Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Data parallelism”

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 1,405 records · Page 78
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Evaluation of intracoronary hemodynamics identifies perturbations in vorticity

Coronary artery disease (CAD) is highly prevalent and associated with adverse events. Challenges have emerged in the treatment of intermediate coronary artery stenoses. These lesions are often interrogated with fractional flow reserve (FFR) testing to determine if a stenosis is likely to be causative for ischemia in a cardiac territory. This invasive test requires insertion of a pressure wire into a coronary vessel. Recently computational fluid dynamics (CFD) has been used to noninvasively assess fractional flow reserve in vessels reconstructed from medical imaging data. However, many of these simulations are unable to provide additional information about intravascular hemodynamics, including velocity, endothelial shear stress (ESS), and vorticity. We hypothesized that vorticity, which has demonstrated utility in the assessment of ventricular and aortic diseases, would also be an important hemodynamic factor in CAD. Three-dimensional (3D), patient-specific coronary artery geometries that included all vessels >1 mm in diameter were created from angiography data obtained from 10 patients who underwent diagnostic angiography and FFR testing (n = 9). A massively parallel CFD solver (HARVEY) was used to calculate coronary hemodynamic parameters including pressure, velocity, ESS, and vorticity. These simulations were validated by comparing velocity flow fields from simulation to both velocities derived from in vitro particle image velocimetry and to invasively acquired pressure wire-based data from clinical testing. There was strong agreement between findings from CFD simulations and particle image velocimetry experimental testing (p < 0.01). CFD-FFR was also highly correlated with invasively measured FFR (ρ = 0.77, p = 0.01) with an average error of 5.9 ± 0.1%. CFD-FFR also had a strong inverse correlation with the vorticity (ρ = -0.86, p = 0.001). Simulations to determine the effect of the coronary stenosis on intravascular hemodynamics demonstrated significant differences in velocity and vorticity (both p < 0.05). Further evaluation of an angiographically normal appearing non-FFR coronary vessel in patients with CAD also demonstrated differences in vorticity when compared with FFR vessels (p < 0.05). The use of highly accurate 3D CFD-derived intravascular hemodynamics provides additional information beyond pressure measurements that can be used to calculate FFR. Vorticity is one parameter that is modified by a coronary stenosis and appears to be abnormal in angiographically normal vessels in patients with CAD, highlighting a possible use-case in preventative screening for early coronary disease.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS1988

Adaptive parallel logic networks

Adaptive, self-organizing concurrent systems (ASOCS) that combine self-organization with massive parallelism for such applications as adaptive logic devices, robotics, process control, and system malfunction management, are presently discussed. In ASOCS, an adaptive network composed of many simple computing elements operating in combinational and asynchronous fashion is used and problems are specified by presenting if-then rules to the system in the form of Boolean conjunctions. During data processing, which is a different operational phase from adaptation, the network acts as a parallel hardware circuit.

Martinez, Tony R.↗
DOE OSTI2021

Parallel I/O Evaluation Techniques and Emerging HPC Workloads: A Perspective

Emerging workloads such as artificial intelligence, big data analytics and complex multi-step workflows alongside future exascale applications are anticipated future HPC workloads, which will result in a more diverse I/O system workload and even less predictable I/O behavior and access patterns. Along with the ever increasing gap between the compute and storage performance capabilities, the in-depth understanding of extreme-scale I/O behavior and the I/O performance modeling and prediction are essential tools of the large-scale I/O evaluation process for addressing the needs of extreme-scale hybrid workloads. In this survey article, we focus on the state-of-the-art of the I/O behavior and performance analysis process for HPC systems in a 5-year time window and identify future research challenges.

Neuwirth, Sarah↗
NASA NTRS1991

Highly parallel computer architecture for robotic computation

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.

