Search NASA⌕ Search

SEARCH · Search NASA

Results for “data parallel”

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 1,693 records · Page 94
All sourcesNASA NTRSDOE OSTI
DOE OSTI2015

Optimized Data Transfers Based on the OpenCL Event Management Mechanism

In standard OpenCL programming, hosts are supposed to control their compute devices. Since compute devices are dedicated to kernel computation, only hosts can execute several kinds of data transfers such as internode communication and file access. These data transfers require one host to simultaneously play two or more roles due to the need for collaboration between the host and devices. The codes for such data transfers are likely to be system-specific, resulting in low portability. Here, this article proposes an OpenCL extension that incorporates such data transfers into the OpenCL event management mechanism. Unlike the current OpenCL standard, the main thread running on the host is not blocked to serialize dependent operations. Hence, an application can easily use the opportunities to overlap parallel activities of hosts and compute devices. In addition, the implementation details of data transfers are hidden behind the extension, and application programmers can use the optimized data transfers without any tricky programming techniques. The evaluation results show that the proposed extension can use the optimized data transfer implementation and thereby increase the sustained data transfer performance by about 18% for a real application accessing a big data file.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

THE INTERACTION OF TURBULENCE WITH PARALLEL AND PERPENDICULAR SHOCKS: THEORY AND OBSERVATIONS AT 1 au

Shocks are thought to be responsible for the amplification of turbulence as well as for generating turbulence throughout the heliosphere. We study the interaction of turbulence with parallel and perpendicular shock waves using the six-coupled-equation turbulence transport model of Zank et al. We model a 1D stationary shock wave using a hyperbolic tangent function and the Rankine–Hugoniot conditions for both a reduced model with four coupled equations and the full model. Eight quasi-parallel and five quasi-perpendicular events in the WIND spacecraft data sets are identified, and we compute the fluctuating magnetic and kinetic energy, the energy in forward and backward propagating modes, the total turbulent energy, the normalized residual energy, and the normalized cross helicity upstream and downstream of the observed shocks. We compare the observed fitted values upstream and downstream of the shock with numerical solutions to our model equations. The comparison shows that our theoretical results are in reasonable agreement with observations for both quasi-parallel and perpendicular shocks. We find that (1) the total turbulent energy, the energy in forward and backward propagating modes, and the normalized residual energy increase across the shock, (2) the normalized cross helicity increases or decreases across the shock, and (3) the correlation length increases upstream and downstream of the shock, and slightly flattens or decreases across the shock.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRS1980

Overseas trip report, CV 990 underflight mission

The scanning microwave radiometer-7 simulator, the ocean temperature scanner, and an imaging scatterometer/altimeter operating at 14 GHz were carried onboard the NASA CV-990 over open oceans, sea ice, and continental ice sheets to gather surface truth information. Data flights were conducted over the Norwegian Sea to map the ocean polar front south and west of Bear Island and to transect several Nimbus-7 footprints in a rectangular pattern parallel to the northern shoreline of Norway. Additional flights were conducted to obtain correlative data on the cryosphere parameters and characteristics of the Greenland ice sheet, and study the frozen lakes near Barrow. The weather conditions and flight path way points for each of the nineteen flights are presented in tables and maps.

Gloersen, P.↗
NASA NTRS1994

Statistical study of ULF wave occurrence in the dayside magnetosphere

Ultralow-frequency (ULF) waves are observed almost everywhere in the dayside magnetosphere. The mechanism by which these waves are generated and transformed in the dayside magnetosphere is still not understood. Here we report a statistical study of these waves based on magnetic field data from the International Sun-Earth Explorer 1 (ISEE 1) spacecraft. Data from the first traversal of the spacecraft through the entire dayside magnetosphere have been examined to determine the spatial distribution of wave occurrence. Successive 20-min segments of data were transformed to a field-aligned coordinate system. The parallel component was detrended and all three components of the field spectrally analyzed. Wave occurrence was defined by the presence of significant peaks in the power spectra. Wave events were categorized by three wave frequency bands: Pc 3 with T approximately 10-45 s; Pc 4 with T approximately 45-150 s; the short-period part of the Pc 5 wave band with T approximately 150-324 s. Properties of the spectral peaks were then entered into a data base. The data base was next sorted to determine the spatial occurrence pattern for the waves. Our results show that Pc 3 waves most frequently occur just outside synchronous orbit and are approximately centered on local noon. Pc 4 waves have a similar distribution with its peak further out. Pc 5 waves have high occurrence rate at the two flanks of the magnetosphere. Peaks in spectra obtained near the magnetopause are less clearly defined than those deeper in the magnetosphere.

