SEARCH · Search NASA

Results for “parallel data structures”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 163 records · Page 9

All sources NASA NTRS DOE OSTI

NASA NTRS2011

First Lunar Wake Passage of ARTEMIS: Discrimination of Wake Effects and Solar Wind Fluctuations by 3D Hybrid Simulations

The spacecraft P1 of the new ARTEMIS (Acceleration, Reconnection, Turbulence, and Electrodynamics of the Moon's Interaction with the Sun) mission passed the lunar wake for the first time on February 13, 2010. We present magnetic field and plasma data of this event and results of 3D hybrid simulations. As the solar wind magnetic field was highly dynamic during the passage, a simulation with stationary solar wind input cannot distinguish whether distortions were caused by these solar wind variations or by the lunar wake; therefore, a dynamic real-time simulation of the flyby has been performed. The input values of this simulation are taken from NASA OMNI data and adapted to the P1 data, resulting in a good agreement between simulation and measurements. Combined with the stationary simulation showing non-transient lunar wake structures, a separation of solar wind and wake effects is achieved. An anisotropy in the magnitude of the plasma bulk flow velocity caused by a non-vanishing magnetic field component parallel to the solar wind flow and perturbations created by counterstreaming ions in the lunar wake are observed in data and simulations. The simulations help to interpret the data granting us the opportunity to examine the entire lunar plasma environment and, thus, extending the possibilities of measurements alone: A comparison of a simulation cross section to theoretical predictions of MHD wave propagation shows that all three basic MHD modes are present in the lunar wake and that their expansion governs the lunar wake refilling process.

Wiehle, S.↗

NASA NTRSDate not supplied

Sparse Linear Algebra Toolkit for Computational Aerodynamics

Finding solutions to sparse linear systems of equations is an essential step in Computational Engineering applications of interest to NASA. Linear systems of equations are composed and solved in almost every computational engineering application. The characteristics of linear systems vary greatly from one application to another. Accordingly, there are a wide variety of methods for the solution of linear systems of equations. The operations and methods prepared by the authors are focused on linear systems of interest to NASA, primarily those associated with Computational Fluid Dynamics (CFD), Aeroelasticity, and Aeroacoustics. The Sparse Linear Algebra Toolkit (SLAT) is a coordinated collection of software featuring operations, methods, and data structures that are useful when solving sparse linear systems of equations on modern computer architectures. The implemented operations and methods are designed and tuned for parallelism in shared memory, in distributed memory, and across the hybrid combination of distributed-shared memory. The toolkit includes novel methods and implementations for modern architectures and facilitates development of new approaches for meeting NASA’s evolving computational engineering challenges using evolving computer architectures that are not available in vendor libraries. In this paper, significant features and interfaces within SLAT are presented and verified for simulations performed with NASA’s CFD solver, FUN3D. The runtime and scaling performance of the Generalized Minimum Residual (GMRES) method implemented in SLAT is analyzed for the linear subproblems within the solution of turbulent Navier-Stokes equations employed in the simulation of high-lift configurations. Prior to this work, the SPARSKIT GMRES implementation was the only Krylov subspace method available within FUN3D. A strong scaling study shows the SLAT GMRES implementation facilitates accurate Reynolds-averaged Navier-Stokes CFD solutions between 15% and 56% faster than the SPARSKIT GMRES implementation.

Stephen L Wood↗

NASA NTRS2001

Embedded Coplanar Strips Traveling-Wave Photomixers

The electric field distribution in photomixers with electrodes deposited on the surface has already been calculated. It was shown that the strength of the electric field diminishes rapidly with depth. It was argued that the resulting reduction of the effective interaction volume of the device lowers the optical-to-heterodyne conversion. In this paper, we will present the results of our investigation on the influence of the electrode placement on the performance of photomixers. We have fabricated and measured traveling-wave photomixer devices which have both embedded and surface electrodes - the nominal spacing between the electrodes was 2 micrometers. Devices were made using either low-temperature-grown (LTG)-GaAs or ErAs:GaAs as the photoconductive material. The dark current, photocurrent, and radio frequency (RF) emission were measured at nominally 1 THz. The experimental data show a surprising difference in the behavior of ErAs:GaAs devices when the electrodes are embedded. A factor of two increase in RF radiation is observed for electric fields < 20 kV/cm. No such improvement was observed for the LTG-GaAs devices. We argue that the distinctive behavior of the two photoconductive materials is due to differences in the crystal structure - LTG-GaAs is isotropic, while ErAs:GaAs is uniaxial. We find that the carrier mobility in-plane (parallel) to the ErAs layers in the ErAs:GaAs superlattice is larger than orthogonal to these layers. The data indicate that carrier velocity overshoot is responsible for the excess radiation produced for the embedded electrode ErAs:GaAs devices.

