Search NASA⌕ Search

SEARCH · Search NASA

Results for “high performance and distributed computing”

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 325 records · Page 18
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Physics for neuromorphic computing

Neuromorphic computing takes inspiration from the brain to create energy-efficient hardware for information processing, capable of highly sophisticated tasks. Systems built with standard electronics achieve gains in speed and energy by mimicking the distributed topology of the brain. Scaling-up such systems and improving their energy usage, speed and performance by several orders of magnitude requires a revolution in hardware. Here, we discuss how including more physics in the algorithms and nanoscale materials used for data processing could have a major impact in the field of neuromorphic computing. We review striking results that leverage physics to enhance the computing capabilities of artificial neural networks, using resistive switching materials, photonics, spintronics and other technologies. We discuss the paths that could lead these approaches to maturity, towards low-power, miniaturized chips that could infer and learn in real time.

device physics↗
DOE OSTI2025

Distributed Stochastic Optimization of a Neural Representation Network for Time-Space Tomography Reconstruction

4D time-space reconstruction of dynamic events or deforming objects using X-ray computed tomography (CT) is an important inverse problem in non-destructive evaluation. Conventional back-projection based reconstruction methods assume that the object remains static for the duration of several tens or hundreds of X-ray projection measurement images (reconstruction of consecutive limited-angle CT scans). However, this is an unrealistic assumption for many in-situ experiments that causes spurious artifacts and inaccurate morphological reconstructions of the object. To solve this problem, we propose to perform a 4D time-space reconstruction using a distributed implicit neural representation (DINR) network that is trained using a novel distributed stochastic training algorithm. Our DINR network learns to reconstruct the object at its output by iterative optimization of its network parameters such that the measured projection images best match the output of the CT forward measurement model. Here, we use a forward measurement model that is a function of the DINR outputs at a sparsely sampled set of continuous valued 4D object coordinates. Unlike previous neural representation architectures that forward and back propagate through dense voxel grids that sample the object's entire time-space coordinates, we only propagate through the DINR at a small subset of object coordinates in each iteration resulting in an order-of-magnitude reduction in memory and compute for training. DINR leverages distributed computation across several compute nodes and GPUs to produce high-fidelity 4D time-space reconstructions. We use both simulated parallel-beam and experimental cone-beam X-ray CT datasets to demonstrate the superior performance of our approach.

36 MATERIALS SCIENCE↗
DOE OSTI2014

SU-E-I-01: A Fast, Analytical Pencil Beam Based Method for First Order X-Ray Scatter Estimation of Kilovoltage Cone Beam X-Rays

Purpose: To analytically estimate first-order x-ray scatter for kV cone beam x-ray imaging with high computational efficiency. Methods: In calculating first-order scatter using the Klein-Nishina formula, we found that by integrating the point-to-point scatter along an interaction line, a “pencil-beam” scatter kernel (BSK) can be approximated to a quartic expression when the imaging field is small. This BSK model for monoenergetic, 100keV x-rays has been verified on homogeneous cube and cylinder water phantoms by comparing with the exact implementation of KN formula. For heterogeneous medium, the water-equivalent length of a BSK was acquired with an improved Siddon's ray-tracing algorithm, which was also used in calculating pre- and post- scattering attenuation. To include the electron binding effect for scattering of low-kV photons, the mean corresponding scattering angle is determined from the effective point of scattered photons of a BSK. The behavior of polyenergetic x-rays was also investigated for 120kV x-rays incident to a sandwiched infinite heterogeneous slab phantom, with the electron binding effect incorporated. Exact computation and Monte Carlo simulations were performed for comparisons, using the EGSnrc code package. Results: By reducing the 3D volumetric target (o(n{sup 3})) to 2D pencil-beams (o(n{sup 2})), the computation expense can be generally lowered by n times, which our experience verifies. The scatter distribution on a flat detector shows high agreement between the analytic BSK model and exact calculations. The pixel-to-pixel differences are within (-2%, 2%) for the homogeneous cube and cylinder phantoms and within (0, 6%) for the heterogeneous slab phantom. However, the Monte Carlo simulation shows increased deviation of the BSK model toward detector periphery. Conclusion: The proposed BSK model, accommodating polyenergetic x-rays and electron binding effect at low kV, shows great potential in efficiently estimating the first-order scatter from small imaging fields. We are investigating more thoroughly to improve performance and explore applications.