Cao, M.↗
NASA NTRS1987

Contextual classification on the massively parallel processor

Classifiers are often used to produce land cover maps from multispectral Earth observation imagery. Conventionally, these classifiers have been designed to exploit the spectral information contained in the imagery. Very few classifiers exploit the spatial information content of the imagery, and the few that do rarely exploit spatial information content in conjunction with spectral and/or temporal information. A contextual classifier that exploits spatial and spectral information in combination through a general statistical approach was studied. Early test results obtained from an implementation of the classifier on a VAX-11/780 minicomputer were encouraging, but they are of limited meaning because they were produced from small data sets. An implementation of the contextual classifier is presented on the Massively Parallel Processor (MPP) at Goddard that for the first time makes feasible the testing of the classifier on large data sets.

Tilton, James C.↗
NASA NTRS1985

Evaluation of existing and proposed computer architectures for future ground-based systems

Parallel processing architectures and techniques used in current supercomputers are described and projections are made of future advances. Presently, the von Neumann sequential processing pattern has been accelerated by having separate I/O processors, interleaved memories, wide memories, independent functional units and pipelining. Recent supercomputers have featured single-input, multiple data stream architectures, which have different processors for performing various operations (vector or pipeline processors). Multiple input, multiple data stream machines have also been developed. Data flow techniques, wherein program instructions are activated only when data are available, are expected to play a large role in future supercomputers, along with increased parallel processor arrays. The enhanced operational speeds are essential for adequately treating data from future spacecraft remote sensing instruments such as the Thematic Mapper.

Schulbach, C.↗
DOE OSTI2023

HEPnOS: a Specialized Data Service for High Energy Physics Analysis

In this paper, we present HEPnOS, a distributed data service for managing data produced by high-energy physics (HEP) experiments. Using HEPnOS, HEP applications can use HPC resources more effciently than traditional fle-based applications. The fle-based model leads to a rigid, chunk-based allocation of computational resources and limits the number of cores that can be used concurrently by an HEP application. The fundamental problem is that organizing domain-specifc data into fles inadvertently introduces a single, artifcial, confated tuning parameter that puts key optimization goals into confict: larger fle sizes reduce metadata overhead and thus improve I/O effciency, but smaller fle sizes provide more opportunity for workfow parallelism and load balancing. In this work, we introduce a domain-specifc data service that decouples that constraint so that data can be accessed and processed in its natural granularity while still maintaining I/O effciency. By removing the constraints introduced by fle handling we are able to obtain better scaling and make effcient use of more cores for processing a fxed-sized data sample. We demonstrate the improved scalability by using an application developed in the fle-based paradigm and comparing it to a version modifed to use HEPnOS.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

Phoenix: A Scalable Streaming Hypergraph Analysis Framework

We present Phoenix, a scalable hypergraph analytics framework for data analytics and knowledge discovery that was implemented on the leadership class computing platforms at Oak Ridge National Laboratory (ORNL). Our software framework comprises a distributed implementation of a streaming server architecture which acts as a gateway for various hypergraph generators/external sources to connect. Phoenix has the capability to utilize diverse hypergraph generators, including HyGen, a very large-scale hypergraph generator developed by ORNL. Phoenix incorporates specific algorithms for efficient data representation by exploiting hidden structures of the hypergraphs. Our experimental results demonstrate Phoenix’s scalable and stable performance on massively parallel computing platforms. Phoenix’s superior performance is due to the merging of high-performance computing with data analytic.