Wyss, R. A.↗

NASA NTRS2016

An Integrated Circuit for Radio Astronomy Correlators Supporting Large Arrays of Antennas

Radio telescopes that employ arrays of many antennas are in operation, and ever larger ones are being designed and proposed. Signals from the antennas are combined by cross-correlation. While the cost of most components of the telescope is proportional to the number of antennas N, the cost and power consumption of cross-correlationare proportional to N2 and dominate at sufficiently large N. Here we report the design of an integrated circuit (IC) that performs digital cross-correlations for arbitrarily many antennas in a power-efficient way. It uses an intrinsically low-power architecture in which the movement of data between devices is minimized. In a large system, each IC performs correlations for all pairs of antennas but for a portion of the telescope's bandwidth (the so-called "FX" structure). In our design, the correlations are performed in an array of 4096 complex multiply-accumulate (CMAC) units. This is sufficient to perform all correlations in parallel for 64 signals (N=32 antennas with 2 opposite-polarization signals per antenna). When N is larger, the input data are buffered in an on-chipmemory and the CMACs are re-used as many times as needed to compute all correlations. The design has been synthesized and simulated so as to obtain accurate estimates of the IC's size and power consumption. It isintended for fabrication in a 32 nm silicon-on-insulator process, where it will require less than 12mm2 of silicon area and achieve an energy efficiency of 1.76 to 3.3 pJ per CMAC operation, depending on the number of antennas. Operation has been analyzed in detail up to N = 4096. The system-level energy efficiency, including board-levelI/O, power supplies, and controls, is expected to be 5 to 7 pJ per CMAC operation. Existing correlators for the JVLA (N = 32) and ALMA (N = 64) telescopes achieve about 5000 pJ and 1000 pJ respectively usingapplication-specific ICs in older technologies. To our knowledge, the largest-N existing correlator is LEDA atN = 256; it uses GPUs built in 28 nm technology and achieves about 1000 pJ. Correlators being designed for the SKA telescopes (N = 128 and N = 512) using FPGAs in 16nm technology are predicted to achieve about 100 pJ.

ASIC↗

NASA NTRS1984

Parametric analysis of synthetic aperture radar data acquired over truck garden vegetation

An airborne X-band SAR acquired multipolarization and multiflight pass SAR images over a truck garden vegetation area. Based on a variety of land cover and row crop direction variations, the vertical (VV) polarization data contain the highest contrast, while cross polarization contains the least. When the radar flight path is parallel to the row direction, both horizontal (HH) and VV polarization data contain very high return which masks out the specific land cover that forms the row structure. Cross polarization data are not that sensitive to row orientation. The inclusion of like and cross polarization data help delineate special surface features (e.g., row crop against non-row-oriented land cover, very-rough-surface against highly row-oriented surface).

Wu, S. T.↗

NASA NTRS1992

Optical distributed sensors for feedback control: Characterization of photorefractive resonator

The aim of the project was to explore, define, and assess the possibilities of optical distributed sensing for feedback control. This type of sensor, which may have some impacts in the dynamic control of deformable structures and the monitoring of small displacements, can be divided into data acquisition, data processing, and control design. Analogue optical techniques, because they are noninvasive and afford massive parallelism may play a significant role in the acquisition and the preprocessing of the data for such a sensor. Assessing these possibilities was the aim of the first stage of this project. The scope of the proposed research was limited to: (1) the characterization of photorefractive resonators and the assessment of their possible use as a distributed optical processing element; and (2) the design of a control system utilizing signals from distributed sensors. The results include a numerical and experimental study of the resonator below threshold, an experimental study of the effect of the resonator's transverse confinement on its dynamics above threshold, a numerical study of the resonator above threshold using a modal expansion approach, and the experimental test of this model. A detailed account of each investigation, including methodology and analysis of the results are also included along with reprints of published and submitted papers.