07 ISOTOPE AND RADIATION SOURCES↗
DOE OSTI2025

Parallelizing autotuning for HPC applications: Unveiling the potential of the speculation strategy in Bayesian optimization

In the exascale computing era, tuning High-Performance Computing (HPC) applications has become a significant computational challenge. Although Bayesian optimization (BO) has emerged as a promising tool for HPC performance tuning, the BO workflow is inherently sequential (i.e., one function evaluation at a time) and cannot leverage the huge amount of parallel resources present in modern supercomputers, resulting in a considerable underutilization of their computational capabilities. This paper explores the trade-off between search quality and parallelism in BO, investigating a diverse set of methods. Building upon both previous approaches from the literature and novel methodologies introduced in this work, our study provides a deep analysis to accelerate BO performance tuning. By examining a set of synthetic functions and practical HPC applications, our exploration analyzes the interaction among various BO methods for parallelization, the quantity of parallel resources, the runtime distribution of target HPC applications, and the costs associated with different search orchestration mechanisms that have been overlooked in previous studies. Compared to sequential BO, our novel methodology achieves comparable quality while demonstrating robust scalability in search time as the amount of parallel resources increases; it also outperforms a state-of-the-art tuner, which supports parallelization, achieving up to 3.67x faster search time. We provide high-value insights for practitioners seeking to leverage the power of parallel computing for efficient HPC application tuning. Additionally, to further assist researchers in accelerating the performance tuning of their HPC applications, we provide an extension of an existing open-source tuning framework that incorporates our methods.

Bayesian optimization↗
DOE OSTI2017

An experimental investigation of gas fuel injection with X-ray radiography

In this paper, an outward-opening compressed natural gas, direct injection fuel injector has been studied with single-shot x-ray radiography. Three dimensional simulations have also been performed to compliment the x-ray data. Argon was used as a surrogate gas for experimental and safety reasons. This technique allows the acquisition of a quantitative mapping of the ensemble-average and standard deviation of the projected density throughout the injection event. Two dimensional, ensemble average and standard deviation data are presented to investigate the quasi-steady-state behavior of the jet. Upstream of the stagnation zone, minimal shot-to-shot variation is observed. Downstream of the stagnation zone, bulk mixing is observed as the jet transitions to a subsonic turbulent jet. From the time averaged data, individual slices at all downstream locations are extracted and an Abel inversion was performed to compute the radial density distribution, which was interpolated to create three dimensional visualizations. The Abel reconstructions reveal that upstream of the stagnation zone, the gas forms an annulus with high argon density and large density gradients. Inside this annulus, a recirculation region with low argon density exists. Downstream, the jet transitions to a fully turbulent jet with Gaussian argon density distributions. This experimental data is intended to serve as a quantitative benchmark for simulations.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2024

Scaling the SciDAC QuantOm Workflow

As part of the Scientific Discovery through Advanced Computing (SciDAC) program, the Quantum Chromodynamics Nuclear Tomography (QuantOM) project aims to analyze data from Deep Inelastic Scattering (DIS) experiments conducted at Jefferson Lab and the upcoming Electron Ion Collider. The DIS data analysis is performed on an event-level by combining the input from theoretical and experimental nuclear physics into a single, composable workflow. The optimization itself (I.e. fitting the experimental data with theoretical predictions) is carried out by a machine / deep learning algorithm. The size of the acquired DIS data as well as the complexity of the workflow itself require that the analysis is performed across multiple GPUs on high performance computing systems, such as Polaris at Argonne National Laboratory. This presentation discusses the novelties and challenges that came along with parallelizing this workflow. Recent results are compared to common distributed training techniques.

