SEARCH · Search NASA
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.
Six computers operate synchronously and are interconnected by three independent data buses. Processors control one subsystem. Some can control buses to transfer data at 1 megabit per second. Every 2.5 msec each processor examines list of things to do during next interval. This spacecraft control system could be adapted for controlling complex industrial processes.
Optical parallel processors have the potential for aiding the transfer of information over networks.
One-dimensional Fast Fourier Transform (FFT) operations work fastest on grids whose size is divisible by a power of two. Because of this, padding grids (that are not already sized to a power of two) so that their size is the next highest power of two can speed up operations. While this works well for one-dimensional grids, it does not work well for two-dimensional grids. For a two-dimensional grid, there are certain pad sizes that work better than others. Therefore, the need exists to generalize a strategy for determining optimal pad sizes. There are three steps in the FFT algorithm. The first is to perform a one-dimensional transform on each row in the grid. The second step is to transpose the resulting matrix. The third step is to perform a one-dimensional transform on each row in the resulting grid. Steps one and three both benefit from padding the row to the next highest power of two, but the second step needs a novel approach. An algorithm was developed that struck a balance between optimizing the grid pad size with prime factors that are small (which are optimal for one-dimensional operations), and with prime factors that are large (which are optimal for two-dimensional operations). This algorithm optimizes based on average run times, and is not fine-tuned for any specific application. It increases the amount of times that processor-requested data is found in the set-associative processor cache. Cache retrievals are 4-10 times faster than conventional memory retrievals. The tested implementation of the algorithm resulted in faster execution times on all platforms tested, but with varying sized grids. This is because various computer architectures process commands differently. The test grid was 512 512. Using a 540 540 grid on a Pentium V processor, the code ran 30 percent faster. On a PowerPC, a 256x256 grid worked best. A Core2Duo computer preferred either a 1040x1040 (15 percent faster) or a 1008x1008 (30 percent faster) grid. There are many industries that can benefit from this algorithm, including optics, image-processing, signal-processing, and engineering applications.
We present the description of a processor which has been implemented to generate map products starting from C-band interferometric data. The first stage of the processor consists of the conventional interferometric SAR processing producing a Digital Elevation Model (DEMs) and a SAR brightness image in sensor coordinates. In the second stage of processing, a land use classification map is obtained by using the DEM, brightness, and interferometric correlation layers. Auxiliary layers which include a drainage layer, a height gradient layer, a height error layer, an estimated penetration layer, and a shaded relief layer are also computed. In the final step, all UTM collocated layers are combined in a GIS system which allows for both hard copy map products and for digital applications.
The NASA/JPL airborne SAR (AIRSAR) system can collect cross track interferometric data for topographic mapping applications (TOPSAR) at both C- and L-Band.
Results are presented for a study conducted in order to characterize the quality and consistency of the SEASAT Synthetic Aperture Radar image data set. The study evaluated a large spectrum of parameters, including the resolution, position location accuracy, and the amplitude calibration of the imagery produced from the raw data by processors at several institutions. It is demonstrated that this data set can be used to obtain resolution up to the theoretical limit of 6 m x 21 m, absolute position location to within 200 m, and relative calibration of the image data to within + or - 2 dB.
In a computer having a large number of single instruction multiple data (SIMD) processors, each of the SIMD processors has two sets of three individual processor elements controlled by a master control unit and interconnected among a plurality of register file units where data is stored. The register files input and output data in synchronism with a minor cycle clock under control of two slave control units controlling the register file units connected to respective ones of the two sets of processor elements. Depending upon which ones of the register file units are enabled to store or transmit data during a particular minor clock cycle, the processor elements within an SIMD processor are connected in rings or in pipeline arrays, and may exchange data with the internal bus or with neighboring SIMD processors through interface units controlled by respective ones of the two slave control units.
While the primary function of the network in a parallel computer is to communicate data between processors, it is often useful if the network can also perform rudimentary calculations. That is, some simple processing ability in the network itself, particularly for performing parallel prefix computations, can reduce both the volume of data being communicated and the computational load on the processors proper. Unfortunately, typical implementations of such networks require a large fraction of the hardware budget, and so combining networks are viewed as being impractical. The FFP Machine has such a combining network, and various characteristics of the machine allow a good deal of simplification in the network design. Despite being simple in construction however, the network relies on many subtle details to work correctly. This paper describes an executable model of the network which will serve several purposes. It provides a complete and detailed description of the network which can substantiate its ability to support necessary functions. It provides an environment in which algorithms to be run on the network can be designed and debugged more easily than they would on physical hardware. Finally, it provides the foundation for exploring the design of the message receiving facility which connects the network to the individual processors.
