Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Traditional Machine Learning”

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 775 records · Page 43
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Toward designing effective exascale scientific computing workflows: experiences and best practices

Many fields within scientific computing have embraced advances in big-data analysis and machine learning, which often requires the deployment of large, distributed and complicated workflows that may combine training neural networks, performing simulations, running inference, and performing database queries and data analysis in asynchronous, parallel and pipelined execution frameworks. Such a shift has brought into focus the need for scalable, efficient workflow management solutions with reproducibility, error and provenance handling, traceability, and checkpoint-restart capabilities, among other needs. Here, we discuss challenges and best-practices for deploying exascale-generation computational science workflows on resources at the Oak Ridge Leadership Computing Facility (OLCF). We present our experiences with large-scale deployment of distributed workflows on the Summit supercomputer, including for bioinformatics and computational biophysics, materials science, and deep learning model optimization. We also present problems and solutions created by working within a Python-centric software base on traditional HPC systems, and discuss steps that will be required before the convergence of HPC, AI, and data science can be fully realized. Our results point to a wealth of exciting new possibilities for harnessing this convergence to tackle new scientific challenges.

Coletti, Mark↗
DOE OSTI2021

Class imbalance in out-of-distribution datasets: Improving the robustness of the TextCNN for the classification of rare cancer types

In the last decade, the widespread adoption of electronic health record documentation has created huge opportunities for information mining. Natural language processing (NLP) techniques using machine and deep learning are becoming increasingly widespread for information extraction tasks from unstructured clinical notes. Disparities in performance when deploying machine learning models in the real world have recently received considerable attention. In the clinical NLP domain, the robustness of convolutional neural networks (CNNs) for classifying cancer pathology reports under natural distribution shifts remains understudied. In this research, we aim to quantify and improve the performance of the CNN for text classification on out-of-distribution (OOD) datasets resulting from the natural evolution of clinical text in pathology reports. We identified class imbalance due to different prevalence of cancer types as one of the sources of performance drop and analyzed the impact of previous methods for addressing class imbalance when deploying models in real-world domains. Our results show that our novel class-specialized ensemble technique outperforms other methods for the classification of rare cancer types in terms of macro F1 scores. We also found that traditional ensemble methods perform better in top classes, leading to higher micro F1 scores. Based on our findings, we formulate a series of recommendations for other ML practitioners on how to build robust models with extremely imbalanced datasets in biomedical NLP applications.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

Exploring PV Circularity by Modeling Socio-Technical Dynamics of Modules' End-of-Life Management: Preprint

The circular economy (CE) tackles environmental and resource scarcity issues by maximizing value retention in the economy. The concept implies design strategies such as reducing the use of materials or improving products’ durability and end-of-life (EOL) strategies such as reusing products and components and recycling materials. With an estimated 80 million tons of global cumulative EOL photovoltaic (PV) modules, applying CE principles to the PV industry could alleviate resource scarcity issues while also providing economic benefits. However, transitioning to a CE may imply changes in organizations and consumer behaviors. In this context, assessment of CE strategies may require accounting for behavioral change, a requirement that methods from complex system science such as agent-based modeling meet. Thus, this paper uses an agent-based modeling (ABM) approach to study circularity in the photovoltaics supply chain. Four types of agents are represented in the ABM: PV owners, installers, recyclers, and manufacturers. Moreover, five possible EOL options – including three CE strategies – are modeled. Departing from traditional techno-economic analysis, the model includes techno-economic factors as well as social factors to model EOL management decisions. Results show that each dollar decrease in the recycling fees improves the recycling rate by roughly 1.1%. However, excluding social factors underestimates the effect that lower recycling prices have on material circularity.