Fijany, Amir↗
NASA NTRS1986

An analysis of the role of drift waves in equatorial spread F

An account is given of results of rocket measurements of the wave number spectrum of equatorial spread F irregularities, with emphasis on wavelengths less than 100 m. The measurements were made from two sounding rockets launched from Peru as part of Project Condor. The Condor density fluctuation spectra display a break at a wavelength near 100 m, identical to that found in the PLUMEX experiment (Kelley et al., 1982). The Condor data also confirm a subrange in which the density and the wave potential obey the Boltzmann relation - a strong indication of the presence of low-frequency electrostatic waves with finite wavelength parallel to the magnetic field, perhaps low-frequency drift waves as proposed by Kelley et al. The Condor data are also consistent with the previous conjecture that drift waves only exist above 300 km altitude. To investigate the difference in spectra observed over two altitude ranges, the data must be fitted to a form for the power spectrum taken from Keskinen and Ossakow (1981). The fitted spectrum, along with empirically determined growth and dissipation rates, is used to calculate the energy pumped into the spectrum at long wavelengths as well as the energy dissipated at shorter wavelengths. It is found that the energy is balanced by classical collisional effects in the low-altitude case, but energy balance in the high-altitude case requires an enhanced dissipation of about 500 times that due to classical diffusion. The model is consistent with, but does not uniquely imply, an inverse cascade of drift wave turbulence in equatorial spread F.

Labelle, J.↗
NASA NTRS2021

Extending the Licklider Transmission Protocol to Multi-Band Links

Most deep space missions return data to Earth using links operating at a single frequency band. Indeed, their data requirements are low enough that bandwidth regulations do not constrain the system. In contrast, spacecraft such as Kepler or Europa Clipper are transitioning to a new operational paradigm where engineering and science data are transmitted through simultaneous links operating at different frequency bands (henceforth termed multi-band links). This ensures, for instance, that critical data is correctly received using a well characterized X-band link, while science data at a much larger data rate can be returned efficiently (both in terms of bandwidth and energy) through a Ka-band link.Having a spacecraft establish two simultaneous links with a ground station opens a large span of potential improvements for space communications and mission operations. In this paper, we consider the problem of running a Licklider Transmission Protocol (LTP) session over a multi-band link. LTP is an implementation of a selective Automatic Repeat reQuest (ARQ) protocol, i.e. it ensures correct delivery of data over an error prone link with potentially long propagation delays. To maximize its efficiency in deep space environments, LTP operates in deferred-ACK mode and is typically included as one of the core protocols in the Delay Tolerant Networking (DTN) suite.The contributions of this paper are as follows: First, we propose an extension to LTP for multi-band links (denoted MBLTP) and sketch how it can be implemented without modifying the current definition of LTP data units. Next, we develop bounds on the performance of MBLTP when transmitting a single data file over a multi-band link. Three metrics are considered, file expected delivery time, total energy spent and bundle jitter. The results of the analytic model are first benchmarked against simulations to ensure validity, and then compared against the performance of both traditional LTP and Parallel LTP (PLTP).We demonstrate that MBLTP can significantly reduce the latency and jitter with which data products are delivered to destination over a deep space link compared to LTP at moderate energy cost. Similarly, we also demonstrate that MBLTP outperforms PLTP in all considered metrics.

Sanchez Net, Marc↗
NASA NTRS2003

Scalability of a Low-Cost Multi-Teraflop Linux Cluster for High-End Classical Atomistic and Quantum Mechanical Simulations

Scalability of a low-cost, Intel Xeon-based, multi-Teraflop Linux cluster is tested for two high-end scientific applications: Classical atomistic simulation based on the molecular dynamics method and quantum mechanical calculation based on the density functional theory. These scalable parallel applications use space-time multiresolution algorithms and feature computational-space decomposition, wavelet-based adaptive load balancing, and spacefilling-curve-based data compression for scalable I/O. Comparative performance tests are performed on a 1,024-processor Linux cluster and a conventional higher-end parallel supercomputer, 1,184-processor IBM SP4. The results show that the performance of the Linux cluster is comparable to that of the SP4. We also study various effects, such as the sharing of memory and L2 cache among processors, on the performance.