Kurte, Kuldeep↗
NASA NTRS2000

The High Mosaicity Illusion: Revealing the True Physical Characteristics of Macromolecular Crystals

An experimental system and software have been developed for simultaneously measuring the diffraction resolution and mosaic spread of macromolecular crystals. Hundreds of reflection profiles over a wide resolution range were rapidly measured by using a charge coupled device (CCD) area detector in combination with superfine phi slicing data collection. The contributions of the X-ray beam to the reflection widths were minimized by using a highly-parallel, highly-monochromatic synchrotron source. These contributions and Lorentz effects were evaluated and deconvoluted from the recorded data. Data collection and processing is described. From one degree of superfine phi slice data collected on a crystal of manganese superoxide dismutase the mosaicity of 261 reflections were measured. The average mosaicity was 0.0101 degrees (0.0035) at the full-width-at-half-maximum (FWHM) and ranged from 0.0011 degrees to 0.0188 degrees. Each reflection profile was individually fit with two gaussian profiles with the first gaussian contributing 55% and the second contributing 35% of the reflection. On average, the mosaicity of the first gaussian was 0.0054 degrees (0.0015) and the second was 0.0061 degrees (0.0023). The mosaicity of the crystal was anisotropic with fh, f k, and fl values of 0.0068 degrees, 0.0140 degrees and 0.0046 degrees, respectively at the FWHM. The anisotropic mosaicity analysis indicates that the crystal is the most perfect in the I direction which corresponds to the favored growth direction of the crystal.

Bellamy, Henry↗
NASA NTRS1998

Distributed Parallel Processing and Dynamic Load Balancing Techniques for Multidisciplinary High Speed Aircraft Design

Multidisciplinary design optimization (MDO) for large-scale engineering problems poses many challenges (e.g., the design of an efficient concurrent paradigm for global optimization based on disciplinary analyses, expensive computations over vast data sets, etc.) This work focuses on the application of distributed schemes for massively parallel architectures to MDO problems, as a tool for reducing computation time and solving larger problems. The specific problem considered here is configuration optimization of a high speed civil transport (HSCT), and the efficient parallelization of the embedded paradigm for reasonable design space identification. Two distributed dynamic load balancing techniques (random polling and global round robin with message combining) and two necessary termination detection schemes (global task count and token passing) were implemented and evaluated in terms of effectiveness and scalability to large problem sizes and a thousand processors. The effect of certain parameters on execution time was also inspected. Empirical results demonstrated stable performance and effectiveness for all schemes, and the parametric study showed that the selected algorithmic parameters have a negligible effect on performance.

Krasteva, Denitza T.↗
NASA NTRS2025

Influence of Microgravity on Bacterial Pathogen Virulence and Immune Cell Function—Relevance for Infectious Disease Risk During Spaceflight

Spaceflight has measurable impacts on astronaut immune profiles as well as the virulence patterns of bacterial pathogens. Data with respect to human immunity indicate diminished T and NK cell function, altered cytokine profiles, persistent inflammation, and latent herpesvirus reactivation. Furthermore, evaluation of International Space Station (ISS) crewmembers gives evidence of compromised immunity, including atypical allergy, infectious disease, and dermatitis. Data with respect to certain human bacterial pathogens suggest modified virulence that may be enhanced. It is therefore critical to examine this interaction of immune dysregulation and increased microbial virulence and whether it might synergistically increase the risk of infectious disease to crew members. The goal of this study is to use modeled microgravity to study the impact of medically significant ISS bacteria that may have altered virulence on the immune response of the host. This study consists of two primary aims to assess this relationship. First, immune cells will be collected from healthy test subjects and cultured in static or in modeled microgravity conditions together with either control pathogens or with microgravity conditioned pathogens that were grown in bioreactors. Second, immune cells will be collected from astronauts before, during, and after flight and co-cultured with the control or the microgravity conditioned bacteria. Three pathogens were selected for this investigation: Pseudomonas aeruginosa , Salmonella enterica serovar Enteritidis and Burkholderia cepacia . Previously, the optimal cell to bacteria ratios that produce the greatest immune cell responses have been derived for these three bacteria. Cellular activation, as determined by the induction of cell surface activation markers and cytokine profiles, will be measured. Interactions between cells and bacteria will be assessed using fluorescent and electron microscopy techniques. This study will provide critical information to help understand how microgravity alters microbial virulence and the associated infectious disease risk to crewmembers during spaceflight missions. Over the past year, the Immunology and Microbiology laboratories at NASA Johnson Space Center have collaborated to process the astronaut subject samples to complete the last year of the flight portion of this study. This included the completion of four astronaut subjects full mission sampling sets. Sample processing included innate and adaptive cell flow cytometry as well as analysis of cytokine concentrations in the supernatant. The ground control segment of the study will take place during FY25 which involves parallel infections run under static and clinostat conditions. The acquired data sets over this 3-year study are now being analyzed to provide a comprehensive set of results and conclusions that will contribute to a more effective risk assessment for astronauts during spaceflight regarding this host-pathogen context.