Indebetouw, Guy↗

NASA NTRSDate not supplied

Performance Optimization Methods for a Memory-Bound, Unstructured-Grid CFD Application on Massively Parallel GPU Platforms

Computational performance of the FUN3D unstructured-grid computational fluid dynamics (CFD) application on massively parallel GPU environments is memory-bound and highly dependent upon efficient reads from and atomic updates to the irregular cell-, edge-, and node-based data structures. In this talk, we present recent efforts into optimizing select performance-critical kernels on NVIDIA Tesla V100 and A100 GPUs and AMD CDNA MI100 GPUs. A novel use of L2 cache residency controls and asynchronous loads into on-chip shared memory are explored on the A100 GPU for the sparse iterative solver, which is dominated by mixed-precision, sparse matrix vector multiplication. Demonstrations show that these methods improve global memory bandwidth utilization by 13.5% on the A100 GPU. Several techniques are also presented that use registers and/or shared memory to facilitate array transposition and aggregation which combine to reduce the frequency and increase the cache efficiency of floating-point atomic updates to the irregular data structures. These methods are demonstrated to improve the kernel throughput by nearly 500% on select kernels on the AMD MI100 over atomic updates directly to global memory. Overall, both V100 and A100 GPUs outperformed the MI100 GPU on kernels dominated by double-precision atomic updates; however, the techniques demonstrated here reduced the performance gap and improved the MI100 performance.

GPU CPU unstructured CFD memory↗

NASA NTRS1981

The SMART MIL-STD-1553 bus adapter hardware manual

The SMART Multiplexer Interface Adapter, (SMIA) a complete system interface for message structure of the MIL-STD-1553, is described. It provides buffering and storage for transmitted and received data and handles all the necessary handshaking to interface between parallel 8-bit data bus and a MIL-STD serial bit stream. The bus adapter is configured as either a bus controller of a remote terminal interface. It is coupled directly to the multiplex bus, or stub coupled through an additional isolation transformer located at the connection point. Fault isolation resistors provide short circuit protection.

Ton, T. T.↗

NASA NTRS1982

Conversion and matched filter approximations for serial minimum-shift keyed modulation

Serial minimum-shift keyed (MSK) modulation, a technique for generating and detecting MSK using series filtering, is ideally suited for high data rate applications provided the required conversion and matched filters can be closely approximated. Low-pass implementations of these filters as parallel inphase- and quadrature-mixer structures are characterized in this paper in terms of signal-to-noise ratio (SNR) degradation from ideal and envelope deviation. Several hardware implementation techniques utilizing microwave devices or lumped elements are presented. Optimization of parameter values results in realizations whose SNR degradation is less than 0.5 dB at error probabilities of .000001.

Ziemer, R. E.↗

NASA NTRS2006

Aluminian Low-Ca Pyroxene in a Ca-Al-rich Chondrule from the Semarkona Meteorite

A Ca-AI-rich chondrule (labeled G7) from the Semarkona LL3.0 ordinary chondrite (OC) consists of 73 vol% glassy mesostasis, 22 vol% skeletal forsterite. 3 vol% fassaite (i.e., Al-Ti diopside), and 2 vol% Al-rich, low-Ca pyroxene. The latter phase, which contains up to 16.3 wt% A1203, is among the most AI-rich, low-Ca pyroxene grains ever reported. It is inferred that 20% of the tetrahedral sites and 13% of the octahedral sites in this grain are occupied by Al. Approximately parallel optical extinction implies that the Al-rich, low-Ca pyroxene grains are probably orthorhombic, consistent with literature data that show that A1203 stabilizes the orthoenstatite structure relative to protoenstatite at low pressure. The order of crystallization in the chondrule was forsterite, AI-rich low-Ca pyroxene, and fassaite; the residual liquid vitrified during chondrule quenching. Phase relationships indicate that, for a G7-composition liquid at equilibrium, spinel and anorthite should crystallize early and orthopyroxene should not crystallize at all. The presence of AI-rich orthopyroxene in G7 is due mainly to the kinetic failure of anorthite to crystallize; this failure was caused by quenching of the G7 precursor droplet. Aluminum preferentially enters the relatively large B tetrahedra of orthopyroxene; because only one tetrahedral size occurs in fassaite, this phase contains higher mean concentrations of Al2O3 than the Al-rich orthopyroxene (17.8 and 14.7 wt%, respectively). Chondrule G7 may have formed by remelting an amoeboid olivine inclusion that entered the OC region of the solar nebula during an episode of chondrule formation.

