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 1,801 records · Page 100
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Characterization and Optimization of the Fitting of Quantum Correlation Functions

This case study presents a characterization and optimization of an application code for extracting parton distribution functions from high energy electron-proton scattering data. Profiling this application code reveals that the phase-space density computation accounts for 93% of the overall execution time for a single iteration on a single core. When executing multiple iterations in parallel on a multicore system, the application spends 78% of its overall execution time idling due to load imbalance. We address these issues by first transforming the application code from Python to C++ and then tackling the application load imbalance via a hybrid scheduling strategy that combines dynamic and static scheduling. These techniques result in a 62% reduction in CPU idle time and a 2.46x speedup in overall execution time per node. In addition, the typically enabled power-management mechanisms in supercomputers (e.g., AMD Turbo Core, Intel Turbo Boost, and RAPL) can significantly impact intra-node scalability when more than 50% of the CPU cores are used. This finding underscores the importance of understanding system interactions with power management, as they can adversely impact application performance, and highlights the necessity of intra-node scaling tests to identify performance degradation that inter-node scaling tests might otherwise overlook.

Chuang, Pi-Yueh [Virginia Tech,Dept. of Computer S↗
DOE OSTI2016

A Prototype Gamma Tomosynthesis System for Molecular Breast Imaging

We are developing an improved gamma camera for molecular breast imaging (MBI) that allows tomographic imaging and improved signal-to-noise over conventional planar methods. We have built a novel gamma camera for MBI that utilizes limited angle tomography, similar to x-ray tomosynthesis, to generate images with depth information and improved signal-to-noise. A unique feature of the gamma tomosynthesis system is a variable-angle slant hole (VASH) collimator. This collimator allows the camera to remain flush against the compression paddle during the tomographic acquisition, which achieves high spatial resolution and simplified detector motion. We constructed a VASH collimator from a stack of tungsten sheets, each sheet containing a matrix of holes created by photo-etching. With the holes of the sheets aligned, a parallel-hole collimator is created; shearing the sheets creates a variable angle slant hole collimator. The shearing is controlled by a motorized mechanism. The collimator was mounted on a commercial MBI camera and projection data were acquired of a set of capillary tube line sources over a 50° range of angles. The projection data were reconstructed using the iterative MLEM method. Results demonstrated the ability of the system to resolve in the depth dimension, and spatial resolution matched expected values. We conclude that the proposed gamma tomosynthesis system has the potential to provide improved signal-to-noise over conventional MBI methods and allow a reduction in the administered radioactivity and patient dose.

Gilland, David↗
DOE OSTI2017

Machine and Application Aware Partitioning for Adaptive Mesh Refinement Applications. In: HPDC '17 Proceedings of the 26th International Symposium on High-Performance Parallel and Distributed Computing

Load balancing and partitioning are critical when it comes to parallel computations. Popular partitioning strategies based on space filling curves focus on equally dividing work. The partitions produced are independent of the architecture or the application. Given the ever-increasing relative cost of data movement and increasing heterogeneity of our architectures, it is no longer sufficient to only consider an equal partitioning of work. Minimizing communication costs are equally if not more important. Our hypothesis is that an unequal partitioning that minimizes communication costs significantly can scale and perform better than conventional equal-work partitioning schemes. This tradeoff is dependent on the architecture as well as the application. We validate our hypothesis in the context of a finite-element computation utilizing adaptive mesh-refinement. Our central contribution is a new partitioning scheme that minimizes the overall runtime of subsequent computations by performing architecture and application-aware non-uniform work assignment in order to decrease time to solution, primarily by minimizing data-movement. We evaluate our algorithm by comparing it against standard space-filling curve based partitioning algorithms and observing time-to-solution as well as energy-to-solution for solving Finite Element computations on adaptively refined meshes. We demonstrate excellent scalability of our new partition algorithm up to $262,144$ cores on ORNL's Titan and demonstrate that the proposed partitioning scheme reduces overall energy as well as time-to-solution for application codes by up to 22.0%.

