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 739 records · Page 41
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

A Task Based Approach for Co-Scheduling Ensemble Workloads on Heterogeneous Nodes

Scientific workflows consist of multiple, connected applications, with data and results flowing from one to another in a pipeline. Traditionally, such workflows are executed in sequential order, storing intermediate data in storage disks. Co-scheduling application workflows concurrently on the same compute nodes would greatly reduce the cost of moving data to/from storage and allow real-time analysis of intermediate results. Nevertheless, most parallel programming runtimes do not allow seamless integration of various applications in a scientific workflow, in part due to the complexity of managing data and resources. The situation is even more complicated for heterogeneous systems. In this work we extend the Minos Computing Library (MCL) runtime to accelerate pipe-lined and parallel workloads where multiple applications are running in the same system. MCL’s asynchronous task library and runtime dynamically manages resources to allow co-scheduling of multiple processes sharing heterogeneous resources. In addition, we design a custom ex- tension of the Open Compute Language (OpenCL) to enable multiple processes to share device memory. We enable MCL to coordinate these shared buffers to allow for easy, fast data sharing between applications. Using malleable micro-benchmarks and two application workflows that combine scientific simulation and AI-based analysis, we show that our method outperforms traditional approaches.

Index Terms—Parallel systems, Scheduling and Task ↗
DOE OSTI2016

Particle Number Concentrations for HI-SCALE Field Campaign Report

In support of the Holistic Interactions of Shallow Clouds, Aerosols, and Ecosystems (HI-SCALE) project to study new particle formation in the atmosphere, a pair of custom water condensation particle counters were provided to the second intensive field campaign, from mid-August through mid-September 2017, at the U.S. Department of Energy Southern Great Plains Atmospheric Radiation Measurement (ARM) Climate Research Facility observatory. These custom instruments were developed by Aerosol Dynamics, Inc. (Hering et al. 2017) to detect particles into the nanometer size range. Referred to as “versatile water condensation particle counter (vWCPC)”, they are water-based, laminar-flow condensational growth instruments whose lower particle size threshold can be set based on user-selected operating temperatures. For HI-SCALE, the vWCPCs were configured to measure airborne particle number concentrations in the size range from approximately 2nm to 2μm. Both were installed in the particle sizing system operated by Chongai Kuang of Brookhaven National Laboratory (BNL). One of these was operated in parallel to a TSI Model 3776, upstream of the mobility particle sizing system, to measure total ambient particle concentrations. The airborne particle concentration data from this “total particle number vWCPC” (Ntot-vWCPC) system has been reported to the ARM database. The data are reported with one-second resolution. The second vWCPC was operated in parallel with the BNL diethylene glycol instrument to count particles downstream of a separate differential mobility size analyzer. Data from this “DMA-vWCPC” system was logged by BNL, and will eventually be provided by that laboratory.

HI-SCALE, aerosol particle number concentration, v↗
DOE OSTI2015

The SprayList: a scalable relaxed priority queue

High-performance concurrent priority queues are essential for applications such as task scheduling and discrete event simulation. Unfortunately, even the best performing implementations do not scale past a number of threads in the single digits. This is because of the sequential bottleneck in accessing the elements at the head of the queue in order to perform a DeleteMin operation. In this paper, we present the SprayList, a scalable priority queue with relaxed ordering semantics. Starting from a non-blocking SkipList, the main innovation behind our design is that the DeleteMin operations avoid a sequential bottleneck by ``spraying'' themselves onto the head of the SkipList list in a coordinated fashion. The spraying is implemented using a carefully designed random walk, so that DeleteMin returns an element among the first O(p log^3 p) in the list, with high probability, where p is the number of threads. We prove that the running time of a DeleteMin operation is O(log^3 p), with high probability, independent of the size of the list. Our experiments show that the relaxed semantics allow the data structure to scale for high thread counts, comparable to a classic unordered SkipList. As a result, we observe that, for reasonably parallel workloads, the scalability benefits of relaxation considerably outweigh the additional work due to out-of-order execution.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

