "backbone" buses, implement redundant processors, and support distributed processing. An example of the architecture using the MOPS6000 from the Southwest Research Institute is also described. Experiments will be done in the Flight System Testbed (FST) of the Jet Propulsion Laboratory to benchmark the MOPS6000 and to demonstrate its multi-mission capability.

Avionics↗
NASA NTRS1993

Implementation of a mesh adaptive scheme based on an element-level error indicator

We investigate the formulation and application of element-level error indicators based on parametrized variational principles. The qualifier 'element-level' means that no information from adjacent elements is used for error estimation. This property is ideally suited to drive adaptive mesh refinement on parallel computers where access to neighboring elements resident on different processors may incur significant computational overhead. Furthermore, such indicators are not affected by physical jumps at junctures or interfaces. An element-level indicator has been derived from the higher-order element energy and applied to r and h mesh adaptation of meshes in plates and shell structures. We report on our initial experiments with a cylindrical shell that intersects with fist plates forming a simplified 'wing-body intersection' benchmark problem.

Keating, Scott↗
NASA NTRS1989

Spectral element simulations of laminar and turbulent flows in complex geometries

Spectral element methods are high-order weighted residual techniques based on spectral expansions of variables and geometry for the Navier-Stokes (NS) and transport equations. Here, practical aspects of these methods and their efficient implementation are examined, and several examples of flows in truly complex geometries are presented. The spectral element discretization for NS equations is introduced, and the convergence of the method is addressed. An efficient data management scheme is discussed in the context of parallel processing computations. The method is validated by comparing the spectral element solutions with the exact eigensolutions for the Orr-Sommerfeld equations in two and three dimensions. Computer-aided flow visualizations are presented for an impulsive flow past a sharp edge wedge. Three-dimensional states of channel flow disrupted by an array of cylindrical eddy promoters are studied, and the results of a direct simulation of the turbulent flow in a plane channel are presented.

Karniadakis, George EM↗
NASA NTRS1992

Users manual for program SSFREQ intermediate mode stability curves: Developed for use on a PC computer

The piping in a liquid rocket can assume complex configurations due to multiple tanks, multiple engines, and structures that must be piped around. The capability to handle some of these complex configurations have been incorporated into the SSFREQ code. The capability to modify the input on line has been implemented. The configurations allowed include multiple tanks, multiple engines, the splitting of a pipe into equal segments going to different (or the same) engines. This program will handle the following type elements: straight pipes, bends, inline accumulators, tuned stub accumulators, Helmholtz resonators, parallel resonators, pumps, split pipes, multiple tanks, and multiple engines.

Armstrong, Wilbur C.↗
NASA NTRS1992

Users manual for program ADMIT: Admittance and pressure transfer function developed for use on a PC computer

The piping in a liquid rocket can assume complex configurations due to multiple tanks, multiple engines, and structures that must be piped around. The capability to handle some of these complex configurations have been incorporated into the ADMIT code. The capability to modify the input on line has been implemented. The configurations allowed include multiple tanks, multiple engines, the splitting of a pipe into unequal segments going to different (or the same) engines. This program will handle the following type elements: straight pipes, bends, inline accumulators, tuned stub accumulators, Helmholtz resonators, parallel resonators, pumps, split pipes, multiple tanks, and multiple engines.

Armstrong, Wilbur C.↗
NASA NTRS2004

An Expert System for the Development of Efficient Parallel Code

We have built the prototype of an expert system to assist the user in the development of efficient parallel code. The system was integrated into the parallel programming environment that is currently being developed at NASA Ames. The expert system interfaces to tools for automatic parallelization and performance analysis. It uses static program structure information and performance data in order to automatically determine causes of poor performance and to make suggestions for improvements. In this paper we give an overview of our programming environment, describe the prototype implementation of our expert system, and demonstrate its usefulness with several case studies.

Jost, Gabriele↗
NASA NTRS2007

An Autonomous Flight Safety System

The Autonomous Flight Safety System (AFSS) being developed by NASA s Goddard Space Flight Center s Wallops Flight Facility and Kennedy Space Center has completed two successful developmental flights and is preparing for a third. AFSS has been demonstrated to be a viable architecture for implementation of a completely vehicle based system capable of protecting life and property in event of an errant vehicle by terminating the flight or initiating other actions. It is capable of replacing current human-in-the-loop systems or acting in parallel with them. AFSS is configured prior to flight in accordance with a specific rule set agreed upon by the range safety authority and the user to protect the public and assure mission success. This paper discusses the motivation for the project, describes the method of development, and presents an overview of the evolving architecture and the current status.