Magnetospheric Multiscale (MMS) mission will study small-scale reconnection structures and their rapid motions from closely spaced platforms using instruments capable of high angular, energy, and time resolution measurements. To meet these requirements, the Fast Plasma Instrument (FPI) consists of eight (8) identical half top-hat electron sensors and eights (8) identical ion sensors and an Instrument Data Processing Unit (IDPU). The sensors (electron or ion) are grouped into pairs whose 6 deg x 180 deg fields-of-view (FOV) are set 90 deg apart. Each sensor is equipped with electrostatic aperture steering to allow the sensor to scan a 45 deg x 180 deg fan about its nominal viewing (0 deg deflection) direction. Each pair of sensors, known as the Dual Electron Spectrometer (DES) and the Dual Ion Spectrometer (DIS), occupies a quadrant on the MMS spacecraft and the combination of the eight electron/ion sensors, employing aperture steering, image the full-sky every 30-ms (electrons) and 150-ms (ions), respectively. To probe the results in the DES complement of a given spacecraft generating 6.5-Mbs(exp -1) of electron data while the DIS generates 1.1-Mbs(exp -1) of ion data yielding an FPI total data rate of 6.6-MBs(exp -1). The FPI electron/ion data is collected by the IDPU then transmitted to the Central Data Instrument Processor (CIDP) on the spacecraft for science interest ranking. Only data sequences that contain the greatest amount of temporal/spatial structure will be intelligently down-linked by the spacecraft. Currently, the FPI data rate allocation to the CIDP is 1.5-Mbs(exp -1). Consequently, the FPI-IDPU must employ data/image compression to meet this CIDP telemetry allocation. Here, we present simulations of the CCSDS 122.0-B-1 algorithm-based compression of the FPI-DES electron data. Compression analysis is based upon a seed of re-processed Cluster/PEACE electron measurements. Topics to be discussed include: review of compression algorithm; data quality; data formatting/organization; and, implications for data/matrix pruning. To conclude a presentation of the base-lined FPI data compression approach is provided.
The Magnetospheric Multiscale (MMS) mission will study small-scale reconnection structures and their rapid motions from closely spaced platforms using instruments capable of high angular, energy, and time resolution measurements. To meet these requirements, the Fast Plasma Instrument (FPI) consists of eight (8) identical half top-hat electron sensors and eight (8) identical ion sensors and an Instrument Data Processing Unit (IDPU). The sensors (electron or ion) are grouped into pairs whose 6 degree x 180 degree fields-of-view (FOV) are set 90 degrees apart. Each sensor is equipped with electrostatic aperture steering to allow the sensor to scan a 45 degree x 180 degree fan about the its nominal viewing (0 deflection) direction. Each pair of sensors, known as the Dual Electron Spectrometer (DES) and the Dual Ion Spectrometer (DIS), occupies a quadrant on the MMS spacecraft and the combination of the eight electron/ion sensors, employing aperture steering, image the full-sky every 30-ms (electrons) and 150-ms (ions), respectively. To probe the diffusion regions of reconnection, the highest temporal/spatial resolution mode of FPI results in the DES complement of a given spacecraft generating 6.5-Mb (raised dot) per second of electron data while the DIS generates 1.1-Mb (raised dot) per second of ion data yielding an FPI total data rate of 6.6-Mb (raised dot) per second. The FPI electron/ion data is collected by the IDPU then transmitted to the Central Data Instrument Processor (CIDP) on the spacecraft for science interest ranking. Only data sequences that contain the greatest amount of temporal/spatial structure will be intelligently down-linked by the spacecraft. This requires a data ranking process known as the burst trigger system. The burst trigger system uses pseudo physical quantities to approximate the local plasma environments. As each pseudo quantity will have a different value, a set of two scaling factors is employed for each pseudo term. These pseudo quantities are then combined at the instrument, spacecraft, and observatory level leading to a final ranking of data based on expected scientific interest. Here, we present simulations of the fixed point burst trigger system for the FPI. A variety of data sets based on previous mission data as well as analytical formulations are tested. Comparisons of floating point calculations versus the fixed point hardware simulation are shown. Analysis of the potential sources of error from overflows, quantization, etc. are examined and mitigation methods are presented. Finally a series of calibration curves are presented, showing the expected error in pseudo quantities based solely on the scale parameters chosen and the expected data range. We conclude with a presentation of the current base-lined FPI burst trigger approach.
An examination of the effects of atomicity (higher resolution) on the performance of array processors (data-flow computers) is presented. Data-flow principles are reviewed, noting the reliance on parallel processing using functional languages to specify sequencing of the operations. Techniques are described for eliminating the necessity of copying whole arrays between processing steps, thereby reducing the number of store cycles. The method involves setting whole columns to specific values rather than individual elements. The individual column values can be processed in parallel, i.e., a locally optimized condition exists. A drawback of the system is the need for more low level arguments, to identify the appropriate processing sequences, and high system complexity.
cThe Xilinx AXI Direct Memory Access (AXI DMA) module is an efficient solution for medium-speed data transfer in Xilinx SoC FPGAs, supporting data rates greater than 1000 Gbps even in very suboptimal operating modes. It facilitates direct transfer of AXI stream data into processor memory without constant software intervention, which reduces overhead and ensures consistent data logging. By utilizing the FPGA's available memory, large circular buffers (1-5 GiB) are used to buffer data and accommodate network limitations, enabling high-rate data bursts. In this study, we measured the performance of AXI DMA under conditions simulating its lowest practical data transfer speeds. The Arbitrary Length Data Sender was used to transmit AXI stream packets at 32-bit width and 100 MHz frequency, a narrow width and slow speed. Results show that the AXI DMA can transfer up to 3192.76 Mbps with large packet sizes but experiences reduced performance for smaller packets, as low as 2.6 Mbps for 4-byte packets. For Ethernet-limited applications, packet sizes between 8,000 and 16,000 bytes provided optimal transfer speeds of 874 to 1600 Mbps. These findings suggest that the AXI DMA is not the limiting factor in systems where packet sizes exceed 8,000 bytes.