Fernando, Milinda↗
DOE OSTI2016

United States Department of Energy severe accident research following the Fukushima Daiichi accidents

The U.S. Department of Energy (DOE) has played a major role in the U.S. response to the events at Fukushima Daiichi. During the first several weeks following the accident, U.S. assistance efforts were guided by results from a significant and diverse set of analyses. In the months that followed, a coordinated analysis activity aimed at gaining a more thorough understanding of the accident sequence was completed using laboratory-developed, system-level best-estimate accident analysis codes, while a parallel analysis was conducted by U.S. industry. A comparison of predictions for Unit 1 from these two studies indicated significant differences between MAAP and MELCOR results for key plant parameters, such as in-core hydrogen production. On that basis, a crosswalk was completed to determine the key modeling variations that led to these differences. In parallel with these activities, it became clear that there was a need to perform a technology gap evaluation on accident-tolerant components and severe accident analysis methodologies with the goal of identifying any data and/or knowledge gaps that may exist given the current state of light water reactor (LWR) severe accident research and augmented by insights from Fukushima. In addition, there is growing international recognition that data from Fukushima could significantly reduce uncertainties related to severe accident progression, particularly for boiling water reactors. On these bases, a group of U. S. experts in LWR safety and plant operations was convened by the DOE Office of Nuclear Energy (DOE-NE) to complete technology gap analysis and Fukushima forensics data needs identification activities. The results from these activities were used as the basis for refining DOE-NE's severe accident research and development (R&D) plan. Finally, this paper provides a high-level review of DOE-sponsored R&D efforts in these areas, including planned activities on accident-tolerant components and accident analysis methods.

61 RADIATION PROTECTION AND DOSIMETRY↗
DOE OSTI2021

Tools for Assessing Performance Project: FY2021 Quarter 4 Report

The rotatable building located at Texas Tech’s Reese Technology Center is approximately 14 m wide (width being defined as more normal to the wind than parallel), 9 m long (aligned more with the wind than perpendicular), and 4 m tall. By placing 29 sonic anemometers downwind of the building, see Fig. 1, this facility provided a chance to collect data concerning both the wake velocity deficit distribution in the downstream and crosssteam directions (relative to the mean wind) behind an isolated building. A preliminary comparison between this data and the recently- developed fast-running diffusive wake model, which was developed based on wind tunnel and LES (performed with JOULES) simulations, for the purpose of either validating this model of understanding potential persistent differences between the idealized wind tunnel or LES conditions and those of full-scale real world phenomena. For this preliminary exploration, we utilized the diffusive wake model implemented in the QUIC model.

42 ENGINEERING↗
DOE OSTI2015

Cloud object store for archive storage of high performance computing data using decoupling middleware

Cloud object storage is enabled for archived data, such as checkpoints and results, of high performance computing applications using a middleware process. A plurality of archived files, such as checkpoint files and results, generated by a plurality of processes in a parallel computing system are stored by obtaining the plurality of archived files from the parallel computing system; converting the plurality of archived files to objects using a log structured file system middleware process; and providing the objects for storage in a cloud object storage system. The plurality of processes may run, for example, on a plurality of compute nodes. The log structured file system middleware process may be embodied, for example, as a Parallel Log-Structured File System (PLFS). The log structured file system middleware process optionally executes on a burst buffer node.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Proper orthogonal decomposition analysis of parallel twin jets measured using particle image velocimetry - 18685

