Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Mathematics and Computing, Information sciences”

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 631 records · Page 35
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

The future of computing beyond Moore’s Law

Moore’s Law is a techno-economic model that has enabled the information technology industry to double the performance and functionality of digital electronics roughly every 2 years within a fixed cost, power and area. Advances in silicon lithography have enabled this exponential miniaturization of electronics, but, as transistors reach atomic scale and fabrication costs continue to rise, the classical technological driver that has underpinned Moore’s Law for 50 years is failing and is anticipated to flatten by 2025. Here, this article provides an updated view of what a post-exascale system will look like and the challenges ahead, based on our most recent understanding of technology roadmaps. It also discusses the tapering of historical improvements, and how it affects options available to continue scaling of successors to the first exascale machine. Lastly, this article covers the many different opportunities and strategies available to continue computing performance improvements in the absence of historical technology drivers. This article is part of a discussion meeting issue ‘Numerical algorithms for high-performance computational science’.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Leveraging graph clustering techniques for cyber‐physical system analysis to enhance disturbance characterisation

Abstract Cyber‐physical systems have behaviour that crosses domain boundaries during events such as planned operational changes and malicious disturbances. Traditionally, the cyber and physical systems are monitored separately and use very different toolsets and analysis paradigms. The security and privacy of these cyber‐physical systems requires improved understanding of the combined cyber‐physical system behaviour and methods for holistic analysis. Therefore, the authors propose leveraging clustering techniques on cyber‐physical data from smart grid systems to analyse differences and similarities in behaviour during cyber‐, physical‐, and cyber‐physical disturbances. Since clustering methods are commonly used in data science to examine statistical similarities in order to sort large datasets, these algorithms can assist in identifying useful relationships in cyber‐physical systems. Through this analysis, deeper insights can be shared with decision‐makers on what cyber and physical components are strongly or weakly linked, what cyber‐physical pathways are most traversed, and the criticality of certain cyber‐physical nodes or edges. This paper presents several types of clustering methods for cyber‐physical graphs of smart grid systems and their application in assessing different types of disturbances for informing cyber‐physical situational awareness. The collection of these clustering techniques provide a foundational basis for cyber‐physical graph interdependency analysis.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Digital Biosecurity Pilot Project

Security models for the bioeconomy have largely been developed on risk profiles borrowed from the financial industry and, in some cases, industrial control systems. The bioeconomy, however, has unique characteristics that require domain-specific knowledge of the risks to environmental, health and economic impact from directly targeted attacks. Key questions need to be assessed and answered for any facility engaged in bioproduction. These include: how much technical information must the attacker possess in order to target a specific process or facility? Which processes cause the largest economic impact if disrupted? Can an attacker lead companies down the wrong path of research, leading to irrecoverable losses of time, resources and capital? Can attackers disrupt venture capital strategies and affect financial returns? What are the resources required to attack key workflows, and subsequently what is the cost of defense? What strategies are effective against such attackers, and what are their cost? Can government provide an active role in assurance of material, process and results for critical bioeconomic infrastructure?

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS2003

Intelligent Machines in the 21st Century: Automating the Processes of Inference and Inquiry

The last century saw the application of Boolean algebra toward the construction of computing machines, which work by applying logical transformations to information contained in their memory. The development of information theory and the generalization of Boolean algebra to Bayesian inference have enabled these computing machines. in the last quarter of the twentieth century, to be endowed with the ability to learn by making inferences from data. This revolution is just beginning as new computational techniques continue to make difficult problems more accessible. However, modern intelligent machines work by inferring knowledge using only their pre-programmed prior knowledge and the data provided. They lack the ability to ask questions, or request data that would aid their inferences. Recent advances in understanding the foundations of probability theory have revealed implications for areas other than logic. Of relevance to intelligent machines, we identified the algebra of questions as the free distributive algebra, which now allows us to work with questions in a way analogous to that which Boolean algebra enables us to work with logical statements. In this paper we describe this logic of inference and inquiry using the mathematics of partially ordered sets and the scaffolding of lattice theory, discuss the far-reaching implications of the methodology, and demonstrate its application with current examples in machine learning. Automation of both inference and inquiry promises to allow robots to perform science in the far reaches of our solar system and in other star systems by enabling them to not only make inferences from data, but also decide which question to ask, experiment to perform, or measurement to take given what they have learned and what they are designed to understand.

