SEARCH · Search NASA

Results for “Scientific Capabilities”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 199 records · Page 11

All sources NASA NTRS DOE OSTI

DOE OSTI2022

Hvac: Removing I/O Bottleneck for Large-Scale Deep Learning Applications

Scientific communities are increasingly adopting deep learning (DL) models in their applications to accelerate scientific discovery processes. However, with rapid growth in the computing capabilities of HPC supercomputers, large-scale DL applications have to spend a significant portion of training time performing I/O to a parallel storage system. Previous research works have investigated optimization techniques such as prefetching and caching. Unfortunately, there exist non-trivial challenges to adopting the existing solutions on HPC supercomputers for large-scale DL training applications, which include non-performance and/or failures at extreme scale, lack of portability and generality in design, complex deployment methodology, and being limited to a specific application or dataset. To address these challenges, we propose High-Velocity AI Cache (HVAC), a distributed read-cache layer that targets and fully exploits the node-local storage or near node-local storage technology. HVAC seamlessly accelerates read I/O by aggregating node-local or near node-local storage, avoiding metadata lookups and file locking while preserving portability in the application code. We deploy and evaluate HVAC on 1,024 nodes (with over 6000 NVIDIA V100 GPUS) of the Summit supercomputer. In particular, we evaluate the scalability, efficiency, accuracy, and load distribution of HVAC compared to GPFS and XFS-on-NVMe. With four different DL applications, we observe an average 25 % performance improvement atop GPFS and 9% drop against XFS-on-NVMe, which scale linearly and are considered the performance upper bound. We envision HVAC as an important caching library for upcoming HPC supercomputers such as Frontier.

Khan, Awais↗

DOE OSTI2025

Bias-Variance Trade-Off in Physics-Informed Neural Networks with Randomized Smoothing for High-Dimensional PDEs

Physics-Informed Neural Networks (PINNs) have triggered a paradigm shift in scientific computing, leveraging mesh-free properties and robust approximation capabilities. While proving effective for low-dimensional partial differential equations (PDEs), the computational cost of PINNs remains a hurdle in high-dimensional scenarios. This is particularly pronounced when computing high-order and high-dimensional derivatives in the physics-informed loss. Randomized Smoothing PINN (RS-PINN) introduces Gaussian noise for stochastic smoothing of the original neural net model, enabling the use of Monte Carlo methods for derivative approximation, which eliminates the need for costly automatic differentiation. Despite its computational efficiency, especially in the approximation of high-dimensional derivatives, RS-PINN introduces biases in both loss and gradients, negatively impacting convergence, especially when coupled with stochastic gradient descent (SGD) algorithms. We present a comprehensive analysis of biases in RS-PINN, attributing them to the nonlinearity of the Mean Squared Error (MSE) loss as well as the intrinsic nonlinearity of the PDE itself. We propose tailored bias correction techniques, delineating their application based on the order of PDE nonlinearity. The derivation of an unbiased RS-PINN allows for a detailed examination of its advantages and disadvantages compared to the biased version. Specifically, the biased version has a lower variance and runs faster than the unbiased version, but it is less accurate due to the bias. To optimize the bias-variance trade-off, we combine the two approaches in a hybrid method that balances the rapid convergence of the biased version with the high accuracy of the unbiased version. In addition to methodological contributions, we present an enhanced implementation of RS-PINN. Extensive experiments on diverse high-dimensional PDEs, including Fokker-Planck, Hamilton-Jacobi-Bellman (HJB), viscous Burgers’, Allen-Cahn, and Sine-Gordon equations, illustrate the bias-variance trade-off and highlight the effectiveness of the hybrid RS-PINN. Empirical guidelines are provided for selecting biased, unbiased, or hybrid versions, depending on the dimensionality and nonlinearity of the specific PDE problem.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

MEUMAPPS (C++ Version)