Lersch, Daniel↗
DOE OSTI2015

Multi-scale Material Parameter Identification Using LS-DYNA® and LS-OPT®

Ever-tightening regulations on fuel economy, and the likely future regulation of carbon emissions, demand persistent innovation in vehicle design to reduce vehicle mass. Classical methods for computational mass reduction include sizing, shape and topology optimization. One of the few remaining options for weight reduction can be found in materials engineering and material design optimization. Apart from considering different types of materials, by adding material diversity and composite materials, an appealing option in automotive design is to engineer steel alloys for the purpose of reducing plate thickness while retaining sufficient strength and ductility required for durability and safety. A project to develop computational material models for advanced high strength steel is currently being executed under the auspices of the United States Automotive Materials Partnership (USAMP) funded by the US Department of Energy. Under this program, new Third Generation Advanced High Strength Steel (i.e., 3GAHSS) are being designed, tested and integrated with the remaining design variables of a benchmark vehicle Finite Element model. The objectives of the project are to integrate atomistic, microstructural, forming and performance models to create an integrated computational materials engineering (ICME) toolkit for 3GAHSS. The mechanical properties of Advanced High Strength Steels (AHSS) are controlled by many factors, including phase composition and distribution in the overall microstructure, volume fraction, size and morphology of phase constituents as well as stability of the metastable retained austenite phase. The complex phase transformation and deformation mechanisms in these steels make the well-established traditional techniques obsolete, and a multi-scale microstructure-based modeling approach following the ICME [0]strategy was therefore chosen in this project. Multi-scale modeling as a major area of research and development is an outgrowth of the Comprehensive Test Ban Treaty of 1996 which banned surface testing of nuclear devices [1]. This had the effect that experimental work was reduced from large scale tests to multiscale experiments to provide material models with validation at different length scales. In the subsequent years industry realized that multi-scale modeling and simulation-based design were transferable to the design optimization of any structural system. Horstemeyer [1] lists a number of advantages of the use of multiscale modeling. Among these are: the reduction of product development time by alleviating costly trial-and-error iterations as well as the reduction of product costs through innovations in material, product and process designs. Multi-scale modeling can reduce the number of costly large scale experiments and can increase product quality by providing more accurate predictions. Research tends to be focussed on each particular length scale, which enhances accuracy in the long term. This paper serves as an introduction to the LS-OPT and LS-DYNA methodology for multi-scale modeling. It mainly focuses on an approach to integrate material identification using material models of different length scales. As an example, a multi-scale material identification strategy, consisting of a Crystal Plasticity (CP) material model and a homogenized State Variable (SV) model, is discussed and the parameter identification of the individual material models of different length scales is demonstrated. The paper concludes with thoughts on integrating the multi-scale methodology into the overall vehicle design.

Stander, Nielen↗
DOE OSTI2014

Grids, virtualization, and clouds at Fermilab

Fermilab supports a scientific program that includes experiments and scientists located across the globe. To better serve this community, in 2004, the (then) Computing Division undertook the strategy of placing all of the High Throughput Computing (HTC) resources in a Campus Grid known as FermiGrid, supported by common shared services. In 2007, the FermiGrid Services group deployed a service infrastructure that utilized Xen virtualization, LVS network routing and MySQL circular replication to deliver highly available services that offered significant performance, reliability and serviceability improvements. This deployment was further enhanced through the deployment of a distributed redundant network core architecture and the physical distribution of the systems that host the virtual machines across multiple buildings on the Fermilab Campus. In 2010, building on the experience pioneered by FermiGrid in delivering production services in a virtual infrastructure, the Computing Sector commissioned the FermiCloud, General Physics Computing Facility and Virtual Services projects to serve as platforms for support of scientific computing (FermiCloud 6 GPCF) and core computing (Virtual Services). Lastly, this work will present the evolution of the Fermilab Campus Grid, Virtualization and Cloud Computing infrastructure together with plans for the future.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Investigation of boundary-layer wind predictions during nocturnal low-level jet events using the Weather Research and Forecasting model

Simulations of two periods featuring three consecutive low level jet (LLJ) events in the US Upper Great Plains during the autumn of 2011 were conducted to explore the impacts of various setup configurations and physical process models on simulated flow parameters within the lowest 200 m above the surface, using the Weather Research and Forecasting (WRF) model. Sensitivities of simulated flow parameters to the horizontal and vertical grid spacing, planetary boundary layer (PBL) and land surface model (LSM) physics options, were assessed. Data from a Light Detection and Ranging (lidar) system, deployed to the Weather Forecast Improvement Project (WFIP; Finley et al. 2013) were used to evaluate the accuracy of simulated wind speed and direction at 80 m above the surface, as well as their vertical distributions between 120 and 40 m, covering the typical span of contemporary tall wind turbines. All of the simulations qualitatively captured the overall diurnal cycle of wind speed and stratification, producing LLJs during each overnight period, however large discrepancies occurred at certain times for each simulation in relation to the observations. 54-member ensembles encompassing changes of the above discussed configuration parameters displayed a wide range of simulated vertical distributions of wind speed and direction, and potential temperature, reflecting highly variable representations of stratification during the weakly stable overnight conditions. Root mean square error (RMSE) statistics show that different ensemble members performed better and worse in various simulated parameters at different times, with no clearly superior configuration . Simulations using a PBL parameterization designed specifically for the stable conditions investigated herein provided superior overall simulations of wind speed at 80 m, demonstrating the efficacy of targeting improvements of physical process models in areas of known deficiencies. However, the considerable magnitudes of the RMSE values of even the best performing simulations indicate ample opportunities for further improvements.