Knuth, Kevin H.↗
DOE OSTI2022

Towards a more general understanding of the algorithmic utility of recurrent connections

Lateral and recurrent connections are ubiquitous in biological neural circuits. Yet while the strong computational abilities of feedforward networks have been extensively studied, our understanding of the role and advantages of recurrent computations that might explain their prevalence remains an important open challenge. Foundational studies by Minsky and Roelfsema argued that computations that require propagation of global information for local computation to take place would particularly benefit from the sequential, parallel nature of processing in recurrent networks. Such “tag propagation” algorithms perform repeated, local propagation of information and were originally introduced in the context of detecting connectedness, a task that is challenging for feedforward networks. Here, we advance the understanding of the utility of lateral and recurrent computation by first performing a large-scale empirical study of neural architectures for the computation of connectedness to explore feedforward solutions more fully and establish robustly the importance of recurrent architectures. In addition, we highlight a tradeoff between computation time and performance and construct hybrid feedforward/recurrent models that perform well even in the presence of varying computational time limitations. We then generalize tag propagation architectures to propagating multiple interacting tags and demonstrate that these are efficient computational substrates for more general computations of connectedness by introducing and solving an abstracted biologically inspired decision-making task. Our work thus clarifies and expands the set of computational tasks that can be solved efficiently by recurrent computation, yielding hypotheses for structure in population activity that may be present in such tasks.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Data Science and Machine Learning in Education

The growing role of data science (DS) and machine learning (ML) in high-energy physics (HEP) is well established and pertinent given the complex detectors, large data, sets and sophisticated analyses at the heart of HEP research. Moreover, exploiting symmetries inherent in physics data have inspired physics-informed ML as a vibrant sub-field of computer science research. HEP researchers benefit greatly from materials widely available materials for use in education, training and workforce development. They are also contributing to these materials and providing software to DS/ML-related fields. Increasingly, physics departments are offering courses at the intersection of DS, ML and physics, often using curricula developed by HEP researchers and involving open software and data used in HEP. In this white paper, we explore synergies between HEP research and DS/ML education, discuss opportunities and challenges at this intersection, and propose community activities that will be mutually beneficial.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2025

Derivation of physical equations for high-speed laser welding using large language models

It is challenging to formulate complex physical phenomena that occur in a manufacturing process, particularly when the available data are limited, rendering conventional data-driven approaches ineffective. This study aims to predict humping onset in high-speed laser welding by introducing a novel framework, namely text-to-equations generative pre-trained transformer (T2EGPT). This method leverages the capabilities of large language models (LLMs), in combination with sparse experimental data and enriched literature data, to derive an interpretable and generalizable equation for predicting humping initiation. By capturing key correlations among physical parameters, T2EGPT generates a compact and dimensionless expression that accurately predicts hump formation. The equation reveals that humping arises from the interplay between inertia-driven backward melt flow and capillary-driven surface stabilization, where inertial forces drive molten metal backward and capillary forces resist surface deformation. Furthermore, compared to traditional data-driven models, T2EGPT demonstrates enhanced predictive accuracy and cross-material transferability. More broadly, this study highlights the potential of LLMs to integrate textual information with data-driven discovery, enabling the extraction of physical laws in data-scarce scientific domains.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Multi-fidelity information fusion with concatenated neural networks