The Intensity Dependent Spread (IDS) is an adaptive algorithm which is modified according to the local intensity in the scene. (This results in a nonlinear process which cannot take advantage of rather nice linear transform methods.) The computation is similar to a neural net whereby intensity information is moving from each input pixel to a set of surrounding output pixels in a manner described by Cornsweet and Yellott. A prototype of a very large scale integration IDS processor is being developed and implemented in a workstation environment. The workstation consists of a SUN 3/260 and a DATACUBE pipeline processor. The IDS prototype is a board set which operates in the DATA CUBE processor. The SUN 3/260 performs control, background processing, IDS simulation and image display functions.
This paper develops a new model, using matrices, for the analysis of fault-tolerant multiprocessor systems. The relationship between processors computing useful data, the output data, and the check processors is defined in terms of matrix entries. Unlike the matrix-based models proposed previously for the analysis of digital systems, this model uses only numerical computations rather than logical operations for the analysis of a system. Algorithms to evaluate the fault detection and location capability of the system are proposed which are much less complex than the existing ones. The new model is used to analyze some fault-tolerant architectures proposed for signal-processing applications.
The aerodynamic computer code, OVERFLOW, with a multi-zone overset grid feature, has been parallelized to enhance its performance on distributed and shared memory paradigms. Practical application benchmarks have been set to assess the efficiency of code's parallelism on high-performance architectures. The code's performance has also been experimented with in the context of the distributed computing paradigm on distant computer resources using the Information Power Grid (IPG) toolkit, Globus. Two parallel versions of the code, namely OVERFLOW-MPI and -MLP, have developed around the natural coarse grained parallelism inherent in a multi-zonal domain decomposition paradigm. The algorithm invokes a strategy that forms a number of groups, each consisting of a zone, a cluster of zones and/or a partition of a large zone. Each group can be thought of as a process with one or multithreads assigned to it and that all groups run in parallel. The -MPI version of the code uses explicit message-passing based on the standard MPI library for sending and receiving interzonal boundary data across processors. The -MLP version employs no message-passing paradigm; the boundary data is transferred through the shared memory. The -MPI code is suited for both distributed and shared memory architectures, while the -MLP code can only be used on shared memory platforms. The IPG applications are implemented by the -MPI code using the Globus toolkit. While a computational task is distributed across multiple computer resources, the parallelism can be explored on each resource alone. Performance studies are achieved with some practical aerodynamic problems with complex geometries, consisting of 2.5 up to 33 million grid points and a large number of zonal blocks. The computations were executed primarily on SGI Origin 2000 multiprocessors and on the Cray T3E. OVERFLOW's IPG applications are carried out on NASA homogeneous metacomputing machines located at three sites, Ames, Langley and Glenn. Plans for the future will exploit the distributed parallel computing capability on various homogeneous and heterogeneous resources and large scale benchmarks. Alternative IPG toolkits will be used along with sophisticated zonal grouping strategies to minimize the communication time across the computer resources.
Self repairing, multiple pool processor and computer for spacecraft central data system
Analysis of the postbuckling behavior of composite shell structures pose many difficult and challenging problems in the field of structural mechanics. Current analysis methods perform well for most cases in predicting the postbuckling response of undamaged components. To predict component behavior accurately at higher load levels, the analysis must include the effects of local material failures. The CSM testbed software system is a highly modular structural analysis system currently under development at Langley Research Center. One of the primary goals of the CSM testbed is to provide a software environment for the development of advanced structural analysis methods and modern numerical methods which will exploit advanced computer architecture such as parallel-vector processors. Development of a progressive failure analysis method consists of the design and implementation of a processor which will perform the ply-level progressive failure analysis and the development of a geometrically nonlinear analysis procedure which incorporates the progressive failure processor. Regarding the development of the progressive failure processor, two components are required: failure criteria and a degradation model. For the initial implementation, the failure criteria of Hashin will be used. For a matrix failure which typically indicates the development of transverse matrix cracks, the ply properties will be degraded. Work to date includes the design of the progressive failure analysis processor and initial plans for the controlling geometrically nonlinear analysis procedure. The implementation of the progressive failure analysis has begun. Access to the model database and the Hashin failure criteria are completed. Work is in progress on the input/output operations for the processor related data and the finite element model updating procedures. In total the progressive failure processor is approximately one-third complete.
Exponential downweighting of past data in single stage, weighted least squares trajectory processor for coasting spacecraft