Rubin, Alan E.↗

NASA NTRS2000

Functionality Improvements to Overaero

The functionality of the overset, static aeroelasticity, Navier-Stokes flow solver OVERAERO was increased by adding capability to the flow solver and enhancing code performance. Improvements were made to the fluids/structure interface, an MLP version of the parallel OVERAERO code was developed, and the OVERAERO-MPI code was ported to the Cray T3E. The OVERFLOW-MPI and OVERAERO-MPI codes were tested successfully on the IPG testbed and a means of reducing communication overhead within OVERFLOW-MPI was investigated. To solve an aeroelastic problem computationally, a structures grid surface definition and a fluids grid surface definition are required. Typically, the structures grid surface has a lower fidelity than the fluids grid surface. Thus, the methods developed to transfer data between the two grid systems are vital to the accuracy and efficiency of the aeroelasticity code. The fluids/structures interface developed for the OVERAERO code was improved to more accurately treat fluids surfaces that bridge between two different structural surfaces. For example, the method allowed the forward portion of a flap track fairing to deform with the wing and the aft end of the fairing to deform with the flap. A tightly-coupled version of the code based on OVERFLOW-MLP was developed to improve code performance on the SGI Origin 2000. This required a new parallelization strategy to couple the fluids and structures codes. The OVERAERO-MPI code was ported to the Cray T3E to extend the usability of the code. The port required extensive use of dynamic memory management techniques to fit large problems within the memory limitations of the T3E. The OVERFLOW-MPI and OVERAERO-MPI codes were tested on the IPG testbed being developed within NASA. For small problems with minimal data transfer between grids, there was little to no performance penalty spreading the computation across two machines. For very large problems, methods were developed to minimize intermachine communication via the grid partitioning scheme. By minimizing the intermachine communication requirements of the problem, it may still be beneficial to run a tightly-coupled flow solver across two machines within the IPG.

Gee, Ken↗

NASA NTRS1989

Parallel asynchronous systems and image processing algorithms

A new hardware approach to implementation of image processing algorithms is described. The approach is based on silicon devices which would permit an independent analog processing channel to be dedicated to evey pixel. A laminar architecture consisting of a stack of planar arrays of the device would form a two-dimensional array processor with a 2-D array of inputs located directly behind a focal plane detector array. A 2-D image data stream would propagate in neuronlike asynchronous pulse coded form through the laminar processor. Such systems would integrate image acquisition and image processing. Acquisition and processing would be performed concurrently as in natural vision systems. The research is aimed at implementation of algorithms, such as the intensity dependent summation algorithm and pyramid processing structures, which are motivated by the operation of natural vision systems. Implementation of natural vision algorithms would benefit from the use of neuronlike information coding and the laminar, 2-D parallel, vision system type architecture. Besides providing a neural network framework for implementation of natural vision algorithms, a 2-D parallel approach could eliminate the serial bottleneck of conventional processing systems. Conversion to serial format would occur only after raw intensity data has been substantially processed. An interesting challenge arises from the fact that the mathematical formulation of natural vision algorithms does not specify the means of implementation, so that hardware implementation poses intriguing questions involving vision science.

Coon, D. D.↗

NASA NTRS1974

The global characteristics of atmosphere emissions in the lower thermosphere and their aeronomic implications

The green line of atomic oxygen and the Herzberg bands of molecular oxygen as observed from the OGO-4 airglow photometer are discussed in terms of their spatial and temporal distributions and their relation to the atomic oxygen content in the lower thermosphere. Daily maps of the distribution of emissions show considerable structure (cells, patches, and bands) with appreciable daily changes. When data are averaged over periods of several days in length, the resulting patterns have occasional tendencies to follow geomagnetic parallels. The Seasonal variations are characterized by maxima in both the Northern and Southern Hemispheres in October, with the Northern Hemisphere having substantially higher emission rates. Formulae are derived relating the vertical column emission rates of the green line and the Herzberg bands to the atomic oxygen peak density. Global averages for the time period for these data (August 1967 to January 1968), when converted to maximum atomic oxygen densities near 95 km, have a range of 2.0 x 10 to the 11th power/cu cm 2.7 x 10 to the 11th power/cu cm.

Reed, E. I.↗

NASA NTRS1982

Description and planned use of a data distribution evaluation system for fiber optic data buses

A general description of a laboratory data distribution evaluation system (DDES) is given and some plans to use the system. The DDES is a microprocessor-based evaluation system consisting of three identical terminals. The DDES provides the capability for evaluating different system protocols and data word structures by making appropriate software changes, and different transmission medias and modulation schemes by making front-end hardware changes. A data multiplexing standard for fiber optic data transmission which parallels MIL-STD-1553B has proposed four different data modulation schemes. These modulation schemes will be described along with the plans to evaluate each of them. NASA-Langley has also been developing a data distribution approach which utilizes optical wavelength division multiplexing (WDM). A four-port, four-wavelength WDM system will be described along with the planned use of the DDES to evaluate its performance.