Recently, computational modeling has shifted towards the use of statistical inference, deep learning, and other data-driven modeling frameworks. Although this shift in modeling holds promise in many applications like design optimization and real-time control by lowering the computational burden, training deep learning models needs a huge amount of data. This big data is not always available for scientific problems and leads to poorly generalizable data-driven models. This gap can be furnished by leveraging information from physics-based models. Exploiting prior knowledge about the problem at hand, this study puts forth a physics-guided machine learning (PGML) approach to build more tailored, effective, and efficient surrogate models. For our analysis, without losing its generalizability and modularity, we focus on the development of predictive models for laminar and turbulent boundary layer flows. In particular, we combine the self-similarity solution and power-law velocity profile (low-fidelity models) with the noisy data obtained either from experiments or computational fluid dynamics simulations (high-fidelity models) through a concatenated neural network. We illustrate how the knowledge from these simplified models results in reducing uncertainties associated with deep learning models applied to boundary layer flow prediction problems. The proposed multi-fidelity information fusion framework produces physically consistent models that attempt to achieve better generalization than data-driven models obtained purely based on data. While we demonstrate our framework for a problem relevant to fluid mechanics, its workflow and principles can be adopted for many scientific problems where empirical, analytical, or simplified models are prevalent. In line with grand demands in novel PGML principles, this work builds a bridge between extensive physics-based theories and data-driven modeling paradigms and paves the way for using hybrid physics and machine learning modeling approaches for next-generation digital twin technologies.

42 ENGINEERING↗
DOE OSTI2021

Reproducibility Practice in High-Performance Computing: Community Survey Results

The integrity of science and engineering research is grounded in assumptions of rigor and transparency on the part of those engaging in such research. HPC community effort to strengthen rigor and transparency take the form of reproducibility efforts. In a recent survey of the SC conference community, we collected information about the SC Reproducibility Initiative activities. Here, we present the survey results in this paper. Results show that the reproducibility initiative activities have contributed to higher levels of awareness on the part of SC conference technical program participants, and hint at contributing to greater scientific impact for the published papers of the SC conference series. Stringent point-of-manuscript-submission verification is problematic for reasons we point out, as are inherent difficulties of computational reproducibility in HPC. Future efforts should better decouple the community educational goals from goals that specifically strengthen a research work's potential for long-term impact through reuse 5-10 years down the road.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

A fast particle-based approach for calibrating a 3-D model of the Antarctic ice sheet

We consider the scientifically challenging and policy-relevant task of understanding the past and projecting the future dynamics of the Antarctic ice sheet. The Antarctic ice sheet has shown a highly nonlinear threshold response to past climate forcings. Triggering such a threshold response through anthropogenic greenhouse gas emissions would drive drastic and potentially fast sea level rise with important implications for coastal flood risks. Previous studies have combined information from ice sheet models and observations to calibrate model parameters. These studies have broken important new ground but have either adopted simple ice sheet models or have limited the number of parameters to allow for the use of more complex models. These limitations are largely due to the computational challenges posed by calibration as models become more computationally intensive or when the number of parameters increases. Here, we propose a method to alleviate this problem: a fast sequential Monte Carlo method that takes advantage of the massive parallelization afforded by modern high-performance computing systems. We use simulated examples to demonstrate how our sample-based approach provides accurate approximations to the posterior distributions of the calibrated parameters. The drastic reduction in computational times enables us to provide new insights into important scientific questions, for example, the impact of Pliocene era data and prior parameter information on sea level projections. These studies would be computationally prohibitive with other computational approaches for calibration such as Markov chain Monte Carlo or emulation-based methods. We also find considerable differences in the distributions of sea level projections when we account for a larger number of uncertain parameters. For example, based on the same ice sheet model and data set, the 99th percentile of the Antarctic ice sheet contribution to sea level rise in 2300 increases from 6.5 m to 13.1 m when we increase the number of calibrated parameters from three to 11. With previous calibration methods, it would be challenging to go beyond five parameters. Here, this work provides an important next step toward improving the uncertainty quantification of complex, computationally intensive and decision-relevant models.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1999

