Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “hierarchical data”

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 523 records · Page 29
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

DICER: Data Intensive Computing Environment and Runtime for Evaluating Unprecedented Scale of Geospatial-Temporal Human Mobility Data

With the significant increase in sources and volume of human mobility data through commercial data vendors as well as microsimulation of cities, the scale of geospatial-temporal data to analyze and assess for mobility characterization has grown to the level of Big Data. There are mobility related commercial organizations deploying scalable computing, but often the system architecture, workflow, and intermediate processing components are not fully disclosed in relevant scope. Current research literature has a notable lack of studies demonstrating architectures and workflows for human mobility analytics that are implemented on a TeraByte scale of geospatial-temporal data. In this context, this paper presents a hyperscale-level system solution named DICER (Data Intensive Computing Environment and Runtime) for processing and analytics of geospatial-temporal data at big data scale. Although the cluster computing architecture of DICER with Apache Spark job running on Kubernetes cluster is not new, there are innovations in the workflow, hierarchical processing logic, and a wide range of intermediate preprocessing and mobility metrics calculation. We have performed case studies to validate the effectiveness of DICER system solution by performing detailed analytics and assessment of human mobility microsimulation output at three different scopes and scale, including a usecase with 16.97 TeraByte and 259.2 Billion rows of data. In addition, we have presented another case study of utilizing DICER to perform the same mobility processing and comparative analytics on large-scale commercially available geospatial-temporal data. All these case studies validate the efficiency and usefulness of DICER in computing population mobility characteristics from geospatial-temporal trajectory data at an unprecedented scale (not only just data volume, but also combination of: number of user entities, temporal frequency, spatial resolution, data duration).

De, Debraj↗
NASA NTRS1999

"Tools For Analysis and Visualization of Large Time- Varying CFD Data Sets"

During the four years of this grant (including the one year extension), we have explored many aspects of the visualization of large CFD (Computational Fluid Dynamics) datasets. These have included new direct volume rendering approaches, hierarchical methods, volume decimation, error metrics, parallelization, hardware texture mapping, and methods for analyzing and comparing images. First, we implemented an extremely general direct volume rendering approach that can be used to render rectilinear, curvilinear, or tetrahedral grids, including overlapping multiple zone grids, and time-varying grids. Next, we developed techniques for associating the sample data with a k-d tree, a simple hierarchial data model to approximate samples in the regions covered by each node of the tree, and an error metric for the accuracy of the model. We also explored a new method for determining the accuracy of approximate models based on the light field method described at ACM SIGGRAPH (Association for Computing Machinery Special Interest Group on Computer Graphics) '96. In our initial implementation, we automatically image the volume from 32 approximately evenly distributed positions on the surface of an enclosing tessellated sphere. We then calculate differences between these images under different conditions of volume approximation or decimation.

Wilhelms, Jane↗
DOE OSTI2026

From natural language to control signals: a conceptual framework for semantic channel finding in complex experimental infrastructure

Modern experimental platforms such as particle accelerators, fusion devices, telescopes, and industrial process control systems expose tens to hundreds of thousands of control and diagnostic channels, accumulated over decades of hardware evolution. Operators and AI systems alike depend on informal expert knowledge, inconsistent naming conventions, and scattered documentation to locate the signals required for monitoring, troubleshooting, and automated control, creating a persistent bottleneck for reliability, scalability, and emerging language-model-driven interfaces. We formalize semantic channel finding, the task of mapping natural-language intent to concrete control-system signals, as a general problem in complex experimental infrastructure, and introduce a four-paradigm conceptual framework to guide architecture selection based on facility-specific data regimes. The paradigms span (i) direct in-context lookup over small, curated channel dictionaries, (ii) constrained hierarchical navigation through structured trees, (iii) interactive agent exploration using iterative reasoning and tool-based database queries, and (iv) ontology-grounded semantic search that decouples channel meaning from facility-specific naming conventions. We demonstrate the practical feasibility of each paradigm through proof-of-concept implementations at four operational facilities spanning two orders of magnitude in scale: from compact free-electron lasers to large synchrotron light sources, operating under diverse control-system architectures ranging from clean hierarchical naming schemes to legacy environments with decades of heterogeneous conventions. Where evaluated against expert-curated operational queries, these instantiations achieve 90%–97% accuracy, validating the framework’s applicability across real-world deployment scenarios. To accelerate adoption across the broader scientific and industrial control-system community, we release open-source, plug-and-play implementations of all three interactive paradigms-direct lookup, hierarchical navigation, and middle-layer exploration-within the Osprey framework, together with tools for channel database generation, interactive testing, and minimal-configuration deployment. This work establishes semantic channel finding as a foundational capability for human-centric and agentic AI interfaces at large-scale facilities, providing both a systematic framework for architecture design and practical resources to enable adoption without building custom infrastructure from scratch.