Kikuchi, Hideaki↗
DOE OSTI2020

PPINN: Parareal physics-informed neural network for time-dependent PDEs

Physics-informed neural networks (PINNs) encode physical conservation laws and prior physical knowledge into the neural networks, ensuring the correct physics is represented accurately while alleviating the need for supervised learning to a great degree. While effective for relatively short-term time integration, when long time integration of the time-dependent PDEs is sought, the time–space domain may become arbitrarily large and hence training of the neural network may become prohibitively expensive. To this end, we develop a parareal physics-informed neural network (PPINN), hence decomposing a long-time problem into many independent short-time problems supervised by an inexpensive/fast coarse-grained (CG) solver. In particular, the serial CG solver is designed to provide approximate predictions of the solution at discrete times, while initiate many fine PINNs simultaneously to correct the solution iteratively. There is a two-fold benefit from training PINNs with small-data sets rather than working on a large-data set directly, i.e., training of individual PINNs with small-data is much faster, while training the fine PINNs can be readily parallelized. Consequently, compared to the original PINN approach, the proposed PPINN approach may achieve a significant speed-up for long-time integration of PDEs, assuming that the CG solver is fast and can provide reasonable predictions of the solution, hence aiding the PPINN solution to converge in just a few iterations. To investigate the PPINN performance on solving time-dependent PDEs, we first apply the PPINN to solve the Burgers equation, and subsequently we apply the PPINN to solve a two-dimensional nonlinear diffusion–reaction equation. Furthermore, our results demonstrate that PPINNs converge in a few iterations with significant speed-ups proportional to the number of time-subdomains employed.

42 ENGINEERING↗
DOE OSTI2022

Enabling machine learning-ready HPC ensembles with Merlin

With the growing complexity of computational and experimental facilities, many scientific researchers are turning to machine learning (ML) techniques to analyze large scale ensemble data. With complexities such as multi-component workflows, heterogeneous machine architectures, parallel file systems, and batch scheduling, care must be taken to facilitate this analysis in a high performance computing (HPC) environment. Here, we present Merlin, a workflow framework to enable large ML-friendly ensembles of scientific HPC simulations. By augmenting traditional HPC with distributed compute technologies, Merlin aims to lower the barrier for scientific subject matter experts to incorporate ML into their analysis. As a producer–consumer workflow model, Merlin enables multi-machine, cross-batch job, dynamically allocated yet persistent workflows capable of utilizing surge-compute resources. Key features of Merlin are a flexible HPC-centric interface, low per-task overhead, multi-tiered fault recovery, and a hierarchical sampling algorithm that allows for $\mathscr{O}$(N) task execution and $\mathscr{O}$(N ln N) task queuing to ensembles of millions of tasks. In addition to Merlin’s design, we test the algorithm’s performance in an HPC center and demonstrate the ability to enqueue 40 million simulations in 100 s, with a 30 millisecond per-task overhead that is independent of ensemble size. Finally, we describe some example applications that Merlin has enabled on leadership-class HPC resources, such as the ML-augmented optimization of nuclear fusion experiments and the calibration of infectious disease models to study the progression of and possible mitigation strategies for COVID-19.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Interface capturing simulations of bubble population effects in PWR subchannels

As the computational power of high-performance computing (HPC) facilities grows, so too does the feasibility of using first principle based simulation to study turbulent two-phase flows within complex pressurized water reactor (PWR) geometries. Direct numerical simulation (DNS), integrated with an interface capturing method, allows for the collection of high-fidelity numerical data using advanced analysis techniques. The research presented here employs the massively parallel, finite-element based, unstructured mesh code, PHASTA, to simulate a set of two-phase bubbly flows through PWR subchannel geometries including auxiliary structures (spacer grids and mixing vanes). The main objective of the presented work is to analyze bubble dynamics and turbulence interactions at varying bubble concentrations to support the development of advanced two-phase flow closure models. Turbulent two-phase flows in PWR subchannels were simulated at hydraulic Reynolds numbers of 81,000 with bubble concentrations of 3%–15% by gas volume fraction (768–3928 resolved bubbles, respectively) and compared against a 1% void fraction case (262 bubbles) that had been previously simulated. The finite element mesh utilized for the study at higher bubble concentrations was composed of 1.55 billion elements, compared to the previous study which employed 1.11 billion elements, ensuring all turbulence scales and individual bubbles within the flow are fully resolved. For each case, the resolved initial bubble size was 0.65 mm in diameter (resolved with 25 grid points across the diameter). The simulations were analyzed to find flow features such as the mean velocity profile, bubble relative velocity and the effect of the bubbles on the turbulent conditions.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2024