Many materials, metal alloys in particular, have features on the on micrometer or nanometer scale that have a large impact on the properties of the material. These features are known as the microstructure of the material. Understanding why and how the microstructure forms in a material is of fundamental scientific interest as well as of significant technological interest. The capability to predict microstructure evolution in a material allows the intentional design of microstructures and hence the intentional design of material properties. The phase-field method is one of the leading methods for predicting microstructure evolution. One of the most significant problems for phase-field models is their computational expense. Even limited phase-field simulations can easily require thousands of CPU core-hours to complete, which significantly limits their use. This code provides both a general framework for creating scalable, GPU-accelerated phase-field model applications as well as several applications themselves. The code is capable of using hundreds of GPUs efficiently, which greatly reduces the time required to perform simulations. The code is written with an emphasis on performance portability, that is the ability for the code to run efficiently on a number of different computing architectures without modification of the source code. The performance portability of this code is primarily enabled through the use of two libraries, Kokkos (performance portable data structures and execution patterns) and heFFTe (performance portable distributed 3D fast Fourier transforms). The code consists of a core library, applications, and tests. The core library includes shared functionality between applications. This includes interfaces with fast Fourier transform (FFT) libraries such as heFFTe, data structures based on Kokkos, file input and output capabilities, and a solver for infinitesimal strain mechanical equilibrium problems. Five applications are included in the code. The flagship application is the MEUMAPPS-SS application, which implements the Kim-Kim-Suzuki phase-field model for precipitation for an arbitrary number of phases and components in a metal alloy. Five simpler applications are also included that solve the Eshelby inclusion problem, Allen-Cahn equation, the coupled Allen-Cahn and diffusion equations, and the Cahn-Hilliard equation. The code includes two applications to solve the Cahn-Hilliard equation, one with constant-step-size first-order time integration and the second with adaptive high-order time integration.

DeWitt, Stephen [Oak Ridge National Lab. (ORNL), O↗

DOE OSTI2021

Adapting a Low-Cost and Open-Source Commercial Pipetting Robot for Nanoliter Liquid Handling

Nanoliter liquid handling capabilities in bioanalytical workflows can reduce reagent costs and improve sample processing efficiency. Here we report the successful modification of a low-cost commercially available pipetting robot, the Opentrons 1 (OT1), using inexpensive 3D-printed parts to accommodate low-nanoliter volumes. The pipetting accuracy and precision of the modified OT1 for 50 nL volumes of water was 0.2% and 2%, respectively. The accuracy and precision at 20 nL was 1% and 3.8%, respectively. The platform was successfully applied to the preparation of low-nanogram proteomic samples for liquid chromatography-mass spectrometry analysis. The modified OT1 is thus an effective platform for nanoliter liquid handling and, at a total cost of less than $6000 including the commercial liquid handler and all modifications, is far less expensive and more versatile than other platforms with similar capabilities, placing this technology within reach of a far broader scientific community.

47 OTHER INSTRUMENTATION↗

DOE OSTI2020

Argonne Leadership Computing Facility (2020 Operational Assessment Report)

In 2004, the U.S. Department of Energy’s (DOE’s) Advanced Scientific Computing Research (ASCR) program founded the Leadership Computing Facility (LCF) with a mission to provide the world’s most advanced computational resources to the open science community. The LCF is a huge investment in the nation’s scientific and technological future, inspired by a growing demand for capability computing and its impact on science and engineering.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Independent Review of Groundwater Remediation Strategy for Hexavalent Chromium and RDX Groundwater Plumes at Los Alamos National Laboratory (Rev. 1)

Site operations at the Los Alamos National Laboratory (LANL) resulted in the release of oxidized chromium, Cr(VI), into Sandia Canyon from cooling tower effluent from 1956 until 1972. The chromium traveled with the surface water approximately 3 miles downstream before migrating below ground surface. Chromium concentrations exceed 50 μg/L in the upper portion of the aquifer. Another LANL groundwater plume of concern is associated with RDX (Royal Demolition Explosives, 1,3,5-trinitro-1,3,5-triazine). Between 1951 and 1996, RDX was released to the mesa-top facilities' process water outfall, adjacent and underlying soils, and alluvial sediments, along with surface water in Cañon de Valle. Between 2000 and 2010, two remedial actions were deployed, removing much of the near-surface RDX, however, recharge due to precipitation has transported RDX into the perched-intermediate zone and into the regional aquifer. The report documents an independent technical review by scientists from the Department of Energy (DOE) Network of National Laboratories for Environmental Management and Stewardship (NNLEMS) to provide recommendations for potential near term actions to address and optimize remediation for both the Cr(VI) and RDX plumes. The proposed near-term remedial actions include design of pump and treat systems for Cr(VI) and monitoring and study for natural attenuation for RDX. The review assesses existing data, conceptual and numerical modeling, and it recommends a technical integration process to support identifying and implementing strategic, effective and efficient remedies. The DOE Environmental Management Los Alamos Field Office (EM-LA) and their cleanup contractor Newport News Nuclear-BWTX, LLC Los Alamos (N3B) provided the information required for the review. Interviews were also conducted with regulators to obtain the full spectrum of technical, regulatory and scientific perspectives. The independent review team was impressed by the capabilities, experiences, innovativeness, and insightfulness of the technical representatives from both the regulator, the New Mexico Environment Department (NMED) and N3B. Incorporation of vadose zone flow pathways in the conceptual site model (CSM) and configuring the numerical modeling for the site was generally state-of-the-practice (or better). This could be considered state-of-the-art by addressing uncertainties related to spatial extent of hydraulic windows. The reviews from the regulators were thorough and often provided useful concepts for consideration and future/study resolutions. The overarching consensus recommendation of independent review team is that the LANL groundwater plumes should be addressed in context of the emerging "management of complex sites" paradigm.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