Parallel adaptive wavelet collocation method for PDEs

A parallel adaptive wavelet collocation method for solving a large class of Partial Differential Equations is presented. The parallelization is achieved by developing an asynchronous parallel wavelet transform, which allows one to perform parallel wavelet transform and derivative calculations with only one data synchronization at the highest level of resolution. The data are stored using tree-like structure with tree roots starting at a priori defined level of resolution. Both static and dynamic domain partitioning approaches are developed. For the dynamic domain partitioning, trees are considered to be the minimum quanta of data to be migrated between the processes. This allows fully automated and efficient handling of non-simply connected partitioning of a computational domain. Dynamic load balancing is achieved via domain repartitioning during the grid adaptation step and reassigning trees to the appropriate processes to ensure approximately the same number of grid points on each process. The parallel efficiency of the approach is discussed based on parallel adaptive wavelet-based Coherent Vortex Simulations of homogeneous turbulence with linear forcing at effective non-adaptive resolutions up to 2048{sup 3} using as many as 2048 CPU cores.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2015

Prestack reverse time migration for 3D marine reflection seismic data

Prestack reverse time migration (RTM) is a method for imaging the subsurface using the inner product of wavefield extrapolation in shot domain and in receiver domain. It is well known that RTM is better for preserving amplitudes and phases than other prestack migrations. Since 3D seismic data is huge data volume and it needs heavy computing works, it requires parallel computing in order to have a meaningful depth image of the 3D subsurface. We implemented a parallelized version of 3D RTM for prestack depth migration. The results of numerical example for 3D SEG/EAGE salt model showed good agreement with the original geological model. We applied RTM to offshore 3D seismic reflection data. The study area is 12 × 25 km with 120 survey lines. Shot and receiver spacing is 25 m and 12.5 m. The line spacing is 100 m. Shot gathers were preprocessed to enhance signal to noise ratio and velocity model was calculated from conventional stack velocity. Both of them were used to obtain 3D image using RTM. The results show reasonable subsurface image.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2021

Improving I/O Performance for Exascale Applications through Online Data Layout Reorganization

The applications being developed within the U.S. Exascale Computing Project (ECP) to run on imminent Exascale computers will generate scientific results with unprecedented fidelity and record turn-around time. Many of these codes are based on particle-mesh methods and use advanced algorithms, especially dynamic load-balancing and mesh-refinement, to achieve high performance on Exascale machines. Yet, as such algorithms improve parallel application efficiency, they raise new challenges for I/O logic due to their irregular and dynamic data distributions. Thus, while the enormous data rates of Exascale simulations already challenge existing file system write strategies, the need for efficient read and processing of generated data introduces additional constraints on the data layout strategies that can be used when writing data to secondary storage. We review these I/O challenges and introduce two online data layout reorganization approaches for achieving good tradeoffs between read and write performance. We demonstrate the benefits of using these two approaches for the ECP particle-in-cell simulation WarpX, which serves as a motif for a large class of important Exascale applications. Here, we show that by understanding application I/O patterns and carefully designing data layouts we can increase read performance by more than 80 percent.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

A Work-Efficient Parallel Sparse Matrix-Sparse Vector Multiplication Algorithm

We design and develop a work-efficient multithreaded algorithm for sparse matrix-sparse vector multiplication (SpMSpV) where the matrix, the input vector, and the output vector are all sparse. SpMSpV is an important primitive in the emerging GraphBLAS standard and is the workhorse of many graph algorithms including breadth-first search, bipartite graph matching, and maximal independent set. As thread counts increase, existing multithreaded SpMSpV algorithms can spend more time accessing the sparse matrix data structure than doing arithmetic. Our shared-memory parallel SpMSpV algorithm is work efficient in the sense that its total work is proportional to the number of arithmetic operations required. The key insight is to avoid each thread individually scan the list of matrix columns. Our algorithm is simple to implement and operates on existing column-based sparse matrix formats. It performs well on diverse matrices and vectors with heterogeneous sparsity patterns. A high-performance implementation of the algorithm attains up to 15x speedup on a 24-core Intel Ivy Bridge processor and up to 49x speedup on a 64-core Intel KNL manycore processor. In contrast to implementations of existing algorithms, the performance of our algorithm is sustained on a variety of different input types include matrices representing scale-free and high-diameter graphs.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Self-potential tomography preconditioned by particle swarm optimization—Self-potential monitoring and streamflow data acquired March 26–September 14, 2023 at East Fork Poplar Creek near Oak Ridge Tennessee