NASA Aeronautics Multidisciplinary Analysis and Design Fellowship Program

This program began as a grant from NASA Headquarters, NGT-10025, which was in effect from 10/l/93 until 10/31/96. The remaining funding for this effort was transferred from NASA Headquarters to NASA Langley and a new grant NGT-1-52155 was issued covering the period II/l/96 to 5/15/99. This report serves as the final report of NGT-1-52155. For a number of years, Virginia Tech had been on the forefront of research in the area of multidisciplinary analysis and design. In June of 1994, faculty members from aerospace and ocean engineering, engineering science and mechanics, mechanical engineering, industrial engineering, mathematics and computer sciences, at Virginia Tech joined together to form the Multidisciplinary Analysis and Design (MAD) Center for Advanced Vehicles. The center was established with the single goal: to perform research that is relevant to the needs of the US industry and to foster collaboration between the university, government and industry. In October of 1994, the center was chosen by NASA headquarters as one of the five university centers to establish a fellowship program to develop a graduate program in multidisciplinary analysis and design. The fellowship program provides full stipend and tuition support for seven U. S. students per year during their graduate studies. The grant is currently being administered by the NMO Branch of NASA Langley. To advise us regarding the problems faced by the industry, an industrial advisory board has been formed consisting of representatives from industry as well as government laboratories. The present membership includes major aerospace companies: Aurora Flight Sciences, Boeing: Philadelphia, Boeing: Long Beach, Boeing: Seattle, Boeing: St. Louis, Cessna, Ford, General Electric, Hughes, Lockheed-Martin: Palo Alto, Northrop-Grumman, Sikorsky, smaller, aerospace software companies: Aerosoft, Phoenix Integration and Proteus Engineering, along with representatives from government agencies, including: NASA Ames, Langley and Lewis. The function of the advisory board is to channel information from its member companies to faculty members concerning problems that need research attention in the general area of multidisciplinary design optimization (MDO). The faculty and their graduate students make presentations to the board on their research. The board makes recommendations on the research and suggests new areas and problems which need attention. Many students participating in the program spend 3-6 months in industry working on their research projects. We are completing the fifth year of the fellowship program and have had four advisory board meetings in Blacksburg. Ten students have spent the three month periods in industry. In addition to the research element of the MAD Center efforts we also have an academic component. We have developed a menu of design-related graduate courses and two new courses: one in Aerospace Manufacturing and another in MDO. Some of the MAD Center activities are described on the world-wide web at http://www.aoe.vt.edu/mads.html The MAD Center represents an innovative approach for joint Industry-Government-University cooperation in the development of a comprehensive program in engineering education which addresses the design needs of industry. The following charts list detail of the grant: mission of the MAD center, faculty members, purpose of the advisory board, board members, summary of the graduate and undergraduate program, history of the fellowship program, mission of the fellowship program, requirements of MAD fellows, course requirements, students supported, advisory board participants, and MAD center research papers

Grossman, B.↗
DOE OSTI2022

FINETUNA: fine-tuning accelerated molecular simulations

Abstract Progress towards the energy breakthroughs needed to combat climate change can be significantly accelerated through the efficient simulation of atomistic systems. However, simulation techniques based on first principles, such as density functional theory (DFT), are limited in their practical use due to their high computational expense. Machine learning approaches have the potential to approximate DFT in a computationally efficient manner, which could dramatically increase the impact of computational simulations on real-world problems. However, they are limited by their accuracy and the cost of generating labeled data. Here, we present an online active learning framework for accelerating the simulation of atomic systems efficiently and accurately by incorporating prior physical information learned by large-scale pre-trained graph neural network models from the Open Catalyst Project. Accelerating these simulations enables useful data to be generated more cheaply, allowing better models to be trained and more atomistic systems to be screened. We also present a method of comparing local optimization techniques on the basis of both their speed and accuracy. Experiments on 30 benchmark adsorbate-catalyst systems show that our method of transfer learning to incorporate prior information from pre-trained models accelerates simulations by reducing the number of DFT calculations by 91%, while meeting an accuracy threshold of 0.02 eV 93% of the time. Finally, we demonstrate a technique for leveraging the interactive functionality built in to Vienna ab initio Simulation Package (VASP) to efficiently compute single point calculations within our online active learning framework without the significant startup costs. This allows VASP to work in tandem with our framework while requiring 75% fewer self-consistent cycles than conventional single point calculations. The online active learning implementation, and examples using the VASP interactive code, are available in the open source FINETUNA package on Github.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Deep Bayesian local crystallography

Abstract The advent of high-resolution electron and scanning probe microscopy imaging has opened the floodgates for acquiring atomically resolved images of bulk materials, 2D materials, and surfaces. This plethora of data contains an immense volume of information on materials structures, structural distortions, and physical functionalities. Harnessing this knowledge regarding local physical phenomena necessitates the development of the mathematical frameworks for extraction of relevant information. However, the analysis of atomically resolved images is often based on the adaptation of concepts from macroscopic physics, notably translational and point group symmetries and symmetry lowering phenomena. Here, we explore the bottom-up definition of structural units and symmetry in atomically resolved data using a Bayesian framework. We demonstrate the need for a Bayesian definition of symmetry using a simple toy model and demonstrate how this definition can be extended to the experimental data using deep learning networks in a Bayesian setting, namely rotationally invariant variational autoencoders.