DOE OSTI2024

Developing Multiphysics, Integrated, High-Fidelity, Massively Parallel Computational Capabilities for Fusion Applications Using MOOSE

As the need for fusion as a clean, sustainable, and abundant energy source grows internationally, so does the need for multiphysics, computational tools to model, study, and predict the complex interactions between plasma, materials, and engineering processes. These tools have a crucial role to play in solving scientific and engineering challenges and accelerating fusion energy deployment. To address these needs, modeling capabilities should enable massively parallel, multiphysics, fully integrated high-fidelity simulations of fusion systems. Additional attributes, such as being open source and modular while maintaining high software quality assurance standards will maximize impact by ensuring accessibility for all and wide acceptance, rapid expansion and development, as well as reliability, efficiency, and robustness. In this paper, we describe how the Multiphysics Object-Oriented Simulation Environment (MOOSE) framework, which has a track record of success in the fission space thanks to the attributes listed above, can be leveraged in the fusion energy field. We highlight key successes of the MOOSE application in the fission space and describe how MOOSE has been and is being applied to fusion applications in the United States---e.g., Tritium Migration Analysis Program, version 8 (TMAP8), MOOSE Fusion Module, Fusion ENergy Integrated multiphys-X (FENIX)---and the United Kingdom---e.g., AURORA, Achlys, Apollo. These efforts aim to establish a suite of tools that can be further extended to accelerate fusion energy deployment.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2023

Enterprise Risks for Scientific Software in the Post Exascale Era

The Department of Energy’s Office of Science’s Advanced Scientific Computing Research (ASCR) Office hosts three high performance computing Facilities and one high performance networking Facility. ASCR’s Facilities Division and these four Facilities established a joint working group to address the most severe threats to the software capabilities that can prevent the Facilities from supporting a diverse array of mission critical scientific research carried out by thousands of researchers from national labs, academia, and industry. This report outlines the risk categories, threat matrix calculations, and number of risks in each of the categories along with the threat levels. Two categories—programming environment/tools and system software—account for 13 out of the 20 risks encountered in the risk register. A significant risk to ASCR’s software ecosystem also arises from the possible loss of personnel and expertise past the end of the exascale computing project. The working group recommends that the software enterprise risks are reevaluated regularly, and are considered in future decision making regarding system acquisition, knowledge sharing, partnership development, and strategic planning

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Workshop on Manufacturing and Integration Challenges for Analog and Neuromorphic Computing

The U.S. Department of Energy’s (DOE’s) Advanced Manufacturing Office (AMO) held its third virtual workshop on Semiconductor R&D for Energy Efficiency on August 11-13, 2021. This public workshop – titled “Manufacturing and Integration Challenges for Analog and Neuromorphic Computing” – brought together more than 180 leading scientific and technical experts to identify challenges and opportunities to improve manufacturing capabilities for analog and neuromorphic computing architectures to significantly increase energy efficiency of microelectronic devices. The workshop featured perspectives of researchers from national laboratories, universities, government agencies, and industry with expertise in novel and emerging devices, processes, and metrology and characterization. This workshop is intended to inform an AMO research, development, demonstration, and deployment (RDD&D) plan to significantly reduce semiconductor energy use by 2030. It was co-sponsored by the Semiconductor Research Corporation (SRC) and the DOE Office of Science’s Advanced Scientific Computing Research (ASCR).

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

Secure API-Driven Research Automation to Accelerate Scientific Discovery

The Secure Scientific Service Mesh (S3M) provides API-driven infrastructure to accelerate scientific discovery through automated research workflows. By integrating near real-time streaming capabilities, intelligent workflow orchestration, and fine-grained authorization within a service mesh architecture, S3M enables secure and flexible programmatic access to high performance computing (HPC) resources. This framework allows intelligent agents and experimental facilities to dynamically provision resources and execute complex workflows, accelerating experimental lifecycles, and enabling AI-augmented autonomous science. S3M establishes a modern foundation for scientific computing infrastructure that significantly reduces traditional barriers between researchers, computational resources, and experimental facilities.