The mixing of turbulent jets has been found in several locations of High Temperature Gas-cooled Reactors (HTGRs) including upper and lower plenums, where jets impinge into open space from coolant channels. In the present work, a parallel twin jets experimental facility was used to generate velocity field of mixing jets by means of Particle Image Velocimetry (PIV), which can be eventually used for validation of Computational Fluid Dynamics (CFD) simulations. The facility has two fixed parallel rectangular jet nozzles (5.8 mm in width, 87.8 mm in length) submerged in the middle of the water tank, which allows an assumption of free jet mixing. In order to compare the experimental data produced to existing data, Re = 4000 {approx} 5000 at the jet outlet was selected. In the past, several statistically meaningful data have been reported using Laser Doppler Velocimetry (LDV) and Hot Wire Anemometer (HWA), which can support Reynolds Averaged Navier-Stokes turbulent models. However, recent CFD simulations using Large Eddy Simulation (LES) require high resolution experimental data in time and space. Thus, in this study, in addition to statistically enough number of data sets for RANS models validation, high temporal and spatial resolution data were collected for the validation of LES calculations. In order to study the turbulent structure Reynolds decomposition and Proper Orthogonal Decomposition (POD) were applied to the data obtained. These decomposition analyses visualized multi-scale vortices and the transportation of eddies. In terms of twin jets structure, the present experimental data showed good agreement with other techniques for the important characteristic parameters including the length of the converging region and the location of the merging point. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2021

A physics-informed and hierarchically regularized data-driven model for predicting fluid flow through porous media

This paper presents a new deep learning data-driven model for predicting structure dependent pore-fluid velocity fields in rock. The model is based on a Convolutional Auto-Encoder (CAE) artificial neural network capable of learning from image data generated by direct numerical simulations of fluid flow through pore-structures, such as by Lattice Boltzmann or molecular dynamics methods. The main novelty of the model in comparison to previous CAE-based data-driven approaches consists of three parts. The first is a methodology for decomposing the full-domain of the porous media into sub-regions, or “sub-domains”, in order to reduce the overall size of the CAE, batch process the sub-domains in parallel, and enable the CAE to learn local and generalizable nonlinear mappings of pore-fluid velocities. The second consists of embedding the finite difference solutions of the incompressible Navier-Stokes and continuity equations into convolutional layers prior to the CAE in order to provide the CAE with knowledge of fluid dynamics physics (PhyFlow). The third main novelty is that the training of the CAE is regularized with a hierarchical loss function that encourages the learning of fluid flow patterns (in a way similar to ranked modes in principal component analysis), ranking from most to least important. This is shown to increase the stability in learning, reduce over-fitting, and promote interpretability of the CAE neural network layers (HierCAE). The comprehensive new data-driven model, which we call the PhyFlow-HierCAE model, is shown to exhibit improved accuracy and generalizability of flow field predictions over conventional CAE models, attributable to the embedded physical knowledge and the hierarchical regularization, as well as realize orders of magnitude speed-ups in computation times as a surrogate for the direct numerical simulations. Examples of training and forward predictions on unseen pore-structures are provided and evaluated for data from Lattice Boltzmann and molecular dynamics simulations of pore-fluid flow. The model is shown to be a fast and accurate emulator (or “surrogate”) for predicting effective permeability of unseen pore-structures based on learning from relatively small direct numerical simulation datasets.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2017

PLATO (Parallel Load Assignment Tool): A Parallel Workload Partitioner for Particle-based Methods

A key first step in many large-scale high-performance scientific computing applications is the distribution of the computational domain across all participating processors of a parallel computing platform such as a shared memory machine, cluster or a tightly-coupled supercomputer. The primary goal of such a partitioning step is to ensure that the computational work performed by each process remains load-balanced throughout the execution of the parallel application. In this project, a general-purpose parallel partitioning tool, code-named PLATO, was designed and implemented. PLATO assumes that the workload is a set of points in an n-dimensional metric space. These points are abstractions of computational units that may represent point-like computational units such as celestial objects in astrophysics, atoms in materials sciences and molecules in biology/chemistry, or may represent more abstract computational units such as grid points in finite difference methods, grid elements in finite element methods or n-dimensional points in the feature space of a host of other applications. PLATO supports several key and useful capabilities, namely: (a) multiple load partitioning strategies, (b) support for arbitrary user-defined processor topology, (c) pre-fetching data from neighboring processors based on user-defined parameters, and (d) support for arbitrary number and types of attributes associated with each point in the computational domain. PLATO is implemented in the C programming language using Message Passing Interface (MPI) for inter-processor communications and its parallel performance tested on up to 48 processors of a shared memory machine. This report includes a detailed performance study of PLATO using parallel speedup and iso-efficiency as the two main metrics of parallel performance. In addition, a user manual and DOXYGEN generated code documentation have been included for completeness.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Expert and operator perspectives on barriers to energy efficiency in data centers