Spencer, J. L.↗

NASA NTRS1975

Study of relation between Pc 3 micropulsations and magnetosheath fluctuations, and multisatellite investigation of earth's bow shock

The validity is investigated of a suggested model according to which Pc 3 and/or Pc 4 micropulsations are excited by magnetosheath field (and plasma) fluctuations arising in the quasi-parallel structure of the subsolar bow shock. The influence of solar wind plasma parameters on local shock structure and on the configuration of the entire bow shock system is included. Simultaneous data from two or more spacecraft and from multiple diagnostics is used to evaluate the geometrical factor, field-to-shock normal angle, or its B-X equivalent, and the principal plasma parameters. Results are presented and discussed.

Greenstadt, E. W.↗

NASA NTRS1985

Experiences with Cray multi-tasking

The issues involved in modifying an existing code for multitasking is explored. They include Cray extensions to FORTRAN, an examination of the application code under study, designing workable modifications, specific code modifications to the VAX and Cray versions, performance, and efficiency results. The finished product is a faster, fully synchronous, parallel version of the original program. A production program is partitioned by hand to run on two CPUs. Loop splitting multitasks three key subroutines. Simply dividing subroutine data and control structure down the middle of a subroutine is not safe. Simple division produces results that are inconsistent with uniprocessor runs. The safest way to partition the code is to transfer one block of loops at a time and check the results of each on a test case. Other issues include debugging and performance. Task startup and maintenance (e.g., synchronization) are potentially expensive.

Miya, E. N.↗

NASA NTRS1989

Characterization of robotics parallel algorithms and mapping onto a reconfigurable SIMD machine

The kinematics, dynamics, Jacobian, and their corresponding inverse computations are six essential problems in the control of robot manipulators. Efficient parallel algorithms for these computations are discussed and analyzed. Their characteristics are identified and a scheme on the mapping of these algorithms to a reconfigurable parallel architecture is presented. Based on the characteristics including type of parallelism, degree of parallelism, uniformity of the operations, fundamental operations, data dependencies, and communication requirement, it is shown that most of the algorithms for robotic computations possess highly regular properties and some common structures, especially the linear recursive structure. Moreover, they are well-suited to be implemented on a single-instruction-stream multiple-data-stream (SIMD) computer with reconfigurable interconnection network. The model of a reconfigurable dual network SIMD machine with internal direct feedback is introduced. A systematic procedure internal direct feedback is introduced. A systematic procedure to map these computations to the proposed machine is presented. A new scheduling problem for SIMD machines is investigated and a heuristic algorithm, called neighborhood scheduling, that reorders the processing sequence of subtasks to reduce the communication time is described. Mapping results of a benchmark algorithm are illustrated and discussed.

Lee, C. S. G.↗

NASA NTRS1998

Analysis of a Memory Device Failure

The recent failure of a vintage memory device presented a unique challenge to failure analysts. Normally device layouts, fabrication parameters and other technical information were available to assist the analyst in the analysis. However, this device was out of production for many years and the manufacturer was no longer in business, so the information was not available. To further complicate this analysis, the package leads were all but removed making additional electrical testing difficult. Under these conditions, new and innovative methods were used to analyze the failure. The external visual exam, radiography, PIND, and leak testing were performed with nominal results. Since electrical testing was precluded by the short lead lengths, the device was delidded to expose the internal structures for microscopic examination. No failure mechanism was identified. The available electrical data suggested an ESD or low level EOS type mechanism which left no visible surface damage. Due to parallel electrical paths, electrical probing on the chip failed to locate the failure site. Two non-destructive Scanning Electron Microscopy techniques, CIVA (Charge Induced Voltage Alteration) and EBIC (Electron Beam Induced Current), and a liquid crystal decoration technique which detects localized heating were employed to aid in the analysis. CIVA and EBIC isolated two faults in the input circuitry, and the liquid crystal technique further localized two hot spots in regions on two input gates. Removal of the glassivation and metallization revealed multiple failure sites located in the gate oxide of two input transistors suggesting machine (testing) induced damage.

Nicolas, David P.↗