Mirocha, Jeff D.↗
DOE OSTI2018

A computationally efficient ductile damage model accounting for nucleation and micro-inertia at high triaxialities

The computational formulation of a micro-mechanical material model for the dynamic failure of ductile metals is presented in this paper. The statistical nature of porosity initiation is accounted for by introducing an arbitrary probability density function which describes the pores nucleation pressures. Each micropore within the representative volume element is modeled as a thick spherical shell made of plastically incompressible material. The treatment of porosity by a distribution of thick-walled spheres also allows for the inclusion of micro-inertia effects under conditions of shock and dynamic loading. The second order ordinary differential equation governing the microscopic porosity evolution is solved with a robust implicit procedure. A new Chebyshev collocation method is employed to approximate the porosity distribution and remapping is used to optimize memory usage. The adaptive approximation of the porosity distribution leads to a reduction of computational time and memory usage of up to two orders of magnitude. Moreover, the proposed model affords consistent performance: changing the nucleation pressure probability density function and/or the applied strain rate does not reduce accuracy or computational efficiency of the material model. The numerical performance of the model and algorithms presented is tested against three problems for high density tantalum: single void, one-dimensional uniaxial strain, and two-dimensional plate impact. Here, the results using the integration and algorithmic advances suggest a significant improvement in computational efficiency and accuracy over previous treatments for dynamic loading conditions.

42 ENGINEERING↗
DOE OSTI2019

A hierarchical approach for modeling X-ray beamlines: application to a coherent beamline

Herein, different approaches to simulate a modern X-ray beamline are considered. Several methodologies with increasing complexity are applied to discuss the relevant parameters that quantify the beamline performance. Parameters such as flux, dimensions and intensity distribution of the focused beam, and coherence properties are obtained from simple analytical calculations to sophisticated computer simulations using ray-tracing and wave optics techniques. A latest-generation X-ray nanofocusing beamline for coherent applications (ID16A at the ESRF) has been chosen to study in detail the issues related to highly demagnifying synchrotron sources and exploiting the beam coherence. The performance of the beamline is studied for two storage rings: the old ESRF-1 (emittance 4000 pm) and the new ESRF-EBS (emittance 150 pm). In addition to traditional results in terms of flux and beam sizes, an innovative study on the partial coherence properties based on the propagation of coherent modes is presented. The different algorithms and methodologies are implemented in the software suite OASYS . These are discussed with emphasis placed upon the their benefits and limitations of each.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2023

Extreme-scale stochastic optimization and simulation via learning-enhanced decomposition and parallelization (Final Technical Report)

Stochastic optimization and simulation models ubiquitously arise in designing and operating complex service/engineering systems. They can be extreme in scale due to high-dimensional data and decisions, and can also involve decisions made sequentially in response to newly revealed data, both causing significant computational challenge. The objective of this research is to explore a unified framework that integrates machine learning with discrete optimization and risk-averse modeling, to improve the efficiency of decomposition paradigms for stochastic optimization and simulations at extreme scale. The models we consider represent a broad class of complex decision-making problems, where 0-1 or continuous decisions are made before and/or after knowing multiple sources of uncertainties that could be correlated. We will employ machine learning methods to dynamically decide and prioritize computational procedures, including cut generation, branching, and bounding of the optimal objective. Furthermore, the research will shed new lights on the traditional decomposition algorithms for extreme-scale computing. Deliverables of the research include new modeling and computational methods for advancing the state-of-the-art research in optimization and simulation, bringing many relevant risk-averse, data-driven optimization problems in practice within the range of tractability. Examples include distributed computing server scheduling and sensor deployment for monitoring critical infrastructures. Success in this effort will enable progress in solving multiple extreme-scale problems in the complex system design and operations arising from DoE missions in energy, environment, and national security.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2023

Software Quality Assurance for High Performance Computing Containers

Software containers are a key channel for delivering portable and reproducible scientific software in high performance computing (HPC) environments. HPC environments are different from other types of computing environments primarily due to usage of the message passing interface (MPI) and drivers for specialized hard- ware to enable distributed computing capabilities. This distinction directly impacts how software containers are built for HPC applications and can complicate software quality assurance efforts including portability and performance. This work introduces a strategy for building containers for HPC applications that adopts layering as a mechanism for software quality assurance. The strategy is demonstrated across three different HPC systems, two of them petaflops scale with entirely different interconnect technologies and/or processor chipsets but running the same container. Performance consequences of the containerization strategy are found to be less than 5-14% while still achieving portable and reproducible containers for HPC systems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Compressed basis GMRES on high-performance graphics processing units