Abstract It was last estimated in 2016 that data centers (DCs) comprise approximately 2% of total US electricity consumption. However, this estimate is currently being updated to account for the massive increase in computing needs due to streaming, cryptocurrency, and artificial intelligence (AI). To prevent energy consumption that tracks with increasing computing needs, it is imperative we identify energy efficiency strategies and investments beyond the low-hanging fruit solutions. In a two-phased research approach, we ask: What non-technical barriers still impede energy efficiency (EE) practices and investments in the data center sector, and what can be done to overcome these barriers? In particular, we are focused on social and organizational barriers to EE. In Phase I, we performed a literature review and found that technical solutions are abundant in the literature, but fail to address the top-down cultural shifts that need to take place in order to adapt new energy efficiency strategies. In Phase II, reported here, we interviewed 16 data center operators/experts to ground-truth our literature findings. Our interview protocols focus on three aspects of DC decision-making: procurement practices, metrics and monitoring, and perceived barriers to energy efficiency. We find that vendors are the key drivers of procurement decisions, advanced efficiency metrics are facility-specific, and there is convergence in the design of advanced facilities due to the heat density of parallelized infrastructure. Our ultimate goals for our research are to design DC decarbonization policies that target organizational structure, empower individual staff, and foster a supportive external market.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2018

Machine Learning Based Parallel I/O Predictive Modeling: A Case Study on Lustre File Systems

Parallel I/O hardware and software infrastructure is a key contributor to performance variability for applications running on large-scale HPC systems. This variability confounds efforts to predict application performance for characterization, modeling, optimization, and job scheduling. We propose a modeling approach that improves predictive ability by explicitly treating the variability and by leveraging the sensitivity of application parameters on performance to group applications with similar characteristics. We develop a Gaussian process-based machine learning algorithm to model I/O performance and its variability as a function of application and file system characteristics. We demonstrate the effectiveness of the proposed approach using data collected from the Edison system at the National Energy Research Scientific Computing Center. The results show that the proposed sensitivity-based models are better at prediction when compared with application-partitioned or unpartitioned models. We highlight modeling techniques that are robust to the outliers that can occur in production parallel file systems. Using the developed metrics and modeling approach, we provide insights into the file system metrics that have a significant impact on I/O performance.

I/O performance variability↗
DOE OSTI2017

PapyrusKV: a high-performance parallel key-value store for distributed NVM architectures

This paper introduces PapyrusKV, a parallel embedded key-value store (KVS) for distributed high-performance computing (HPC) architectures that offer potentially massive pools of nonvolatile memory (NVM). PapyrusKV stores keys with their values in arbitrary byte arrays across multiple NVMs in a distributed system. PapyrusKV provides standard KVS operations such as put, get, and delete. More importantly, PapyrusKV provides advanced features for HPC such as dynamic consistency control, zero-copy workflow, and asynchronous checkpoint/restart. Beyond filesystems, PapyrusKV provides HPC programmers with a high-level interface to exploit distributed NVM in the system, and it transparently organizes data to achieve high performance. Also, it allows HPC applications to specialize PapyrusKV to meet their specific requirements. We empirically evaluate PapyrusKV on three HPC systems with real NVM devices: OLCF's Summitdev, TACC's Stampede, and NERSC's Cori. Our results show that PapyrusKV can offer high performance, scalability, and portability across these various distributed NVM architectures.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Task Parallel Incomplete Cholesky Factorization using 2D Partitioned-Block Layout