Skluzacek, Tyler [ORNL] (ORCID:0000000322424931)↗

DOE OSTI2022

Transforming Energy Through Computational Excellence: Advanced Scientific Visualization Reveals Energy Insights

The National Renewable Energy Laboratory's world-class researchers and analysts, along with the Insight Center (our state-of-the-art scientific visualization facility) make data immersion a reality, allowing users to step into and explore their data. With the rise of large, diverse, and distributed data sets, scientific visualization is now critical to the process of scientific discovery and to managing and analyzing data and extracting insights. NREL provides visualization capabilities and facilities that are supported by state-of-the-art equipment, leading-edge techniques, and expert staff.

data science↗

DOE OSTI2021

International Workshop on Next Generation Gamma-Ray Source

A workshop on The Next Generation Gamma-Ray Sources , sponsored by the Office of Nuclear Physics at the Department of Energy, was held November 17--19, 2016 in Bethesda, Maryland. The goals of the workshop were to identify basic and applied research opportunities at the frontiers of nuclear physics that would be made possible by the beam capabilities of an advanced laser Compton beam facility. To anchor the scientific vision to realistically achievable beam specifications using proven technologies, the workshop brought together experts in the fields of electron accelerators, lasers, and optics to examine the technical options for achieving the beam specifications required by the most compelling parts of the proposed research programs. An international assembly of participants included current and prospective $\gamma$-ray beam users, accelerator and light-source physicists, and federal agency program managers. Sessions were organized to foster interactions between the beam users and facility developers, allowing for information sharing and mutual feedback between the two groups. The workshop findings and recommendations are summarized in this report.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2025

Ultimate brightness of a medium-energy synchrotron light source at operational beam intensity

Synchrotron light sources are key instruments of modern science, providing unique opportunities for groundbreaking studies in diverse scientific disciplines and driving innovation in numerous scientific and technological fields. Fourth-generation light sources provide unprecedented capabilities in imaging, spectroscopy and diffraction techniques. Ultimate brightness is the key to advancing to a smaller scale, faster response, and higher data measurement and processing rate. The brightness is primarily determined by the electron beam emittance and energy spread at operational intensity. A common feature of fourth-generation synchrotrons is the short length of the electron bunches combined with a very small transverse beam size. Consequently, the high particle density leads to strong collective effects that significantly increase the emittance and limit the achievable brightness at operational beam intensity. In this article, we summarize our studies of the emittance and brightness scaled with the beam energy and intensity, taking into account the effects of intrabeam scattering, beam-impedance interaction and bunch lengthening provided by higher-harmonic RF systems to identify optimal combinations of machine and beam parameters.

36 MATERIALS SCIENCE↗

DOE OSTI2024

Distributed-Memory Sparse Deep Neural Network Inference Using Global Arrays

Partitioned Global Address Space (PGAS) models exhibit tremendous promise in developing efficient and productive distributed-memory parallel applications. They have been used extensively in scientific computations due to conveniently offering a ``shared-memory''-like model and convenient interfaces that separate communication with synchronization. Traditionally, PGAS communication models have been applied to dense/contiguously distributed data, but most modern applications depict varied levels of sparsity. Existing PGAS models require certain adaptations to support distributed sparse computations, since associated computations often require matrix arithmetic, in addition to data movement. The Global Arrays toolkit from Pacific Northwest National Laboratory (PNNL) is one of the earliest PGAS models to combine one-sided data communication and distributed matrix operations and is still used in the popular NWChem quantum chemistry suite. Recently, we have expanded the Global Arrays toolkit to support common sparse operations, like sparse matrix-dense matrix multiplies (SpMM), sparse matrix-sparse matrix multiplication (SpGEMM) and Sampled Dense-Dense Matrix Multiplication (SDDMM). As it turns out, these operations are the bedrock of sparse Deep Learning (DL); sparse deep neural networks and Graph Neural Networks (GNNs) have gained increasing attention recently in achieving speedups on training and inference with reduced memory footprints. Unlike scientific applications in High Performance Computing (HPC), modern (distributed-memory capable) DL toolkits often rely on non-standardized and closed-source vendor software optimizations, creating challenges in software-hardware co-design at scale. Our goal is to support a variety of distributed-memory sparse matrix operations and helper functions in the newly created Sparse Global Arrays (SGA), such that it is possible to build portable and productive Machine Learning scenarios for algorithm/software and hardware codesign purposes. Contemporary data-parallel schemes for training/inference are undergoing a major overhaul since model replication limits scalability and causes resource inefficiencies. As such, we have adopted tensor parallelism in decomposing the model and inputs, to mitigate memory issues. Current implementation is built on top of MPI and uses CPUs to maximize the portability across the platforms.