Porting ATLAS Fast Calorimeter Simulation to GPUs with Performance Portable Programming Models

FastCaloSim is a parameterized simulation of the particle energy response and of the energy distribution in the ATLAS calorimeter. It is a relatively small and self-contained package with massive inherent parallelism and captures the essence of GPU offloading via important operations like data transfer, memory initialization, floating point operations, and reduction. It was identified by the High Energy Physics Center for Computational Excellence project as a good testbed for evaluating the performance and ease of portability of programming models. In this paper, we will discuss the results of our evaluation of the porting process to Kokkos, SYCL, Alpaka, OpenMP and std::par (nvc++), and compare performance on NVIDIA, AMD and Intel GPUs, as well as multicore CPUs.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Protein folding from heterogeneous unfolded state revealed by time-resolved X-ray solution scattering

One of the most challenging tasks in biological science is to understand how a protein folds. In theoretical studies, the hypothesis adopting a funnel-like free-energy landscape has been recognized as a prominent scheme for explaining protein folding in views of both internal energy and conformational heterogeneity of a protein. Despite numerous experimental efforts, however, comprehensively studying protein folding with respect to its global conformational changes in conjunction with the heterogeneity has been elusive. Here we investigate the redox-coupled folding dynamics of equine heart cytochrome c (cyt-c) induced by external electron injection by using time-resolved X-ray solution scattering. A systematic kinetic analysis unveils a kinetic model for its folding with a stretched exponential behavior during the transition toward the folded state. With the aid of the ensemble optimization method combined with molecular dynamics simulations, we found that during the folding the heterogeneously populated ensemble of the unfolded state is converted to a narrowly populated ensemble of folded conformations. These observations obtained from the kinetic and the structural analyses of X-ray scattering data reveal that the folding dynamics of cyt-c accompanies many parallel pathways associated with the heterogeneously populated ensemble of unfolded conformations, resulting in the stretched exponential kinetics at room temperature. This finding provides direct evidence with a view to microscopic protein conformations that the cyt-c folding initiates from a highly heterogeneous unfolded state, passes through still diverse intermediate structures, and reaches structural homogeneity by arriving at the folded state.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

MARBLES (Multi-scale Adaptively Refined Boltzmann LatticE Solver) [SWR-23-37]

MARBLES (Multi-scale Adaptively Refined Boltzmann LatticE Solver) is an open-source computational fluid dynamics package powered by the lattice Boltzmann equations and built on AMReX. In the lattice Boltzmann method, local collisions between meso-scale fictitious particles drive the governing equations which enables MARBLES to easily simulate flow around complex and/or moving geometry without the generation of a body-conforming mesh. Using AMReX data structures and operations ensures a high level of computational performance and parallel scaling on heterogenous architectures while also naturally supporting locally enhanced grid resolution and fidelity through automatic mesh refinement. New domains and problem definitions are easily specified through an input file with examples and guidance on all options and variables provided in the MARBLES documentation.

Henry de Frahan, Marc↗
DOE OSTI2021

Efficient exascale discretizations: High-order finite element methods

Efficient exploitation of exascale architectures requires rethinking of the numerical algorithms used in many large-scale applications. These architectures favor algorithms that expose ultra fine-grain parallelism and maximize the ratio of floating point operations to energy intensive data movement. One of the few viable approaches to achieve high efficiency in the area of PDE discretizations on unstructured grids is to use matrix-free/partially assembled high-order finite element methods, since these methods can increase the accuracy and/or lower the computational time due to reduced data motion. In this paper we provide an overview of the research and development activities in the Center for Efficient Exascale Discretizations (CEED), a co-design center in the Exascale Computing Project that is focused on the development of next-generation discretization software and algorithms to enable a wide range of finite element applications to run efficiently on future hardware. CEED is a research partnership involving more than 30 computational scientists from two US national labs and five universities, including members of the Nek5000, MFEM, MAGMA and PETSc projects. We discuss the CEED co-design activities based on targeted benchmarks, miniapps and discretization libraries and our work on performance optimizations for large-scale GPU architectures. We also provide a broad overview of research and development activities in areas such as unstructured adaptive mesh refinement algorithms, matrix-free linear solvers, high-order data visualization, and list examples of collaborations with several ECP and external applications.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