We introduce a task-parallel algorithm for sparse incomplete Cholesky factorization that utilizes a 2D sparse partitioned-block layout of a matrix. Our factorization algorithm follows the idea of algorithms-by-blocks by using the block layout. The algorithm-byblocks approach induces a task graph for the factorization. These tasks are inter-related to each other through their data dependences in the factorization algorithm. To process the tasks on various manycore architectures in a portable manner, we also present a portable tasking API that incorporates different tasking backends and device-specific features using an open-source framework for manycore platforms i.e., Kokkos. A performance evaluation is presented on both Intel Sandybridge and Xeon Phi platforms for matrices from the University of Florida sparse matrix collection to illustrate merits of the proposed task-based factorization. Experimental results demonstrate that our task-parallel implementation delivers about 26.6x speedup (geometric mean) over single-threaded incomplete Choleskyby- blocks and 19.2x speedup over serial Cholesky performance which does not carry tasking overhead using 56 threads on the Intel Xeon Phi processor for sparse matrices arising from various application problems.

2D layout↗
DOE OSTI2023

Utah FORGE: Well 16B(78)-32 Drilling Data

This drilling data for Utah FORGE well 16B(78)-32 include a well survey, core summary, mud and mud temperature logs, daily reports of the drilling process, and additional data from the Pason oil and gas company. Well 16B(78)-32 serves as the production well for reservoir creation, fluid circulation, and demonstration of heat extraction for the FORGE project. It has been drilled as a doublet approximately 300 feet parallel to and above the injection well 16A(78)-32. The proposed total depth was 10,658 feet, which was exceeded. Spudding began on April 26th, 2023. As of June 20th, 2023, the total depth measured 10,947 feet and the vertical depth measured 8,357 feet. Drilling included the trial use of insulated drill pipe (IDP) from Eavor Technologies, which was considered a success. IDP restricts counter-current heat transfer between drilling fluid inside the drill string and hotter returning fluid in the annulus, thereby ensuring the bottomhole assembly (BHA) remains submerged in cool fluid. Eavor rented 350 joints of IDP to FORGE which were run in two consecutive BHAs at the well. The results of this trial are included here.

15 GEOTHERMAL ENERGY↗
DOE OSTI2014

Unresolved fine-scale structure in solar coronal loop-tops

New and advanced space-based observing facilities continue to lower the resolution limit and detect solar coronal loops in greater detail. We continue to discover even finer substructures within coronal loop cross-sections, in order to understand the nature of the solar corona. Here, we push this lower limit further to search for the finest coronal loop substructures, through taking advantage of the resolving power of the Swedish 1 m Solar Telescope/CRisp Imaging Spectro-Polarimeter (CRISP), together with co-observations from the Solar Dynamics Observatory/Atmospheric Image Assembly (AIA). High-resolution imaging of the chromospheric Hα 656.28 nm spectral line core and wings can, under certain circumstances, allow one to deduce the topology of the local magnetic environment of the solar atmosphere where its observed. Here, we study post-flare coronal loops, which become filled with evaporated chromosphere that rapidly condenses into chromospheric clumps of plasma (detectable in Hα) known as a coronal rain, to investigate their fine-scale structure. We identify, through analysis of three data sets, large-scale catastrophic cooling in coronal loop-tops and the existence of multi-thermal, multi-stranded substructures. Many cool strands even extend fully intact from loop-top to footpoint. We discover that coronal loop fine-scale strands can appear bunched with as many as eight parallel strands within an AIA coronal loop cross-section. The strand number density versus cross-sectional width distribution, as detected by CRISP within AIA-defined coronal loops, most likely peaks at well below 100 km, and currently, 69% of the substructure strands are statistically unresolved in AIA coronal loops.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