28 EE - Advanced Manufacturing Office (EE-5A)↗
DOE OSTI2022

Dynamic Earth Energy Storage: Terawatt-year, Grid-scale Energy Storage Using Planet Earth as a Thermal Battery (GeoTES): Phase I Project (Final Report)

Grid-scale energy storage has been identified by the U.S. Department of Energy’s (DOE) Energy Storage Grand Challenge as a necessary technology to support the continued build-out of intermittent renewable energy resources required to attain a carbon-free energy future. To meet this goal, the 2018 Department of Energy Research and Innovation Act mandated the creation of a comprehensive program to accelerate the development and commercialization of next-generation energy storage technologies. One of numerous energy storage technology options is the storage of excess energy as heated geothermal brine in suitable geologic formations. This concept, known as reservoir thermal energy storage (RTES), geologic thermal energy storage (GeoTES), aquifer thermal energy storage (ATES), etc., relies on the storage of thermal energy in geologic formations for recovery and use in large-scale direct use geothermal (e.g., district heating, industrial processes, etc.) and electrical power generation applications. This thermal energy is derived from excess or waste heat from any high-temperature heat source, such as concentrated solar or from conventional thermal/nuclear generation. As such, RTES can potentially play a significant role in meeting the energy storage shortfall in the coming decades. RTES can provide energy arbitrage through both the storage and production of thermal energy stored in geologic formations for direct use applications and can serve as a source of hot fluids that can be used to generate electricity to support peak demand ramping, thus easing stress on transmission and distribution. This energy storage option has geographic benefits in that energy can be stored locally or regionally depending on the various needs/loads. RTES can also be located across an enormous geographic area, without the need for a traditional hydrothermal resource but where thermal gradients and hydrogeology allow economic exploitation of subsurface heat. The work conducted for this project includes (1) a review of lessons learned from past high-temperature RTES international projects; (2) geochemical experimental investigation and numerical simulations of potential domestic sedimentary reservoirs and (3) development of a thermo-hydrological-mechanical (THM) numerical simulation tool for optimizing formation properties and design parameters to maximize thermal energy storage performance.

15 GEOTHERMAL ENERGY↗
DOE OSTI2023

Field Laboratory for Emerging Stacked Unconventional Plays in Central Appalachia

The goal of the Field Laboratory for Emerging Stacked Unconventional Plays (ESUP) in Central Appalachia project was to investigate and characterize the resource potential for multi-play production of emerging unconventional reservoirs in Central Appalachia. Project activities included drilling, logging and coring of a vertical characterization well drilled to basement to approximately 15,000 feet in depth. The data from drilling and well logs as well recovered core samples were used by the research team to support characterization of the geology and potential pay zones within the Nora Gas Field of southwestern Virginia and the greater Central Appalachian Basin. This project was led by Dr. Nino Ripepi of the Virginia Center for Coal and Energy Research (VCCER) in close collaboration with EnerVest Operating LLC (EnerVest). VCCER is housed in the Department of Mining and Minerals Engineering at Virginia Tech and was created by an Act of the Virginia General Assembly on March 30, 1977, as an interdisciplinary study, research, information, and resource facility for the Commonwealth of Virginia. EnerVest is a top tier, low-cost oil and natural gas company with a long history of traditional operating relationships with institutional investors as well as public and private companies. This report summarizes two major research activities: 1) Log and Core Analysis and 2) Numerical Modeling. Core analysis comprises the characterization efforts of the ESUP Field Laboratory and includes a discussion of the well logs run in the deep characterization wells, an initial analysis of those, the number, size and location of cores recovered, core testing and analysis done to date as well as plans for future analysis. Numerical modeling is comprised of seven major modeling efforts, including (1) Multiphysics shale transport modeling work, (2) the developed in-house compositional simulator studies, (3) reservoir modeling studies using GEM software, (4) fracture modeling work using EFRAC3D software and (5) numerical modeling efforts with ABAQUS software, (6) Numerical modeling efforts using FLAC3D and PFC3D, and (7) Automatic Machine Learning (AutoML) Studies.