channel finding↗
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 OSTI2021

Machine Learning Assisted Safety Modeling and Analysis of Advanced Reactors

With the advances in computational power and numerical methods, analysts can now rely on first-principle simulations to predict ultra-fine details in a variety of applications. Advances in machine learning (ML) have produced algorithms that can now learn high-level abstractions via hierarchical models. This project aims to leverage advances in ML techniques and the available high-resolution simulation data to develop a novel modeling and simulation (M\&S) methodology for reactor safety analysis. While application-agnostic ML techniques are available, complex physics constraints need to be incorporated into ML techniques to build ML-based closures for computationally efficient predictive simulations. This project intends to develop a physics-guided data-driven multi-scale methodology for M\&S of advanced reactors. The project focuses on thermal fluid (T/F) phenomena, which play major roles in advanced reactor safety. Specifically, we propose a data-driven coarse-mesh turbulence model based on local flow features for the transient analysis of thermal mixing and stratification in a sodium-cooled fast reactor (SFR). The model has a coarse-mesh setup to ensure computational efficiency, while it is trained by fine-mesh computational fluid dynamics (CFD) data with Reynolds-averaged Navier-Stokes (RANS) turbulence model to ensure accuracy. Three different neural networks are developed and tested for loss-of-flow transients in the hot pool of SFR, i.e. the densely connected convolutional neural network (DCNN), long-short-term-memory network based on proper orthogonal decomposition (POD-LSTM), and the DCNN informed by LSTM (DCNN-LSTM). The performances of these three neural networks are evaluated based on baseline models. The DCNN-LSTM model has been chosen for further hyperparameter optimization. Furthermore, based on a simplified two-dimensional case, uncertainty quantification (UQ) of the developed ML-based closure are investigated with three methods, i.e. Monte Carlo dropout, deep ensemble, and Bayesian neural network. The developed ML-based turbulent viscosity closure relation based on deep ensemble is then integrated into the system analysis module SAM and serves as a term in the conservation equations. Such a SAM-ML based procedure guarantees that the obtained results are consistent with the physical constraints of the thermal-fluid system. The SAM-ML simulation on the same loss-of-flow transient showed comparable accuracy with the CFD simulation but with a much coarser mesh setup. Last but not least, the ML-based closure improvement with the support of higher-fidelity data from large eddy simulation (LES) is discussed. As a first step towards this direction, a baseline LES simulation is performed to obtain comparable data with RANS results. Based on the early results, future investigation on further improving the ML-based closure is discussed. We believe the developed approach that combines scientific machine learning with nuclear system analysis code can benefit the advanced reactor community as more accurate safety analyses will better characterize reactor safety margins and reduce licensing efforts.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
NASA NTRS1985

Using a data base management system for modelling SSME test history data

The usefulness of a data base management system (DBMS) for modelling historical test data for the complete series of static test firings for the Space Shuttle Main Engine (SSME) was assessed. From an analysis of user data base query requirements, it became clear that a relational DMBS which included a relationally complete query language would permit a model satisfying the query requirements. Representative models and sample queries are discussed. A list of environment-particular evaluation criteria for the desired DBMS was constructed; these criteria include requirements in the areas of user-interface complexity, program independence, flexibility, modifiability, and output capability. The evaluation process included the construction of several prototype data bases for user assessement. The systems studied, representing the three major DBMS conceptual models, were: MIRADS, a hierarchical system; DMS-1100, a CODASYL-based network system; ORACLE, a relational system; and DATATRIEVE, a relational-type system.