Krylov methods provide a fast and highly parallel numerical tool for the iterative solution of many large-scale sparse linear systems. To a large extent, the performance of practical realizations of these methods is constrained by the communication bandwidth in current computer architectures, motivating the investigation of sophisticated techniques to avoid, reduce, and/or hide the message-passing costs (in distributed platforms) and the memory accesses (in all architectures). This article leverages Ginkgo’s memory accessor in order to integrate a communication-reduction strategy into the (Krylov) GMRES solver that decouples the storage format (i.e., the data representation in memory) of the orthogonal basis from the arithmetic precision that is employed during the operations with that basis. Given that the execution time of the GMRES solver is largely determined by the memory accesses, the cost of the datatype transforms can be mostly hidden, resulting in the acceleration of the iterative step via a decrease in the volume of bits being retrieved from memory. Together with the special properties of the orthonormal basis (whose elements are all bounded by 1), this paves the road toward the aggressive customization of the storage format, which includes some floating-point as well as fixed-point formats with mild impact on the convergence of the iterative process. We develop a high-performance implementation of the “compressed basis GMRES” solver in the Ginkgo sparse linear algebra library using a large set of test problems from the SuiteSparse Matrix Collection. We demonstrate robustness and performance advantages on a modern NVIDIA V100 graphics processing unit (GPU) of up to 50% over the standard GMRES solver that stores all data in IEEE double-precision.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Destabilizing high-capacity high entropy hydrides via earth abundant substitutions: From predictions to experimental validation

The vast chemical space of high entropy alloys (HEAs) makes trial-and-error experimental approaches for materials discovery intractable and often necessitates data-driven and/or first principles computational insights to successfully target materials with desired properties. In the context of materials discovery for hydrogen storage applications, a theoretical prediction-experimental validation approach can vastly accelerate the search for substitution strategies to destabilize high-capacity hydrides based on benchmark HEAs, e.g. TiVNbCr alloys. Here, in this study, machine learning predictions, corroborated by density functional theory calculations, predict substantial hydride destabilization with increasing substitution of earth-abundant Fe content in the (TiVNb) 75 Cr 25-x Fe x system. The as-prepared alloys crystallize in a single-phase bcc lattice for limited Fe content x < 7, while larger Fe content favors the formation of a secondary C14 Laves phase intermetallic. Short range order for alloys with x < 7 can be well described by a random distribution of atoms within the bcc lattice without lattice distortion. Hydrogen absorption experiments performed on selected alloys validate the predicted thermodynamic destabilization of the corresponding fcc hydrides and demonstrate promising lifecycle performance through reversible absorption/desorption. This demonstrates the potential of computationally expedited hydride discovery and points to further opportunities for optimizing bcc alloy ↔ fcc hydrides for practical hydrogen storage applications.

36 MATERIALS SCIENCE↗
DOE OSTI2020

3D microstructure design of lithium-ion battery electrodes assisted by X-ray nano-computed tomography and modelling

Driving range and fast charge capability of electric vehicles are heavily dependent on the 3D microstructure of lithium-ion batteries (LiBs) and substantial fundamental research is required to optimise electrode design for specific operating conditions. Here we have developed a full microstructure-resolved 3D model using a novel X-ray nano-computed tomography (CT) dual-scan superimposition technique that captures features of the carbon-binder domain. This elucidates how LiB performance is markedly affected by microstructural heterogeneities, particularly under high rate conditions. The elongated shape and wide size distribution of the active particles not only affect the lithium-ion transport but also lead to a heterogeneous current distribution and non-uniform lithiation between particles and along the through-thickness direction. Building on these insights, we propose and compare potential graded-microstructure designs for next-generation battery electrodes. To guide manufacturing of electrode architectures, in-situ X-ray CT is shown to reliably reveal the porosity and tortuosity changes with incremental calendering steps.

33 ADVANCED PROPULSION SYSTEMS↗
DOE OSTI2024

CP violation searches and distributed computing development with the Belle II Experiment at the University of Mississippi and Brookhaven National Laboratory (Final Report)

This report reflects the work performed at the University of Mississippi under the support of DOE EPSCoR grant DE‐SC0021274 during the period of September 2020 through August 2024, including a one-year, no-cost extension. A summary of the research outcomes is given, with reference to the project goals as stated in the proposal. This successful project supported the mission of the DOE High Energy Physics program by leveraging the complimentary expertise of researchers at the University of Mississippi and Brookhaven National Lab to search for CP violation in charmed baryon decays using data from the Belle II experiment and to provide vital support for Belle II distributed computing.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