02 PETROLEUM↗
DOE OSTI2020

Time-series machine-learning error models for approximate solutions to parameterized dynamical systems

This work proposes a machine-learning framework for modeling the error incurred by approximate solutions to parameterized dynamical systems. In particular, we extend the machine-learning error models (MLEM) framework proposed in Ref. Freno and Carlberg (2019) to dynamical systems. The proposed Time-Series Machine-Learning Error Modeling (T-MLEM) method constructs a regression model that maps features – which comprise error indicators that are derived from standard a posteriori error-quantification techniques – to a random variable for the approximate-solution error at each time instance. The proposed framework considers a wide range of candidate features, regression methods, and additive noise models. We consider primarily recursive regression techniques developed for time-series modeling, including both classical time-series models (e.g., autoregressive models) and recurrent neural networks (RNNs), but also analyze standard non-recursive regression techniques (e.g., feed-forward neural networks) for comparative purposes. Finally, numerical experiments conducted on multiple benchmark problems illustrate that the long short-term memory (LSTM) neural network, which is a type of RNN, outperforms other methods and yields substantial improvements in error predictions over traditional approaches.

42 ENGINEERING↗
DOE OSTI2025

Scalable edge clustering of dynamic graphs via weighted line graphs

Timestamped relational datasets consisting of records (or connections) between pairs of entities are ubiquitous in network science. For applications like peer-to-peer communication, email, various social network interactions, and computer network security, it is useful to organize these records into groups based on how and when they are occurring. Weighted line graphs offer a natural way to model how records are related in such datasets but for large real-world graph topologies, building and utilizing the line graph is prohibitively expensive. Here, we present the framework to cluster the edges of a dynamic graph via the associated line graph that contains two major contributions. The first is a method to work with the line graph implicitly and the second is a distributed scale implementation of an agglomerative hierarchical graph clustering algorithm. We outline a novel hierarchical dynamic graph edge clustering approach that efficiently breaks massive relational datasets into small sets of edges containing events at various timescales. This is in stark contrast to traditional graph clustering algorithms that prioritize highly connected (clique-like) community structures. Our approach relies on constructing a sufficient subgraph of a weighted line graph and applying a hierarchical agglomerative clustering. This approach is related to scalable techniques from spatial clustering, nonlinear-dimension reduction, topological data analysis, and draws particular inspiration from HDBSCAN. As an edge clustering, this method yields an overlapping node clustering. Our algorithm is parallelizable and we demonstrate efficient clustering of a billion-scale, real-world dynamic graph into small edge sets that correlate in topology and time. The entire clustering process for a graph with tens of billions of edges takes just a few minutes of run time on 256 nodes of a distributed compute environment. We argue how the output of the edge clustering is useful for a multitude of data visualization and powerful machine learning tasks, both involving the original massive dynamic graph data and metadata associated with the nodes and edges. Finally, we describe how this approach can be extended to dynamic hypergraphs and dynamic graphs/hypergraphs with unstructured data living on vertices and edges.

Data Analysis↗
DOE OSTI2025

Enhancing Multi-Step Reservoir Inflow Forecasting: A Time-Variant Encoder–Decoder Approach

Accurate reservoir inflow forecasting is vital for effective water resource management. Reliable forecasts enable operators to optimize storage and release strategies to meet competing sectoral demands—such as water supply, irrigation, and hydropower scheduling—while also mitigating flood and drought risks. To address this need, in this study, we propose a novel time-variant encoder–decoder (ED) model designed specifically to improve multi-step reservoir inflow forecasting, enabling accurate predictions of reservoir inflows up to seven days ahead. Unlike conventional ED-LSTM and recursive ED-LSTM models, which use fixed encoder parameters or recursively propagate predictions, our model incorporates an adaptive encoder structure that dynamically adjusts to evolving conditions at each forecast horizon. Additionally, we introduce the Expected Baseline Integrated Gradients (EB-IGs) method for variable importance analysis, enhancing interpretability of inflow by incorporating multiple baselines to capture a broader range of hydrometeorological conditions. The proposed methods are demonstrated at several diverse reservoirs across the United States. Our results show that they outperform traditional methods, particularly at longer lead times, while also offering insights into the key drivers of inflow forecasting. These advancements contribute to enhanced reservoir management through improved forecasting accuracy and practical decision-making insights under complex hydroclimatic conditions.