Abernethy, K.↗
DOE OSTI2020

Hierarchical Phenomena in Multicomponent Liquids: Simulation Methods, Analysis, Chemistry

Complex, multicomponent, solutions have often been studied solely through the lens of specific applications of interest. Yet advances to both simulation methodologies (enhanced sampling, etc.) and analysis techniques (network analysis algorithms and others), are creating a trove of data that reveal transcending characteristics across vast compositional phase space. This perspective discusses technical considerations of the reliable and accurate simulations of complex solutions, followed by the advances to analysis algorithms that elucidate coupling of different length and timescale behavior (hierarchical phenomena). The different manifestations of hierarchical phenomena are presented across an array of solution environments, emphasizing fundamental and ongoing science questions. With a more advanced molecular understanding in hand, a quintessential application (solvent extraction) is discussed, where significant opportunities exist to re-imagine the technical scope of an established technology.

Servis, Michael J.↗
DOE OSTI2022

A Hierarchical Local Electricity Market for a DER-rich Grid Edge

We report with increasing penetration of distributed energy resources (DER) in the distribution system, it is critical to design market structures that enable smooth integration of DERs. A hierarchical local electricity market (LEM) structure is proposed in this paper with a secondary market (SM) at the lower level representing secondary feeders and a primary market (PM) at the upper level, representing primary feeders, in order to effectively use DERs to increase grid efficiency and resilience. The lower level SM enforces budget, power balance, and flexibility constraints and accounts for costs related to consumers, such as their disutility, flexibility limits, and commitment reliability, while the upper level PM enforces grid physics constraints such as power balance and capacity limits, and also minimizes line losses. The hierarchical LEM is extensively evaluated using a modified IEEE-123 bus with high DER penetration, with each primary feeder consisting of at least three secondary feeders. Data from a GridLAB-D model is used to emulate realistic power injections and load profiles over the course of 24 hours. The performance of the LEM is illustrated by delineating the family of power-injection profiles across the primary and secondary feeders as well as corresponding local electricity tariffs that vary across the distribution grid. Through numerical simulations, the hierarchical LEM is shown to improve the efficiency of the market in terms of lowering overall costs, including both the distribution-level locational marginal prices (d-LMP) as well as retail tariffs paid by customers. Together, it represents an overall framework for a Distribution System Operator (DSO) who can provide the oversight for the entire LEM.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRS2018

Enriching the Twitter Stream Increasing Data Mining Yield and Quality Using Machine Learning

Social media data streams are important sources of real-time and historical global information for science applications. At the NASA Goddard Earth Sciences Data and Information Services Center (GES DISC), we are exploring the Twitter data stream for its potential in augmenting the validation program of NASA Earth science missions, specifically the Global Precipitation Measurement (GPM) mission. We have implemented a tweet processing infrastructure that outputs classified precipitation tweets. Inputs are "passive" tweets, along with a smaller number of tweets from "active" participants, i.e., those knowingly contributing to our effort. The "active" tweets, presumably of higher quality, enrich the Twitter stream. "Active" sources include data scraped from other social media (e.g., public Facebook posts) and data from existing crowdsourcing programs (e.g., mPING reports). In addition, there is likely relevant precipitation information in images and documents that are the end points of links often included in tweets. Information derived from these "active" sources could then be tweeted into the Twitter stream, thus enriching its quality. The objective of our current work is to mine these tweet­ linked images and documents, using neural networks, to increase the information content and quality related to precipitation. For images, we classified them as either precipitation-related or not. For training and validation, we used images obtained via the Google custom search API. We created two models: (1) by training a simple Convolutional Neural Network and (2) by using transfer learning principles to adapt a pre-trained object recognition model. For documents, both those linked to tweets and the tweet contents, we trained Hierarchical Attention Networks to determine precipitation occurrence, type, and intensity. For training and validation, we used a keyword-filtered tweet data set labelled with ground truth data from Dark Sky (an API to retrieve weather-related labels) and the National Severe Storms Laboratory's Multi­ Radar/Multi-Sensor (MRMS) system. Our results demonstrated the efficacy of our machine learning approaches for enriching the Twitter stream, to derive information potentially useful for validation of earth science satellite data.