Immunology↗
DOE OSTI2017

Modelling I and C systems with relays for simulation, adapting CIM standards - 243

As the nuclear power plants are in operation for many years, a main issue is to integrate simulation activities in parallel to the present I and C design process in order to limit impacts on current operating I and C engineering process and commissioning activities. The objective is to keep the existing I and C data and diagrams format. EDF chose to build an intermediate file independent of the tool used for simulation to exchange and store data. This file has been created adapting CIM standards (IEC 61968-13 and IEC 61970-301). It is an XML file built according to an RDF schema. CIM standards are intended to model transport or distribution electrical networks and used to facilitate system integration and information exchanges. We chose to adapt CIM to model EDF electromechanical relay circuitries in order to re-use all methods and tools developed around CIM standards such as RDF schema and RiseClipse, an open source tool developed by EDF and SUPELEC engineering school that will be used to verify exported I and C data according to a data model and to EDF engineering rules. This paper will describe the ongoing project on simulation in parallel to current I and C design process, focusing on how I and C electromechanical relays data are modelled based on CIM standards, and what benefits that model provides to I and C Verification and Validation activities. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
NASA NTRS1985

Generation of available potential energy and the energy cycle during the global weather experiment

Two parallel sets of analyses, which in one case included and in the other omitted data observed by satellite based and other FGGE special observing systems are examined. The results of our previous work is extended in two separate, but not unrelated, ways. First, from these two parallel analyses, which are labeled FGGE (full FGGE system) and NOSAT (satellite omitted), it was discovered that the two sets of fields were quite close over much of the globe. Locally the influence of satellite based systems led to some differences, particularly over the Southern Hemisphere Oceans. The diabatic heating fields generated by the GLA FGGE analysis was also examined. From these fields, one can ascertain the role of total diabatic heating and of the various diabatic heating components in the atmospheric energy cycle, in particular in the generation of available potential energy.

Salstein, D. A.↗
DOE OSTI2017

Characterizing MPI matching via trace-based simulation

With the increased scale expected on future leadership-class systems, detailed information about the resource usage and performance of MPI message matching provides important insights into how to maintain application performance on next-generation systems. However, obtaining MPI message matching performance data is often not possible without significant effort. A common approach is to instrument an MPI implementation to collect relevant statistics. While this approach can provide important data, collecting matching data at runtime perturbs the application's execution, including its matching performance, and is highly dependent on the MPI library's matchlist implementation. In this paper, we introduce a trace-based simulation approach to obtain detailed MPI message matching performance data for MPI applications without perturbing their execution. Using a number of key parallel workloads, we demonstrate that this simulator approach can rapidly and accurately characterize matching behavior. Specifically, we use our simulator to collect several important statistics about the operation of the MPI posted and unexpected queues. For example, we present data about search lengths and the duration that messages spend in the queues waiting to be matched. Here, data gathered using this simulation-based approach have significant potential to aid hardware designers in determining resource allocation for MPI matching functions and provide application and middleware developers with insight into the scalability issues associated with MPI message matching.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

Compiler-Directed Transformation for Higher-Order Stencils

As the cost of data movement increasingly dominates performance, developers of finite-volume and finite-difference solutions for partial differential equations (PDEs) are exploring novel higher-order stencils that increase numerical accuracy and computational intensity. This paper describes a new compiler reordering transformation applied to stencil operators that performs partial sums in buffers, and reuses the partial sums in computing multiple results. This optimization has multiple effect son improving stencil performance that are particularly important to higher-order stencils: exploits data reuse, reduces floating-point operations, and exposes efficient SIMD parallelism to backend compilers. We study the benefit of this optimization in the context of Geometric Multigrid (GMG), a widely used method to solvePDEs, using four different Jacobi smoothers built from 7-, 13-, 27-and 125-point stencils. We quantify performance, speedup, andnumerical accuracy, and use the Roofline model to qualify our results. Ultimately, we obtain over 4× speedup on the smoothers themselves and up to a 3× speedup on the multigrid solver. Finally, we demonstrate that high-order multigrid solvers have the potential of reducing total data movement and energy by several orders of magnitude.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1961