58 GEOSCIENCES↗
DOE OSTI2025

Additive manufacturing of metal matrix composites

Although Metal matrix composites (MMCs) are superior to most sought-after metallic alloys, their challenging fabricability has limited their widespread use in bulk-form applications. Among the many advanced fabrication techniques, Additive Manufacturing (AM), owing to its unique capabilities to produce near-net shapes, has drawn significant traction in the past two decades, especially for materials that are difficult to process using traditional methods. However, unlike pure metal/alloy systems, MMCs are highly sensitive to the processing conditions prevailing in AM techniques due to factors such as the high melting point of reinforcement particles and the potential for in-situ reactions. Therefore, it may be a while before metal matrix composites are commercially produced via AM. This review will discuss the current state-of-the-art design, fabricability, and performance of various additively manufactured MMCs. A particular focus will be on microstructural evolution and microstructure-property relationships. The most employed AM techniques, such as directed energy deposition, powder bed fusion, binder jetting, sheet lamination, and solid-state friction stir processing, are fundamentally different in terms of thermo-kinetics, forming the perspective for this review. A detailed comparison of microstructural evolution and process parameter optimization, including feedstock preparation methods and the role of machine learning and modeling among the different AM processes, is also presented. Finally, a critical evaluation of emerging AM technologies for MMCs is also provided, highlighting their potential advantages and challenges.

36 - MATERIALS SCIENCE↗
DOE OSTI2024

Flat and Level Analysis Tool (FLAT) for real-time automated segmentation and analysis of concrete slab point clouds

In the United States, the flatness and levelness of concrete floors during construction is traditionally specified by a maximum allowable gap under a 3 meter straightedge. However, the straightedge method is inexact and rarely representative of the entire floor since the technician is free to choose any location on the floor to perform the measurement. In cases requiring a higher degree of precision and repeatability, concrete floor flatness and levelness can be measured using the standard test method ASTM E1155. With the recent introduction of advanced surveying instruments such as robotic theodolites and terrestrial laser scanners (TLS), the means now exist to modernize and expedite the measurement of floor flatness and levelness. This paper details the development and demonstration of a digital tool, named the Flat and Level Analysis Tool (FLAT), to automate and expedite the segmentation and analysis of flatness and levelness from dense point cloud data of concrete floor slabs. Segmentation algorithms were developed using unsupervised machine learning to extract the set of points belonging to the concrete floor slab from a full 360 scan of a construction site. After segmentation, automated analysis algorithms report the results according to the standard method. The developed algorithms were demonstrated on a dense point cloud captured from a concrete slab-on-grade at a construction site. Results show that the digital tool can quickly provide estimates for floor flatness and levelness with minimal human involvement with comparable accuracy to manual methods.

Hayes, Nolan↗
DOE OSTI2023

TEACHING AN OLD ACCELERATOR NEW TRICKS

The Argonne Tandem Linac Accelerator System (ATLAS) has been a National User Facility since 1985. In that time, many of the systems that help operators retrieve, modify, and store beamline parameters have not kept pace with the advancement of technology. Development of a new method of storing and retrieving beamline parameters resulted in the testing and installation of a time-series database as a potential replacement for the traditional relational database. InfluxDB was selected due to its self-hosted Open-Source version availability as well as the simplicity of installation and setup. A program was written to periodically gather all accelerator parameters in the control system and store them in the time-series database. This resulted in over 13,000 distinct data points, captured at 5-minute intervals. A second test captured 35 channels on a 1-minute cadence. Graphing of the captured data is being done on Grafana, an Open-Source version is available that co-exists well with InfluxDB as the back-end. Grafana made visualizing the data simple and flexible. The testing has allowed for the use of modern graphing tools to generate new insights into operating the accelerator, as well as opened the door to building large data sets suitable for Artificial Intelligence and Machine Learning applications.