Bull, James B.↗
NASA NTRS2014

Structural Characterization of Lateral-grown 6H-SiC am-plane Seed Crystals by Hot Wall CVD Epitaxy

The performance of commercially available silicon carbide (SiC) power devices is limited due to inherently high density of screw dislocations (SD), which are necessary for maintaining polytype during boule growth and commercially viable growth rates. The NASA Glenn Research Center (GRC) has recently proposed a new bulk growth process based on axial fiber growth (parallel to the c-axis) followed by lateral expansion (perpendicular to the c-axis) for producing multi-faceted m-plane SiC boules that can potentially produce wafers with as few as one SD per wafer. In order to implement this novel growth technique, the lateral homoepitaxial growth expansion of a SiC fiber without introducing a significant number of additional defects is critical. Lateral expansion is being investigated by hot wall chemical vapor deposition (HWCVD) growth of 6H-SiC am-plane seed crystals (0.8mm x 0.5mm x 15mm) designed to replicate axially grown SiC single crystal fibers. The post-growth crystals exhibit hexagonal morphology with approximately 1500 m (1.5 mm) of total lateral expansion. Preliminary analysis by synchrotron white beam x-ray topography (SWBXT) confirms that the growth was homoepitaxial, matching the polytype of the respective underlying region of the seed crystal. Axial and transverse sections from the as grown crystal samples were characterized in detail by a combination of SWBXT, transmission electron microscopy (TEM) and Raman spectroscopy to map defect types and distribution. X-ray diffraction analysis indicates the seed crystal contained stacking disorders and this appears to have been reproduced in the lateral growth sections. Analysis of the relative intensity for folded transverse acoustic (FTA) and optical (FTO) modes on the Raman spectra indicate the existence of stacking faults. Further, the density of stacking faults is higher in the seed than in the grown crystal. Bundles of dislocations are observed propagating from the seed in m-axis lateral directions. Contrast extinction analysis of these dislocation lines reveals they are edge type basal plane dislocations that track the growth direction. Polytype phase transition and stacking faults were observed by high-resolution TEM (HRTEM), in agreement with SWBXT and Raman scattering.

Single Crystal↗
NASA NTRSDate not supplied

Radiation-Hard Parallel Readout Circuit for Low-Frequency Voltage Signal Measurements

NASA Goddard Space Flight Center (GSFC) has successfully developed and tested a custom-designed low-noise multi-channel digitizer (MCD) application specific integrated circuit (ASIC) for operation in harsh radiation environments. The MCD-ASIC is optimized for low-frequency and low-voltage signal measurements from sensors and transducers. It has 20 input channels where each channel is comprised of auto-zeroed chopper variable-gain amplifier, post amplifier, and a second order ∑∆ modulator. ∑∆ analog-to-digital converter (ADC) relies on oversampling and noise shaping to achieve high-resolution conversion. However, the MCD-ASIC requires digital filtering and decimation to convert the output single bit streams from the ADC to useful data words. A parallel digital platform such as a field-programmable-gate-array (FPGA) is highly suitable to fully leverage the capabilities of the MCD-ASIC. The FPGA controls the MCD-ASIC via serial peripheral interface (SPI) protocol and acquires data from it. A Python-script communicates with the FPGA board through a USB interface on a cross operating platform. Using this architecture, the system is capable of monitoring up to 20 voltage readout channels simultaneously in a real-time manner. Each channel’s parameters can be programmed independently allowing maximum user versatility. In this paper, we present analysis of the analog front-end, the implementation of the digital processing unit on the FPGA, and provide noise performance results from the MCD-ASIC readout.

ASIC↗
NASA NTRS1974

Prediction of compressor stall for distorted and undistorted flow by use of a multistage compressor simulation on the digital computer

A simulation technique is presented for the prediction of compressor stall for axial-flow compressors for clean and distorted inlet flow. The simulation is implemented on the digital computer and uses stage stacking and lumped-volume gas dynamics. The resulting nonlinear differential equations are linearized about a steady-state operating point, and a Routh-Hurwitz stability test is performed on the linear system matrix. Parallel compressor theory is utilized to extend the technique to the distorted inlet flow problem. The method is applied to the eight-stage J85-13 compressor.

Daniele, C. J.↗