Albayrak, Arif↗
DOE OSTI2025

Chemist: A Domain-Specific Language by Chemists for Chemists

Managing the complexity of quantum chemistry (QC) software is key to ensuring it remains accessible, maintainable, and reusable. Noticeably missing from the QC ecosystem are modules targeting bottleneck routines. Here we argue that this is likely due to the difficulty in defining interfaces for such modules. To that end, we introduce the open-source, publicly available Chemist library https://github.com/NWChemEx/Chemist. Chemist is a domain-specific language targeting the QC domain. Chemist has been developed focusing on performance and user-friendliness. Using Chemist, QC tasks are defined using familiar domain concepts such as molecules, wave functions, and operators. The domain objects are hierarchical to ensure a systematic encapsulation of information. Key features of Chemist include: extensibility, the ability to alias existing data, and the ability to succinctly define many common QC tasks. The usefulness of Chemist is demonstrated by discussing the interface of NWChemEx’s Fock build module and by showcasing a proof-of-concept self-consistent field algorithm containing uncertainty propagation.

Algorithms↗
DOE OSTI2025

Chemist

Chemist is a domain-specific language targeting the QC domain. Chemist has been developed focusing on performance and user-friendliness. Using Chemist, QC tasks are defined using familiar domain concepts such as molecules, wave functions, and operators. The domain objects are hierarchical to ensure a systematic encapsulation of information. Key features of Chemist include: extensibility, the ability to alias existing data, and the ability to succinctly define many common QC tasks.

Richard, Ryan [Ames Laboratory (AMES), Ames, IA (U↗
NASA NTRS1974

Data on the noise vibrations of modern traction locomotives

A survey is presented of data on noise and vibration sources in modern locomotives and their influence on engine drivers. An attempt is made hierarchize noise and vibration sources in terms of importance and to correlate the noise level with the influence of noise on the engine drivers' organ of hearing. Some possible recommendations are outlined for reducing the level of these noxae in order to improve the acoustic comfort of engine drivers.

Paslaru, V.↗
DOE OSTI2023

Hierarchical reconstruction of 3D well-connected porous media from 2D exemplars using statistics-informed neural network

The relationships between porous microstructures and transport properties are of fundamental importance in various scientific and engineering applications. Due to the intricacy, stochasticity and heterogeneity of porous media, reliable characterization and modeling of transport properties often require a complete dataset of internal microstructure samples. However, it is often an unbearable cost to acquire sufficient 3D digital microstructures by purely using microscopic imaging systems. Herein this paper presents a machine learning-based technique to hierarchically reconstruct 3D well-connected porous microstructures from one isotropic or several anisotropic low-cost 2D exemplar(s). To compactly characterize the large-scale microstructural features, a Gaussian image pyramid is built for each 2D exemplar. Local morphology patterns are collected from the Gaussian image pyramids, and then they serve as the training data to embed the 2D morphological statistics into feed-forward neural networks at multiple length levels. By using a specially-developed morphology integration scheme, the 3D morphological statistics at different levels can be inferred from the statistics-informed neural networks. Gibbs sampling is adopted to hierarchically reconstruct 3D microstructures by using multi-level 3D morphological statistics, where the large-scale, regional and local morphological patterns are statistically generated and successively added to the same 3D random field. The proposed method is tested on a series of porous media with distinct morphologies, and the statistical equivalence between the reconstructed and the real microstructures is systematically evaluated by comparing morphological descriptors and transport properties. The results demonstrate that the proposed 2D-to-3D microstructure reconstruction method is a universal and efficient approach to generating morphologically and physically realistic samples of porous media.