Novak, D.↗
DOE OSTI2023

Scalable in situ non-destructive evaluation of additively manufactured components using process monitoring, sensor fusion, and machine learning

Laser Powder Bed Fusion (L-PBF) Additive Manufacturing (AM) is among the metal 3D printing technologies most broadly adopted by the manufacturing industry. However, the current industry qualification paradigm for critical-application L-PBF parts relies heavily on expensive non-destructive inspection techniques, which significantly limits the use-cases of L-PBF. In situ monitoring of the process promises a less expensive alternative to ex situ testing, but existing sensor technologies and data analysis techniques struggle to detect sub-surface flaws (e.g., porosity and cracking) on production-scale L-PBF printers. In this work, an in situ NDE (INDE) system was engineered to detect subsurface flaws detected in X-Ray Computed Tomography (XCT) directly from process monitoring data. A multilayer, multimodal data input allowed the INDE system to detect numerous subsurface flaws in the size range of 200–1000µm using a novel human-in-the-loop annotation procedure. Furthermore, a framework was established for generating probability-of-detection (POD) and probability-of-false-alarm (PFA) curves compliant with NDE standards by systematically comparing instances of detected subsurface flaws to post-build XCT data. Here, we also introduce for the first time in the AM in situ sensing literature the a 90/95 – the flaw size corresponding to a 90% detection rate on the lower 95% confidence interval of the POD curve. The INDE system successfully demonstrated POD capabilities commensurate with traditional NDE methods. Traditional ML performance metrics were also shown to be inadequate for assessing the ability of the INDE system’s flaw detection performance. It is the hope of the authors that future studies will adopt the POD and PFA approach outlined here to provide better insight into the utility of process monitoring for AM.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Grand challenges in the design, manufacture, and operation of future wind turbine systems

Abstract. Wind energy is foundational for achieving 100 % renewable electricity production, and significant innovation is required as the grid expands and accommodates hybrid plant systems, energy-intensive products such as fuels, and a transitioning transportation sector. The sizable investments required for wind power plant development and integration make the financial and operational risks of change very high in all applications but especially offshore. Dependence on a high level of modeling and simulation accuracy to mitigate risk and ensure operational performance is essential. Therefore, the modeling chain from the large-scale inflow down to the material microstructure, and all the steps in between, needs to predict how the wind turbine system will respond and perform to allow innovative solutions to enter commercial application. Critical unknowns in the design, manufacturing, and operability of future turbine and plant systems are articulated, and recommendations for research action are laid out. This article focuses on the many unknowns that affect the ability to push the frontiers in the design of turbine and plant systems. Modern turbine rotors operate through the entire atmospheric boundary layer, outside the bounds of historic design assumptions, which requires reassessing design processes and approaches. Traditional aerodynamics and aeroelastic modeling approaches are pressing against the limits of applicability for the size and flexibility of future architectures and flow physics fundamentals. Offshore wind turbines have additional motion and hydrodynamic load drivers that are formidable modeling challenges. Uncertainty in turbine wakes complicates structural loading and energy production estimates, both around a single plant and for downstream plants, which requires innovation in plant operations and flow control to achieve full energy capture and load alleviation potential. Opportunities in co-design can bring controls upstream into design optimization if captured in design-level models of the physical phenomena. It is a research challenge to integrate improved materials into the manufacture of ever-larger components while maintaining quality and reducing cost. High-performance computing used in high-fidelity, physics-resolving simulations offer opportunities to improve design tools through artificial intelligence and machine learning, but even the high-fidelity tools are yet to be fully validated. Finally, key actions needed to continue the progress of wind energy technology toward even lower cost and greater functionality are recommended.