This data release contains self-potential (SP) monitoring data measured on the flood plain of East Fork Poplar Creek (East Fork) in Oak Ridge, Tennessee and streamflow data measured at streamgage EFK5.4 about 310 meters upstream from the SP monitoring site. Additionally, forward and inverse numerical modeling scripts used to model the electrical-potential field on the East Fork flood plain are provided. SP monitoring data included in this data release were measured at 39 different data-collection points on the east flood plain; 30 points were spaced 3-m apart along an 87-m profile parallel to the edge of the streambank, and 9 points were spaced 5-m apart along a 40-m profile approximately perpendicular to the streambank. The two profiles of SP data-collection points intersected at the approximate midpoint of the profile parallel to the streambank. Transient voltages were measured at each data-collection point every 60 seconds between 16:13 Eastern Standard Time (EST) on March 26, 2023, and 11:41 EST on September 14, 2023. Streamflow data included in this data release overlap the time-period of self-potential monitoring and were measured every 900 seconds between 16:23 on March 26, 2023, and 23:53 on September 14, 2023.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Axolotl: a scalable genomics library based on Apache Spark (Axolotl) v1.0.0

Axolotl is a Python library for scalable distributed genome and metagenome data analysis. Existing tools and systems that we rely on are struggling to keep up with the rapid explosion of genomic data. Compounding this issue, developing scalable solutions require a steep learning curve in parallel programming, which presents a barrier to academic researchers. While we do have scalable solutions for specific tasks, we lack comprehensive, end-to-end solutions. It's this gap in our toolkit that we aim to address with Axolotl. The Axolotl library is built for easy parallel processing, efficiently handling multiple tasks or large datasets simultaneously, and scaling up to meet the demands of extensive genomic data analysis.

Wang, Zhong↗
DOE OSTI2016

Data Acquisition with GPUs: The DAQ for the Muon $g$-$2$ Experiment at Fermilab

Graphical Processing Units (GPUs) have recently become a valuable computing tool for the acquisition of data at high rates and for a relatively low cost. The devices work by parallelizing the code into thousands of threads, each executing a simple process, such as identifying pulses from a waveform digitizer. The CUDA programming library can be used to effectively write code to parallelize such tasks on Nvidia GPUs, providing a significant upgrade in performance over CPU based acquisition systems. The muon $g$-$2$ experiment at Fermilab is heavily relying on GPUs to process its data. The data acquisition system for this experiment must have the ability to create deadtime-free records from 700 $\mu$s muon spills at a raw data rate 18 GB per second. Data will be collected using 1296 channels of $\mu$TCA-based 800 MSPS, 12 bit waveform digitizers and processed in a layered array of networked commodity processors with 24 GPUs working in parallel to perform a fast recording of the muon decays during the spill. The described data acquisition system is currently being constructed, and will be fully operational before the start of the experiment in 2017.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2026

From Raw to Curated Data: A Lakehouse Approach for Scientific Workflows

This report provides a technical overview of how to go from raw to curated data in three stages using a lakehouse approach. We focus on the application of open source tools in scientific use cases (while noting parallels to enterprise and commercial alternatives). Our goal is to provide scientific data managers and infrastructure providers with a common frame of reference for understanding and applying modern lakehouse technologies and approaches.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗
DOE OSTI2021

Error-controlled, progressive, and adaptable retrieval of scientific data with multilevel decomposition

Extreme-scale simulations and high-resolution instruments have been generating an increasing amount of data, which poses significant challenges to not only data storage during the run, but also post-processing where data will be repeatedly retrieved and analyzed for a long period of time. The challenges in satisfying a wide range of post-hoc analysis needs while minimizing the I/O overhead caused by inappropriate and/or excessive data retrieval should never be left unmanaged. In this paper, we propose a data refactoring, compressing, and retrieval framework capable of 1) fine-grained data refactoring with regard to precision; 2) incrementally retrieving and recomposing the data in terms of various error bounds; and 3) adaptively retrieving data in multi-precision and multi-resolution with respect to different analysis. With the progressive data re-composition and the adaptable retrieval algorithms, our framework significantly reduces the amount of data retrieved when multiple incremental precision are requested and/or the downstream analysis time when coarse resolution is used. Experiments show that the amount of data retrieved under the same progressively requested error bound using our framework is 64% less than that using state-of-the-art single-error-bounded approaches. Parallel experiments with up to 1, 024 cores and ~ 600 GB data in total show that our approach yields 1.36× and 2.52× performance over existing approaches in writing to and reading from persistent storage systems, respectively.