Results of the Second U.S. Manned Suborbital Space Flight, July 21, 1961

This document presents the results of the second United States manned suborbital space flight. The data and flight description presented form a continuation of the information provided at an open conference held under the auspices of the National Aeronautics and Space Administration, in cooperation with the National Institutes of Health and the National Academy of Sciences, at the U.S. Department of State Auditorium on June 6, 1961. The papers presented herein generally parallel the presentations of the first report and were prepared by the personnel of the NASA Manned Spacecraft Center in collaboration with personnel from other government agencies, participating industry, and universities. The second successful manned suborbital space flight on July 21, 1961, in which Astronaut Virgil I. Grissom was the pilot was another step in the progressive research, development, and training program leading to the study of man's capabilities in a space environment during manned orbital flight. Data and operational experiences gained from this flight were in agreement with and supplemented the knowledge obtained from the first suborbital flight of May 5, 1961, piloted by Astronaut Alan B. Shepard, Jr. The two recent manned suborbital flights, coupled with the unmanned research and development flights, have provided valuable engineering nd scientific data on which the program can progress. The successful active participation of the pilots, in much the same way as in the development and testing of high performance aircraft, has. greatly increased our confidence in giving man a significant role in future space flight activities. It is the purpose of this report to continue the practice of providing data to the scientific community interested in activities of this nature. Brief descriptions are presented of the Project Mercury spacecraft and flight plan. Papers are provided which parallel the presentations of data published for the first suborbital space flight. Additional information is given relating to the operational aspects of the medical support activities for the two manned suborbital space flights.

Source record↗
DOE OSTI2022

Power Converter Circuit Design Automation using Parallel Monte Carlo Tree Search

The tidal waves of modern electronic/electrical devices have led to increasing demands for ubiquitous application-specific power converters. A conventional manual design procedure of such power converters is computation- and labor-intensive, which involves selecting and connecting component devices, tuning component-wise parameters and control schemes, and iteratively evaluating and optimizing the design. To automate and speed up this design process, we propose an automatic framework that designs custom power converters from design specifications using Monte Carlo Tree Search. Specifically, the framework embraces the upper-confidence-bound-tree (UCT), a variant of Monte Carlo Tree Search, to automate topology space exploration with circuit design specification-encoded reward signals. Moreover, our UCT-based approach can exploit small offline data via the specially designed default policy and can run in parallel to accelerate topology space exploration. Further, it utilizes a hybrid circuit evaluation strategy to substantially reduce design evaluation costs. Empirically, we demonstrated that our framework could generate energy-efficient circuit topologies for various target voltage conversion ratios. Compared to existing automatic topology optimization strategies, the proposed method is much more computationally efficient --- the sequential version can generate topologies with the same quality while being up to 67% faster. Here, the parallelization schemes can further achieve high speedups compared to the sequential version.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2024

OPER: Optimality-Guided Embedding Table Parallelization for Large-scale Recommendation Model

With the sharp increasing volume of user data, Deep Learning Recommendation Model (DLRM) becomes an indispensable infrastructure in large technology companies. However, large-scale DLRM on the multi-GPU platform is still inefficient due to unbalanced workload partitioning and intensive inter-GPU communication. To this end, we propose OPER, an OPtimality guided Embedding table placement for large-scale Recommendation model training and inference. OPER explores the potential of mitigating remote memory access latency in DLRM through fine-grained embedding table placement. Specifically, OPER proposes a theoretical modeling that builds up the relationship between EMT placement and the embedding communication latency in both training and inference. OPER proves the NP hardness of finding the optimal embedding table placement and proposes a heuristic algorithm that yields near optimal placement. OPER implements a SHMEM-based embedding table training system and a unified embedding index mapping to support fine-grained embedding table sharding and placement. Comprehensive experiments reveal that OPER achieves on average 3.4× and 5.1× speedup on training and inference respectively over state-of-the-art DLRM frameworks.

Wang, Zheng↗