17 WIND ENERGY↗
DOE OSTI2025

Anomaly Detection in Seismic Data with Deep Learning: Application for Instrument Failure Detection and Forecasting

Seismic data quality assessment (QA) is the first and one of the most important steps before conducting any further data analysis. Traditional methods involve checking various metrics, such as spike detection and power spectral density, by setting strict thresholds or comparing data against synthetic benchmarks. However, these approaches often rely on pre-existing knowledge and assumptions about data anomalies, leading to potential misclassification of unusual cases. Here, in this study, we propose a deep autoencoder model, an unsupervised learning approach that evaluates data quality without making assumptions about normal and anomalous data, which can be used to identify deviations in recorded data that may indicate nascent instrument failure. We test the model with the U.S. International Monitoring System (IMS) seismic stations and demonstrate the capability of detecting anomalies on a monthly scale. This could prompt station operators to examine potential problems early, allowing sufficient time for instrument maintenance to prevent data outages. In addition, we use a new manually selected testing dataset to compare our model performance against two supervised machine learning (ML) approaches and a standard QA package, as baseline models. When applied to the dataset containing known data anomalies, performance of the supervised and unsupervised ML approaches is similar, with an accuracy of 88.1% for our model compared to ∼90% for the supervised ML approach and 78.2% for the standard QA package. Our model outperforms the baseline models when applied to new stations, where new types of data anomalies can be station-specific and not included in the training dataset. Finally, we show model transferability by training the model with data from the Global Seismograph Network only and applying it to the IMS network data. The results suggest that our model is generalizable and can be applied to new stations with good accuracy.

Lin, Jiun-Ting [Lawrence Livermore National Labora↗
DOE OSTI2023

Communication-Constrained Robust Control and Learning of Grid-Connected

The electric grid of things (EGoT) promises great potential for innovative grid services by tapping into vast load flexibility. However, the unique characteristics of EGoT, being a part of the cyber-physical electric power system, present both opportunities and challenges, especially concerning supply-demand balancing, stability, and communication constraints. Traditionally, centralized control was employed to ensure balance and stability in power systems. However, with the massive influx of EGoT devices, new strategies are needed to efficiently coordinate and control these distributed devices for optimal grid operations. While some studies have explored efficiency and economic models, there remains a gap in ensuring reliability under everyday operations and resilience during extreme conditions. Addressing this gap, this project develops the technology for an Energy Service Interface (ESI) that includes novel pricing, control, learning, and distributed optimization algorithms, which will enable utilities to recruit EGoT assets for crucial grid services such as load flexibility, voltage regulation, and situation-awareness. The key novelty of the proposed technology is the careful distribution of learning and control functions across utility and EGoT asset owners such that provably efficient and resilient grid operations are attained while respecting communication and information-exchange constraints. Specifically, the project team develops machine-learning enhanced load modeling methods to allow EGoT asset owners to learn their load capability and flexibility, and develops pricing-based and decentralized learning-based control so that asset owners can coordinate to meet system-wide demand-supply balance and reliability goals. For extreme situations involving high-impact, low-probability catastrophic events (termed the “black-sky” events), the team also develops (1) a “Feeder-Operating Center-on-a-Laptop” (FOCAL) software that can assist utility personnel in leveraging EGoT assets to accelerate the service recovery of damaged feeders, and (2) distributed optimization algorithms that can coordinate the operation points of EGoT devices under severe communication constraints. The proposed technology has been extensively tested and evaluated through simulations and on a testbed. In summary, as we transition into a more interconnected and digital power grid era, our project’s findings and developments offer a pivotal step toward guaranteeing both efficiency and resilience in the face of both everyday operations and rare “black-sky” events.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2023

LipidOz enables automated elucidation of lipid carbon–carbon double bond positions from ozone-induced dissociation mass spectrometry data