36 MATERIALS SCIENCE↗
NASA NTRS2013

Planetary Protection Bioburden Analysis Program

This program is a Microsoft Access program that performed statistical analysis of the colony counts from assays performed on the Mars Science Laboratory (MSL) spacecraft to determine the bioburden density, 3-sigma biodensity, and the total bioburdens required for the MSL prelaunch reports. It also contains numerous tools that report the data in various ways to simplify the reports required. The program performs all the calculations directly in the MS Access program. Prior to this development, the data was exported to large Excel files that had to be cut and pasted to provide the desired results. The program contains a main menu and a number of submenus. Analyses can be performed by using either all the assays, or only the accountable assays that will be used in the final analysis. There are three options on the first menu: either calculate using (1) the old MER (Mars Exploration Rover) statistics, (2) the MSL statistics for all the assays, or This software implements penetration limit equations for common micrometeoroid and orbital debris (MMOD) shield configurations, windows, and thermal protection systems. Allowable MMOD risk is formulated in terms of the probability of penetration (PNP) of the spacecraft pressure hull. For calculating the risk, spacecraft geometry models, mission profiles, debris environment models, and penetration limit equations for installed shielding configurations are required. Risk assessment software such as NASA's BUMPERII is used to calculate mission PNP; however, they are unsuitable for use in shield design and preliminary analysis studies. The software defines a single equation for the design and performance evaluation of common MMOD shielding configurations, windows, and thermal protection systems, along with a description of their validity range and guidelines for their application. Recommendations are based on preliminary reviews of fundamental assumptions, and accuracy in predicting experimental impact test results. The software is programmed in Visual Basic for Applications for installation as a simple add-in for Microsoft Excel. The user is directed to a graphical user interface (GUI) that requires user inputs and provides solutions directly in Microsoft Excel workbooks. This work was done by Shannon Ryan of the USRA Lunar and Planetary Institute for Johnson Space Center. Further information is contained in a TSP (see page 1). MSC- 24582-1 Micrometeoroid and Orbital Debris (MMOD) Shield Ballistic Limit Analysis Program Lyndon B. Johnson Space Center, Houston, Texas Commercially, because it is so generic, Enigma can be used for almost any project that requires engineering visualization, model building, or animation. Models in Enigma can be exported to many other formats for use in other applications as well. Educationally, Enigma is being used to allow university students to visualize robotic algorithms in a simulation mode before using them with actual hardware. This work was done by David Shores and Sharon P. Goza of Johnson Space Center; Cheyenne McKeegan, Rick Easley, Janet Way, and Shonn Everett of MEI Technologies; Mark Manning of PTI; and Mark Guerra, Ray Kraesig, and William Leu of Tietronix Software, Inc. For further information, contact the JSC Innovation Partnerships Office at (281) 483-3809. MSC-24211-1 Spitzer Telemetry Processing System NASA's Jet Propulsion Laboratory, Pasadena, California The Spitzer Telemetry Processing System (SirtfTlmProc) was designed to address objectives of JPL's Multi-mission Image Processing Lab (MIPL) in processing spacecraft telemetry and distributing the resulting data to the science community. To minimize costs and maximize operability, the software design focused on automated error recovery, performance, and information management. The system processes telemetry from the Spitzer spacecraft and delivers Level 0 products to the Spitzer Science Center. SirtfTlmProc is a unique system with automated error notification and recovery, with a real-time continuous service that can go