PeleLMeX: an AMR Low Mach Number Reactive Flow Simulation Code without level sub-cycling

PeleLMeX simulates chemically reacting low Mach number flows with block-structured adaptive mesh refinement (AMR). The code is built upon the AMReX library, which provides the underlying data structures and tools to manage and operate on them across massively parallel computing architectures. PeleLMeX algorithmic features are inherited from its predecessor PeleLM but key improvements allow representation of more complex physical processes. Together with its compressible flow counterpart PeleC, the thermo-chemistry library PelePhysics and the multi-physics library PeleMP, it forms the Pele suite of open-source reactive flow simulation codes.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Practical procedures for sensor quality assessment

Sensors are increasingly deployed for process monitoring and control. These produce on-line measurements at a high frequency, in parallel with low-frequency laboratory measurements. Compared to laboratory practices, sensor data quality assessment and control practices are far less structured at most utilities. This leads to inaccurate sensor data with unknown uncertainty factors.This chapter shows how to establish standard operating procedures (SOPs) to support sensor data quality assessment and control and subsequent maintenance actions by producing relevant sensor metadata. Furthermore, SOPs are provided for the most commonly used wastewater quality sensors, inspired by utility and academic best practices. This chapter builds on definitions provided in Chapter 3 and provides additional definitions specifically related to sensors maintenance. Chapter 6 complements the methods in this chapter, which are based on reference measurements, with data-analytical techniques.

Alferes, Janelcy↗
DOE OSTI2024

Indirect reactions and connection with R-matrix theory

It is often the case that nuclear reactions that are important for societal applications or basic science are difficult to measure directly in existing facilities for accelerated beams. The difficulty might be associated with an exceedingly small cross section, as the ones obtained at beam energies well below the Coulomb barrier, or with the impossibility of devising a short-lived target, as is the case for neutron-induced reactions on unstable isotopes. The fruitful line of experimental research addressing this issue with alternative (indirect) reactions has been developed in parallel to the theory needed to make the connection between the observed data and the desired cross sections of the (direct) reaction under study. Within this context, we present here an explicit connection between the indirect cross section and the R-matrix parameters which describe the energy-dependent cross section of the desired direct reaction.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2024

Turbulence, Waves, and Taylor’s Hypothesis for Heliosheath Observations

Magnetic field fluctuations measured in the heliosheath by the Voyager spacecraft are often characterized as compressible, as indicated by a strong fluctuating component parallel to the mean magnetic field. However, the interpretation of the turbulence data faces the caveat that the standard Taylor’s hypothesis is invalid because the solar wind flow velocity in the heliosheath becomes subsonic and slower than the fast magnetosonic speed, given the contributions from hot pickup ions (PUIs) in the heliosheath. We attempt to overcome this caveat by introducing a 4D frequency-wavenumber spectral modeling of turbulence, which is essentially a decomposition of different wave modes following their respective dispersion relations. Isotropic Alfvén and fast mode turbulence are considered to represent the heliosheath fluctuations. We also include two dispersive fast wave modes derived from a three-fluid theory. We find that (1) magnetic fluctuations in the inner heliosheath are less compressible than previously thought, an isotropic turbulence spectral model with about 25% in compressible fluctuation power is consistent with the observed magnetic compressibility in the heliosheath; (2) the hot PUI component and the relatively cold solar wind ions induce two dispersive fast magnetosonic wave branches in the perpendicular propagation limit, PUI fast wave may account for the spectral bump near the proton gyrofrequency in the observable spectrum; (3) it is possible that the turbulence wavenumber spectrum is not Kolmogorov-like although the observed frequency spectrum has a −5/3 power-law index, depending on the partitioning of power among the various wave modes, and this partitioning may change with wavenumber.

79 ASTRONOMY AND ASTROPHYSICS↗