GLEAM: Galaxy Line Emission & Absorption Modeling

We present Galaxy Line Emission & Absorption Modeling (gleam), a Python tool for fitting Gaussian models to emission and absorption lines in large samples of 1D extragalactic spectra. gleam is tailored to work well in batch mode without much human interaction. With gleam, users can uniformly process a variety of spectra, including galaxies and active galactic nuclei, in a wide range of instrument setups and signal-to-noise regimes. gleam also takes advantage of multiprocessing capabilities to process spectra in parallel. With the goal of enabling reproducible workflows for its users, gleam employs a small number of input files, including a central, user-friendly configuration in which fitting constraints can be defined for groups of spectra and overrides can be specified for edge cases. For each spectrum, gleam produces a table containing measurements and error bars for the detected spectral lines and continuum and upper limits for nondetections. For visual inspection and publishing, gleam can also produce plots of the data with fitted lines overlaid. In the present paper, we describe gleam’s main features, the necessary inputs, expected outputs, and some example applications, including thorough tests on a large sample of optical/infrared multi-object spectroscopic observations and integral field spectroscopic data. gleam is developed as an open-source project hosted at https://github.com/multiwavelength/gleam and welcomes community contributions.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2018

Thin film confinement reduces compatibility in symmetric ternary block copolymer/homopolymer blends

Here, the thin film phase behavior of ternary blends consisting of symmetric poly(styrene) (PS)– b –poly(dimethylsiloxane)(PDMS), PS, and PDMS was investigated using X–ray reflectivity (XRR) and atomic force microscopy (AFM). This system is strongly segregated, and the homopolymers are approximately the same length as the corresponding blocks of the copolymer. The XRR and AFM data are used to quantify changes in domain spacing ( L ) and morphology evolution with increasing homopolymer content (Φ H ). In 100 nm thick films, from Φ H = 0 to 0.20, the system maintains a perfect parallel lamellar structure and domains swell as predicted based on theory; however, from Φ H = 0.30 to 0.50, a morphology transition to a “dot pattern” morphology (tentatively identified as perforated lamellae) and mixed morphologies were observed before macrophase separation. In thicker films, dot patterns were observed for a broad range of Φ H before macrophase separation. The absence of the bicontinuous microemulsion phase reported for bulk blends and thin films of perpendicular lamellae and the presence of dot patterns/perforated lamellae are attributed to preferential migration of the PDMS homopolymer to the wetting layers located at the substrate and free air interfaces, which leads to an asymmetric composition within the film and morphology transition.

36 MATERIALS SCIENCE↗
DOE OSTI2018

Orientation of Ferroelectric Domains and Disappearance upon Heating Methylammonium Lead Triiodide Perovskite from Tetragonal to Cubic Phase

We study the spontaneous polarization of the archetypal semiconducting halide perovskite methylammonium lead triiodide (CH 3 NH 3 PbI 3 ) that is currently being investigated for use in thin film solar cells and light-emitting diodes. Using both lateral and vertical piezoresponse force microscopy (PFM) to image polycrystalline thin films, we observed domains in the piezoresponse that reversibly appear and disappear below and above the tetragonal-to-cubic phase transition temperature. Importantly, we observe these domains to exhibit a piezoresponse that is predominantly in-plane for films with the (110) plane oriented parallel to the substrate, providing a measure of the polarization associated with specific crystal planes. We characterize the polarization and its temporal response using both local switching spectroscopy and time-dependent PFM spectra. These data show hysteresis loops with the polarization switching with bias but relaxing back on time scales of several minutes. Furthermore, our results suggest the existence of ferroelectric behavior due to off-center displacement of the Pb 2+ cation, although the local polarization response is complicated by the presence of local ionic and electronic conductivity. Understanding the nature of these domains paves the way for further optimization of optoelectronic devices using CH 3 NH 3 PbI 3 perovskite material.

36 MATERIALS SCIENCE↗

Refine your search to explore more results.