42 ENGINEERING↗
DOE OSTI2021

Resource selection functions based on hierarchical generalized additive models provide new insights into individual animal variation and species distributions

Habitat selection studies are designed to generate predictions of species distributions or inference regarding general habitat associations and individual variation in habitat use. Such studies frequently involve either individually indexed locations gathered across limited spatial extents and analyzed using resource selection functions (RSFs) or spatially extensive locational data without individual resolution typically analyzed using species distribution models. Both analytical methodologies have certain desirable features, but analyses that combine individual- and population-level inference with flexible non-linear functions may provide improved predictions while accounting for individual variation. Here, we describe how RSFs can be fit using hierarchical generalized additive models (HGAMs) using widely available software, providing a means to explore individual variation in habitat associations and to generate species distribution maps. We used GPS tracking data from golden eagles Aquila chrysaetos from across eastern North America with four environmental predictors to generate monthly distribution models. We considered three model structures that assumed different amounts of individual variation in the functional relationship between predictors and habitat use and used k-fold cross-validation to compare model performance. Models accounting for individual variability in shape and smoothness of functional responses performed best. Eagles exhibited the least amount of individual variation in response to land cover variables during winter months, with most individuals more closely adhering to the population-level trend. During the summer months, eagles exhibited more substantial individual variation in shape and smoothness of the functional relationships, suggesting some need to account for individual variation in eagle habitat use for both inferential and predictive purposes, during this time of year. Because they allow users to blend flexible functions with random effects structures and are well-supported by a variety of software platforms, we believe that HGAMs provide a useful addition to the suite of analyses used for modeling habitat associations or predicting species distributions.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS2015

Software Testbed for Developing and Evaluating Integrated Autonomous Subsystems

To implement fault tolerant autonomy in future space systems, it will be necessary to integrate planning, adaptive control, and state estimation subsystems. However, integrating these subsystems is difficult, time-consuming, and error-prone. This paper describes Intelliface/ADAPT, a software testbed that helps researchers develop and test alternative strategies for integrating planning, execution, and diagnosis subsystems more quickly and easily. The testbed's architecture, graphical data displays, and implementations of the integrated subsystems support easy plug and play of alternate components to support research and development in fault-tolerant control of autonomous vehicles and operations support systems. Intelliface/ADAPT controls NASA's Advanced Diagnostics and Prognostics Testbed (ADAPT), which comprises batteries, electrical loads (fans, pumps, and lights), relays, circuit breakers, invertors, and sensors. During plan execution, an experimentor can inject faults into the ADAPT testbed by tripping circuit breakers, changing fan speed settings, and closing valves to restrict fluid flow. The diagnostic subsystem, based on NASA's Hybrid Diagnosis Engine (HyDE), detects and isolates these faults to determine the new state of the plant, ADAPT. Intelliface/ADAPT then updates its model of the ADAPT system's resources and determines whether the current plan can be executed using the reduced resources. If not, the planning subsystem generates a new plan that reschedules tasks, reconfigures ADAPT, and reassigns the use of ADAPT resources as needed to work around the fault. The resource model, planning domain model, and planning goals are expressed using NASA's Action Notation Modeling Language (ANML). Parts of the ANML model are generated automatically, and other parts are constructed by hand using the Planning Model Integrated Development Environment, a visual Eclipse-based IDE that accelerates ANML model development. Because native ANML planners are currently under development and not yet sufficiently capable, the ANML model is translated into the New Domain Definition Language (NDDL) and sent to NASA's EUROPA planning system for plan generation. The adaptive controller executes the new plan, using augmented, hierarchical finite state machines to select and sequence actions based on the state of the ADAPT system. Real-time sensor data, commands, and plans are displayed in information-dense arrays of timelines and graphs that zoom and scroll in unison. A dynamic schematic display uses color to show the real-time fault state and utilization of the system components and resources. An execution manager coordinates the activities of the other subsystems. The subsystems are integrated using the Internet Communications Engine (ICE). an object-oriented toolkit for building distributed applications.