Lipids play essential roles in many biological processes and disease pathology, but unambiguous identification of lipids is complicated by the presence of multiple isomeric species differing by fatty acyl chain length, stereospecifically numbered (sn) position, and position/stereochemistry of double bonds. Conventional liquid chromatography-mass spectrometry (LC-MS/MS) analyses enable the determination of fatty acyl chain lengths (and in some cases sn position) and number of double bonds, but not carbon-carbon double bond positions. Ozone-induced dissociation (OzID) is a gas-phase oxidation reaction that produces characteristic fragments from lipids containing double bonds. OzID can be incorporated into ion mobility spectrometry (IMS)-MS instruments for the structural characterization of lipids, including additional isomer separation and confident assignment of double bond positions. The complexity and repetitive nature of OzID data analysis and lack of software tool support have limited the application of OzID for routine lipidomics studies. Here, we present an open-source Python tool, LipidOz, for the automated determination of lipid double bond positions from OzID-IMS-MS data, which employs a combination of traditional automation and deep learning approaches. Our results demonstrate the ability of LipidOz to robustly assign double bond positions for lipid standard mixtures and complex lipid extracts, enabling practical application of OzID for future lipidomics.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Scalable Deep Learning-Based Microarchitecture Simulation on GPUs

Cycle-accurate microarchitecture simulators are essential tools for designers to architect, estimate, optimize, and manufacture new processors that meet specific design expectations. However, conventional simulators based on discrete-event methods often require an exceedingly long time-to-solution for the simulation of applications and architectures at full complexity and scale. Given the excitement around wielding the machine learning (ML) hammer to tackle various architecture problems, there have been attempts to employ ML to perform architecture simulations, such as Ithemal and SimNet. However, the direct application of existing ML approaches to architecture simulation may be even slower due to overwhelming memory traffic and stringent sequential computation logic. This work proposes the first graphics processing unit (GPU)-based microarchitecture simulator that fully unleashes the potential of GPUs to accelerate state-of-the-art ML-based simulators. First, considering the application traces are loaded from central processing unit (CPU) to GPU for simulation, we introduce various designs to reduce the data movement cost between CPUs and GPUs. Second, we propose a parallel simulation paradigm that partitions the application trace into sub-traces to simulate them in parallel with rigorous error analysis and effective error correction mechanisms. Combined, this scalable GPU-based simulator outperforms by orders of magnitude the traditional CPU-based simulators and the state-of-the-art ML-based simulators, i.e., SimNet and Ithemal.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

A Comparison of Machine Learning Methods of Association Tested on Dense Nodal Arrays

The association of phase picks to form events is one of the fundamental components of seismology. Large and dense sensor networks, such as >1000 geophone arrays (and distributed acoustic sensing), offer unique challenges in association due to the vast numbers of observations and high likelihood of errant picks. In addition, the large number of stations can greatly increase the time it takes to perform the association. For this reason, machine learning (ML) methods might provide a more optimal method of association for such networks. In this work, we examine how well ML methods (e.g., Gaussian mixture model association, PhaseLink, and Graph Earthquake Neural Interpretation Engine) can incorporate dense seismic arrays into regional networks and how well they handle the increasing numbers of stations. Here, we test their capabilities on two dense seismic deployments, one within Rock Valley Nevada (52 nodes and a 9-station sparse local network), and the LArge-n Seismic Survey in Oklahoma dense nodal array (>1800 vertical-component geophones). Processing data from these two different styles of dense seismic deployments allows testing of how the ML algorithms can merge array data with a broader regional network, how they deal with poorly picked phases, and how they handle anthropogenic noise. We compare the ML-associated bulletins to those obtained using the Rapid Earthquake Association and Location algorithm, a more traditional method of association. We find that there are very small differences in results between the methods for small networks (<100 stations) with low pick rates. For large networks (>1000), there are enough errant picks that some of the ML methods start to create false events out of noise. We also find that the ML methods vary in computation time significantly but are all faster than the traditional method tested here.

58 GEOSCIENCES↗