Liang, Xin↗
DOE OSTI2015

Physics-based multiscale coupling for full core nuclear reactor simulation

Numerical simulation of nuclear reactors is a key technology in the quest for improvements in efficiency, safety, and reliability of both existing and future reactor designs. Historically, simulation of an entire reactor was accomplished by linking together multiple existing codes that each simulated a subset of the relevant multiphysics phenomena. Recent advances in the MOOSE (Multiphysics Object Oriented Simulation Environment) framework have enabled a new approach: multiple domain-specific applications, all built on the same software framework, are efficiently linked to create a cohesive application. This is accomplished with a flexible coupling capability that allows for a variety of different data exchanges to occur simultaneously on high performance parallel computational hardware. Examples based on the KAIST-3A benchmark core, as well as a simplified Westinghouse AP-1000 configuration, demonstrate the power of this new framework for tackling—in a coupled, multiscale manner—crucial reactor phenomena such as CRUD-induced power shift and fuel shuffle. 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-SA license

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2023

Multiscale spatial analysis of fracture nodes in two dimensions

Spatial arrangement of fractures as a function of scale is an important component of fracture quantification for inferential and predictive modeling. Available methods that analyze fracture spatial arrangement are based on one-dimensional spacing data; therefore, they are limited to semi-parallel fractures. Such methods cannot be applied to fracture networks in higher dimensions, particularly when fractures have different orientations. Here to characterize fracture arrangements in two dimensions, we propose using Ripley’s K-function, as a method of point pattern analysis, to quantify spatial arrangement of fracture nodes. Fracture nodes, such as barycenters, intersection points, and tips, are point-based representations of fracture locations and connectivity within the fracture network. We introduce formulations for isotropic as well as directional analyses of spatial arrangement. In addition, we derive formulations for edge correction in circular and rectangular study domains. Finally, we demonstrate applications of Ripley’s K-function on two natural fracture datasets. Our proposed method supports quantification and characterization of fracture spatial arrangements that allow practitioners to build representative models of fractures in the subsurface.

02 PETROLEUM↗
DOE OSTI2015

HUBBLE FRONTIER FIELDS FIRST COMPLETE CLUSTER DATA: FAINT GALAXIES AT z ∼ 5-10 FOR UV LUMINOSITY FUNCTIONS AND COSMIC REIONIZATION

We present comprehensive analyses of faint dropout galaxies up to z ∼ 10 with the first full-depth data set of the A2744 lensing cluster and parallel fields observed by the Hubble Frontier Fields (HFF) program. We identify 54 dropouts at z ∼ 5-10 in the HFF fields and enlarge the size of the z ∼ 9 galaxy sample obtained to date. Although the number of highly magnified (μ ∼ 10) galaxies is small because of the tiny survey volume of strong lensing, our study reaches the galaxies' intrinsic luminosities comparable to the deepest-field HUDF studies. We derive UV luminosity functions with these faint dropouts, carefully evaluating by intensive simulations the combination of observational incompleteness and lensing effects in the image plane, including magnification, distortion, and multiplication of images, with the evaluation of mass model dependencies. Our results confirm that the faint-end slope, α, is as steep as –2 at z ∼ 6-8 and strengthen the evidence for the rapid decrease of UV luminosity densities, ρ{sub UV}, at z > 8 from the large z ∼ 9 sample. We examine whether the rapid ρ{sub UV} decrease trend can be reconciled with the large Thomson scattering optical depth, τ{sub e}, measured by cosmic microwave background experiments, allowing a large space of free parameters, such as an average ionizing photon escape fraction and a stellar-population-dependent conversion factor. No parameter set can reproduce both the rapid ρ{sub UV} decrease and the large τ {sub e}. It is possible that the ρ{sub UV} decrease moderates at z ≳ 11, that the free parameters significantly evolve toward high z, or that there exist additional sources of reionization such as X-ray binaries and faint active galactic nuclei.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Genetic-algorithm retrieval of the molecular alignment distribution with high-order harmonics generated from transiently aligned CO 2 molecules