quiescent after periods of inactivity. The software can process 2 GB of telemetry and deliver Level 0 science products to the end user in four hours. It provides analysis tools so the operator can manage the system and troubleshoot problems. It automates telemetry processing in order to reduce staffing costs. This work was done by Alice Stanboli, Elmain M. Martinez, and James M. McAuley of Caltech for NASA's Jet Propulsion Laboratory. For more information, contact iaoffice @jpl.nasa.gov. This software is available for commercial licensing. Please contact Dan Broderick at Daniel.F. Broderick@jpl.nasa.gov. Refer to NPO-47803. NASA Tech Briefs, September 2013 29 This rapid response computer program predicts Orbiter Wing Leading Edge (WLE) damage caused by ice or foam impact during a Space Shuttle launch (Program "IMPACT2"). The program was developed after the Columbia accident in order to assess quickly WLE damage due to ice, foam, or metal impact (if any) during a Shuttle launch. IMPACT2 simulates an impact event in a few minutes for foam impactors, and in seconds for ice and metal impactors. The damage criterion is derived from results obtained from one sophisticated commercial program, which requires hours to carry out simulations of the same impact events. The program was designed to run much faster than the commercial program with prediction of projectile threshold velocities within 10 to 15% of commercial-program values. The mathematical model involves coupling of Orbiter wing normal modes of vibration to nonlinear or linear springmass models. IMPACT2 solves nonlinear or linear impact problems using classical normal modes of vibration of a target, and nonlinear/ linear time-domain equations for the projectile. Impact loads and stresses developed in the target are computed as functions of time. This model is novel because of its speed of execution. A typical model of foam, or other projectile characterized by material nonlinearities, impacting an RCC panel is executed in minutes instead of hours needed by the commercial programs. Target damage due to impact can be assessed quickly, provided that target vibration modes and allowable stress are known. This work was done by Robert Clark, Jr., Paul Cotter, and Constantine Michalopoulos of The Boeing Company for Johnson Space Center. For further information, contact the JSC Innovation Partnerships Office at (281) 483-3809. MSC-24988-1 Wing Leading Edge RCC Rapid Response Damage Prediction Tool (IMPACT2) Lyndon B. Johnson Space Center, Houston, Texas (3) the MSL statistics for only the accountable assays. Other options on the main menu include a data editing form and utility programs that produce various reports requested by the microbiologists and the project, and tools to generate the groupings for the final analyses. The analyses can be carried out in three ways: Each assay can be treated separately, the assays can be collectively treated for the whole zone as a group, or the assays can be collected in groups designated by the JPL Planetary Protection Manager. The latter approach was used to generate the final report because assays on the same equipment or similar equipment can be assumed to have been exposed to the same environment and cleaning. Thus, the statistics are improved by having a larger population, thereby reducing the standard deviation by the square root of N. For each method mentioned above, three reports are available. The first is a detailed report including all the data. This version was very useful in verifying the calculations. The second is a brief report that is similar to the full detailed report, but does not print out the data. The third is a grand total and summary report in which each assay requires only one line. For the first and second reports, most of the calculations are performed in the report section itself. For the third, all the calculations are performed directly in the query bound to the report. All the numerical results were verified by comparing them with Excel templates, then exporting the data from the Planetary Protection Analysis program to Excel.