Testbed↗
NASA NTRS1992

An intelligent decomposition approach for efficient design of non-hierarchic systems

The design process associated with large engineering systems requires an initial decomposition of the complex systems into subsystem modules which are coupled through transference of output data. The implementation of such a decomposition approach assumes the ability exists to determine what subsystems and interactions exist and what order of execution will be imposed during the analysis process. Unfortunately, this is quite often an extremely complex task which may be beyond human ability to efficiently achieve. Further, in optimizing such a coupled system, it is essential to be able to determine which interactions figure prominently enough to significantly affect the accuracy of the optimal solution. The ability to determine 'weak' versus 'strong' coupling strengths would aid the designer in deciding which couplings could be permanently removed from consideration or which could be temporarily suspended so as to achieve computational savings with minimal loss in solution accuracy. An approach that uses normalized sensitivities to quantify coupling strengths is presented. The approach is applied to a coupled system composed of analysis equations for verification purposes.

Bloebaum, Christina L.↗
DOE OSTI2024

Remote Sensing and GIS data at 1km-grid over Chesapeake Bay used in “He et al. 2024, Effects of spatial variability in vegetation phenology, climate, landcover, biodiversity, topography, and soil property on soil respiration across a coastal ecosystem”

The package contains the data layers used in “He et al. 2024, Effects of spatial variability in vegetation phenology, climate, landcover, biodiversity, topography, and soil property on soil respiration across a coastal ecosystem”. The study aims to use multi-source remote sensing and GIS datasets to investigate the spatial heterogeneity and identify spatial zones with similar environmental characteristics and understand the primary driving factors affecting soil respiration within sub-ecosystems of the coastal ecosystem. We employed unsupervised hierarchical clustering analysis to identify spatial regions with distinct environmental characteristics, then determined the main driving factors using Random Forest regression and SHapley Additive exPlanations (SHAP). Spatial data layers include soil respiration, kernel Normalized Difference Vegetation Index (kNDVI) computed from Harmonized Landsat 8 and Sentinel-2 time series, climate variables from the Daymet dataset, land cover, biodiversity, topographical metrics, soil property, and tidal elevation.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

VWC-BERT: Scaling Vulnerability–Weakness–Exploit Mapping on Modern AI Accelerators

Defending cybersystems needs accurate mapping of software and hardware vulnerabilities to generalized descriptions of weaknesses, and weaknesses to exploits. These mappings enable cyber defenders to build plans for effective defense and assessment of potential risks to a cybersystem. With close to 170k vulnerabilities, manual mapping is not a feasible option. However, automated mapping is challenging due to limited training data, computational intractability, and limitations in computational natural language processing. Tools based on breakthroughs in Transformer-based language models have been demonstrated to classify vulnerabilities with high accuracy. We make three key contributions in this paper: (1) We present a new framework, \VWCBERT, that augments the Transformer-based hierarchical multi-class classification framework of Das et al. (\textsc{V2W-BERT}) with the ability to map weaknesses to exploits. (2) We implement \VWCBERT~ on modern AI accelerator platforms using two data parallel techniques for the pre-training phase and demonstrate nearly linear speedups across NVIDIA and Graphcore accelerator platforms. We observe nearly linear speedups for up to 16 V100 and 8 A100 GPUs, and about 3.4$\times$ speedup for A100 relative to V100 GPUs. We also observe excellent speedups on Graphcore, with $5.7\times$ speedup on 128 IPUs relative to 16 IPUs. Enabled by scaling, we also demonstrate higher accuracy using a larger language model, RoBERTa-Large. We show up to 87\% accuracy for strict and up to 98\% accuracy for relaxed classification. (3) We develop a novel parallel link manager for the link prediction phase and demonstrate up to 21$\times$ speedup with 16 V100 GPUs relative to one V100 GPU, and thus reducing the runtime from 2.5 hours to 10 minutes. We believe that generalizability and scalability of \VWCBERT~ will benefit both the theoretical development and practical deployment of novel cyberdefense solutions and vulnerability classification.

Das, Siddhartha Shankar↗