Distributed computing, machine learning↗

DOE OSTI2022

Overview of the MOSAiC expedition—Atmosphere

With the Arctic rapidly changing, the needs to observe, understand, and model the changes are essential. To support these needs, an annual cycle of observations of atmospheric properties, processes, and interactions were made while drifting with the sea ice across the central Arctic during the Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition from October 2019 to September 2020. An international team designed and implemented the comprehensive program to document and characterize all aspects of the Arctic atmospheric system in unprecedented detail, using a variety of approaches, and across multiple scales. These measurements were coordinated with other observational teams to explore cross-cutting and coupled interactions with the Arctic Ocean, sea ice, and ecosystem through a variety of physical and biogeochemical processes. This overview outlines the breadth and complexity of the atmospheric research program, which was organized into 4 subgroups: atmospheric state, clouds and precipitation, gases and aerosols, and energy budgets. Atmospheric variability over the annual cycle revealed important influences from a persistent large-scale winter circulation pattern, leading to some storms with pressure and winds that were outside the interquartile range of past conditions suggested by long-term reanalysis. Similarly, the MOSAiC location was warmer and wetter in summer than the reanalysis climatology, in part due to its close proximity to the sea ice edge. The comprehensiveness of the observational program for characterizing and analyzing atmospheric phenomena is demonstrated via a winter case study examining air mass transitions and a summer case study examining vertical atmospheric evolution. Overall, the MOSAiC atmospheric program successfully met its objectives and was the most comprehensive atmospheric measurement program to date conducted over the Arctic sea ice. The obtained data will support a broad range of coupled-system scientific research and provide an important foundation for advancing multiscale modeling capabilities in the Arctic.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2022

Microphysical-macrophysical interactions in low cloud systems over the Eastern North Atlantic

How low clouds over the oceans respond to changes in atmospheric greenhouse gases and aerosol particles is poorly understood. Climate model disagreement in these responses limits the accuracy of predictions of future climate. Processes governing the behavior of marine low cloud need to be improved in models and comprehensive long-term observations provide important constraints. Cloud structure and impacts on climate are shaped by processes operating on a wide range of time and space scales, which interact to modulate cloud dynamics, boundary layer structure, and precipitation. The Atmospheric Radiation Measurement (ARM) Eastern North Atlantic (ENA) site at Graciosa Island (Azores, 28°W 39°N) began taking measurements in late 2013 to address the need for comprehensive cloud, aerosol, and meteorological observations in a remote marine low cloud environment. The site straddles the boundary between the subtropical and midlatitude climate zones in the eastern North Atlantic and experiences a great diversity of meteorological and cloud conditions. In addition, the ENA site is downwind of the North American continent and is periodically impacted by continental aerosol particles transported there on the prevailing winds. In response to a 2014 proposal call, our team (ENA Site Science Team) laid out a strategy to accelerate research focused on understanding the multi-scale interactions that shape cloud properties by maximizing the utility of the observational capabilities of the ENA site. The proposed work aligned with the scientific directions of the ASR Working Groups. The team, together with collaborators, conducted a program of research focused upon multi-scale cloud processes. The team worked with scientists both within and outside the Atmospheric System Research (ASR) program, and with the ARM infrastructure to ensure that the full potential of the ENA site is realized. The suite of instruments at the ENA site has been increasing since the site’s inception and is now essentially complete. In 2019, we laid out a two-year extension of the research, and were granted an extension of the Site Science award.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2022

ARM FY2023 Radar Plan

The fundamental objective of the U.S. Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) user facility’s radar capability is to provide high-quality radar observations to the scientific user community with an overarching goal of improving the treatment of clouds and precipitation in climate models. ARM has a complement of 32 radars, including scanning and zenith-pointing varieties (Table 1). The effort to maintain and operate these radars is significant and requires a large, distributed staffing commitment. It has not been feasible to ensure the continuous, high-level operations expected by the ARM user community for all radars at all times, so priorities are established to routinely manage select radar operations every fiscal year. These priorities for FY23 are listed herein.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2022

Five years of ForTrilinos ECP

The ForTrilinos subproject of the Exascale Computing Project (ECP) was initiated to bring the capabilities and scalability of the Trilinos numerical solver collection to Fortran scientific application codes. A novel Fortran extension to the Simplified Wrapper and Interface Generator (SWIG) tool, which automatically generates Fortran bindings from existing C/C++ library code, has been applied to key Trilinos solver libraries to generate the new ForTrilinos libraries. SWIG-Fortran has additionally been used to generate new Fortran compatibility layers for additional scientific libraries and applications. This report summarizes the products and impact of the ForTrilinos subproject.

97 MATHEMATICS AND COMPUTING↗