Beaudet, Robert A.↗
DOE OSTI2022

AI-Enhanced Co-Design for Next-Generation Microelectronics: Innovating Innovation [Workshop Report]

In April 5-7, 2022, Sandia National Laboratories hosted a second virtual workshop to further explore the potential for developing AI-enhanced co-design for microelectronics (AICoM). This second piece in an ongoing workshop series again brought together two themes. The first theme, co-design for next generation microelectronics, was drawn from the 2018 Department of Energy Office of Science (DOE SC) “Basic Research Needs for Microelectronics” (BRN) report (DOE/SC, 2018, 2021), which called for a “fundamental rethinking” of the traditional design approach to microelectronics, in which subject matter experts (SMEs) in each microelectronics discipline (materials, devices, circuits, algorithms, etc.) work near-independently. Instead, the BRN called for a non-hierarchical, egalitarian vision of co-design, wherein “each scientific discipline informs and engages the others” in “parallel but intimately networked efforts to create radically new capabilities.” The second theme, exploiting and advancing artificial intelligence (AI) to support co-design for microelectronics, acknowledges the continuing breakthroughs in AI that are currently enhancing and accelerating solutions to traditional design problems in materials synthesis and processing, circuit design, and electronic design automation (EDA).

42 ENGINEERING↗
DOE OSTI2023

Accelerating network layouts using graph neural networks

Graph layout algorithms used in network visualization represent the first and the most widely used tool to unveil the inner structure and the behavior of complex networks. Current network visualization software relies on the force-directed layout (FDL) algorithm, whose high computational complexity makes the visualization of large real networks computationally prohibitive and traps large graphs into high energy configurations, resulting in hard-to-interpret “hairball” layouts. Here we use Graph Neural Networks (GNN) to accelerate FDL, showing that deep learning can address both limitations of FDL: it offers a 10 to 100 fold improvement in speed while also yielding layouts which are more informative. We analytically derive the speedup offered by GNN, relating it to the number of outliers in the eigenspectrum of the adjacency matrix, predicting that GNNs are particularly effective for networks with communities and local regularities. Finally, we use GNN to generate a three-dimensional layout of the Internet, and introduce additional measures to assess the layout quality and its interpretability, exploring the algorithm’s ability to separate communities and the link-length distribution. The novel use of deep neural networks can help accelerate other network-based optimization problems as well, with applications from reaction-diffusion systems to epidemics.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Future Trends in Nuclear Physics Computing

In nuclear physics (NP) today the study of quarks, gluons and their strong interactions extends across a broad research program at a varied range of collaborative scales, from a few collaborators up to large experiments at scales comparable to those typical of high energy physics (HEP). Overall, the software and computing efforts vary accordingly, from pragmatic do-it-yourself approaches among a few, to substantial organized software and computing activities within large experiments. With new experiments starting up and on the horizon [1], and rapidly increasing data volumes [2, 3] and processing demands even at small experiments, the NP community has in recent years been thinking about the next generation of data processing and analysis workflows that will maximize the science output. One context for this discussion has been a series of workshops, “Future Trends in Nuclear Physics Computing” [4]. The most recent in this series took place in Fall 2020, organized by the authors together with colleagues. The workshop focused on identifying the unique aspects of software and computing in NP, and discussing how the NP community could strengthen common efforts and chart a path forward for the next decade, sure to be an exciting one with rich ongoing scientific programs at Brookhaven National Laboratory (BNL), Jefferson Lab (JLab), and other NP facilities, and culminating in datataking at the Electron-Ion Collider (EIC) [5,6,7] in the early 2030s. Without claiming to present a collective view from the workshop and discussions since—fortunately this is not expected of us in this opinion editorial—we offer here our reflections on the topic, informed by the workshop and the summary we authored with our colleagues [8], as well as discussions and developments in the eventful time since.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2026

ZENN: A thermodynamics-inspired computational framework for heterogeneous data–driven modeling

Traditional entropy-based methods—such as cross-entropy loss in classification problems—have long been essential tools for representing the information uncertainty and physical disorder in data and for developing artificial intelligence algorithms. However, the rapid growth of data across various domains has introduced new challenges, particularly the integration of heterogeneous datasets with intrinsic disparities. To address this, we introduce a zentropy-enhanced neural network (ZENN), extending zentropy theory into the data science domain via intrinsic entropy, enabling more effective learning from heterogeneous data sources. ZENN simultaneously learns both energy and intrinsic entropy components, capturing the underlying structure of multisource data. To support this, we redesign the neural network architecture to better reflect the intrinsic properties and variability inherent in diverse datasets. We demonstrate the effectiveness of ZENN on classification tasks and energy landscape reconstructions, showing its superior generalization capabilities and robustness-particularly in predicting high-order derivatives. In image and text classification tasks, ZENN demonstrates superior generalization by introducing a learnable temperature variable that models latent multisource heterogeneity, allowing it to surpass state-of-the-art models on CIFAR-10/100, BBC News, and AG News. As a practical application in materials science, we employ ZENN to reconstruct the Helmholtz energy landscape of Fe3Pt using data generated from density functional theory and capture key material behaviors, including negative thermal expansion and the critical point in the temperature–pressure space. Overall, this work presents a zentropy-grounded framework for data-driven machine learning, positioning ZENN as a versatile and robust approach for scientific problems involving complex, heterogeneous datasets.

36 MATERIALS SCIENCE↗