Using genetic algorithm, we propose a method to retrieve the alignment distribution of transiently aligned CO 2 molecules from the high-order harmonic generation (HHG) spectra. The retrieval method is based on the quantitative rescattering (QRS) model where averaged photorecombination transition dipole can be factored out from the parallel (or perpendicular) harmonic spectra after the propagation of the harmonic fields in the gas medium. We examine how the retrieved alignment distributions are affected by uncertainty in alignment dependent ionization probability and on multiple orbital contribution to the HHG. We further confirm that alignment distribution is more accurately retrieved by using the minima in the HHG spectra driven by a long-wavelength laser. In addition, we show that earlier experimental data on the ratios between the perpendicular and the parallel HHG components of aligned CO 2 molecules are in better agreement with the QRS model if the macroscopic propagation and multiple orbital interference are included in the theoretical calculation.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2022

In-beam γ-ray spectroscopy of 32 Mg via direct reactions

Background: The nucleus 32 Mg (N=20 and Z=12) plays a central role in the so-called “island of inversion,” where in the ground states sd-shell neutrons are promoted to the fp-shell orbitals across the shell gap, resulting in the disappearance of the canonical neutron magic number N=20. Purpose: The primary goals of this work are to extend the level scheme of 32Mg, provide spin-parity assignments to excited states, and discuss the microscopic structure of each state through comparisons with theoretical calculations. Method: In-beam γ-ray spectroscopy of 32 Mg was performed using two direct-reaction probes: one-neutron (two-proton) knockout reactions on 33 Mg ( 34 Si). Final-state exclusive cross sections and parallel momentum distributions were extracted from the experimental data and compared with eikonal-based reaction model calculations combined with shell-model overlap functions. Results: Owing to the remarkable selectivity of the one-neutron and two-proton knockout reactions, a significantly updated level scheme for 32 Mg, which exhibits negative-parity intruder and positive-parity normal states, was constructed. The experimental results were confronted with four different nuclear structure models. Conclusions: In some of these models, different aspects of 32 Mg and the transition into the island of inversion are well described. However, unexplained discrepancies remain, and, even with the help of these state-of-the-art theoretical approaches, the structure of this key nucleus is not yet fully captured.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2024

ICED: An Integrated CGRA Framework Enabling DFVS-Aware Acceleration

oarse-grained reconfigurable arrays (CGRAs) are a promising solution to enable energy-efficient acceleration of applications from different domains. By leveraging reconfiguration at the functional level, they can adapt to significantly different computational patterns. Existing CGRA mapping approaches extract instruction-level parallelism, exploit loop-pipelining opportunities, guarantee the data dependency, and target high throughput of a given loop. However, the recurrence data-dependency in the DFG and the mismatch between required and available computing/communication resources complicate the mapping, and might lead to significant unbalances in the utilization of the CGRA's tiles. This results in wasted power for tiles with low utilization. Applying dynamic voltage and frequency scaling (DVFS) can potentially solve this challenge and improve energy efficiency by adjusting voltage and frequency of different tiles independently. CGRAs have also been successful in accelerating data-dependent streaming applications. However, in these applications, the execution time of each kernel in the pipeline might dynamically vary depending on the characteristics of the input. This also leads to under-utilization of resources for the dynamically changing kernels that do not limit the application throughput. DVFS can also improve energy efficiency for these applications by dynamically changing the voltage and frequency levels of tiles that host non performance-constraining kernels. This paper proposes ICEDTEA -- an integrated DVFS-aware framework to map applications on CGRAs that support power islands. ICEDTEA proposes a CGRA architecture supporting DVFS islands at varying granularity (from a single tile to a group of tiles) and the related DVFS-aware compilation and mapping toolchain. ICEDTEA is the first work that introduces DVFS support for spatio-temporal CGRAs at power-island levels. The experimental evaluation shows that ICEDTEA improves average utilization by 2.3$\times$ and energy-efficiency by 1.32$\times$ over a conventional CGRA. With streaming applications, ICEDTEA improves energy efficiency by 1.12$\times$ over a state-of-the-art CGRA that introduces partial dynamic reconfiguration to adapt to variations in kernels' throughput.

Tan, Cheng↗