Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “graphs”

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 613 records · Page 34

HydraGNN_Predictive_GFM_2024 - Ensemble of predictive graph foundation models for ground state atomistic materials modeling

We provide the ensemble of fifteen pre-trained graph foundation models (GFMs) for atomistic materials modeling applications. Each one of the fifteen GFMs has been trained on five open-source datasets that (once aggregated) amount to over 154 million atomistic structures, which cover over two-thirds of the natural elements of the periodic table and that comprises a broad set of organic and inorganic compounds. This vast set of atomistic structures comprises ground state configurations that are dynamically stable (i.e., equilibrated structures with atomic forces approximately close to zero values) as well as dynamically unstable structures (i.e., non-equilibrium structures with non-negligible non-zero values of atomic forces). The ensemble of datasets aggregated does NOT include excited states. The datasets have been curated to remove atomistic structures with spectral norm of the force tensor above 100 eV/angstrom. Moreover, a linear term of the energy was computed for each dataset using a linear regression model that uses the chemical concentration of each natural element as regressor. The linear term predicted by the linear regression model has been subtracted from each original energy value to perform a re-alignment of the energy values across different electronic structures approximation theories performed to generate the diverse multi-source, multi-fidelity datasets. The folder "ADIOS_files" contains the set of pre-processed datasets in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used for the development and training of GFMs in this work. The "ADIOS_files" directory contains 6 sub-directories named as follows: - ANI1x-v3.bp - MPTrj-v3.bp - OC2020-20M-v3.bp - OC2020-v3.bp - OC2022-v3.bp - qm7x-v3.bp Each sub-directory contains the pre-processed datasets converted in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used to the development, training, and performance testing of the ensemble go predictive graph foundation models. Each GFM was developed using HydraGNN (https://github.com/ORNL/HydraGNN) as underlying graph neural network (GNN) architecture. The multi-task learning (MTL) capability of HydraGNN was used to simultaneously train the GFMs on labeled values for direct predictions of energy (a total system property of an atomistic structure that measures the chemical stability) and atomic forces (an atomic level property of an atomistic structure that measures the dynamical stability). The hyper parameters of the GFM have been tuned using scalable hyperparameter optimization (HPO) algorithms implemented in the software DeepHyper (https://github.com/deephyper/deephyper). The pre-training of each HPO trial was performed using distributed data parallelism (DDP) to scale the training across 128 compute nodes of the exascale OLCF supercomputer Frontier. Each HPO trial was trained only for 10 epochs and an early stopping was performed to avoid wasting significant computational resources on GNN architectures that were clearly underperforming. For each HPO trial, the 'omnistat' tool developed by (AMD Research - Advanced Micro Device) was used to measure the total energy consumption in kWh. The ensemble of GFMs was obtained by selecting the fifteen best performing HPO trials. Four models have been selected for their clear advantage in accuracy, and these are the GFMs with IDs 229, 156, 147, 260. Additional eleven models have been selected based on judicious balance between accuracy and energy consumption needed for training, and these are the GFMs with IDs 165, 78, 137, 1, 175, 171, 181, 67, 179, 167, 351. Each selected GFM of the ensemble was continued to cumulate a total of at most 30 epochs. In some cases, the total number of epochs actually performed was les than 30 due to two combined factors: (1) the size of the GFM (i.e., the number of model parameters to train) and (2) the total wall-clock time for which the computational resources could be allocated on OLCF-Frontier. The "Ensemble_of_models" directory contains 15 sub-directories named as follows: - gfm_0.229 - gfm_0.156 - gfm_0.147 - gfm_0.260 - gfm_0.165 - gfm_0.78 - gfm_0.137 - gfm_0.1 - gfm_0.175 - gfm_0.171 - gfm_0.181 - gfm_0.67 - gfm_0.179 - gfm_0.167 - gfm_0.351 Each one of these sub-directories refers to one of the fifteen HPO trials that have been selected to continue the pre-training with at most 30 epochs. With each sub-directory associated with a specific HPO trial, the following files can be found: - config.json: file for argument parsing to develop and train an HydraGNN architecture - gfm_0.ID_epoch_N.pk: file with model parameters for HPO ID trial after N epochs of training The ensemble of fifteen GFM architectures was used for (1) ensemble averaging to stabilize the predictions of energy and atomic forces after pre-training for post-processing analysis and (2) ensemble uncertainty quantification (UQ). The code used to develop, pre-train, and load the pre-trained models for post-processing analysis is available on the ORNL-GitHub at the following link: https://github.com/ORNL/HydraGNN/tree/Predictive_GFM_2024

36 MATERIALS SCIENCE↗

The Graph Laplacian and the Dynamics of Complex Networks [Slides]

In this talk, we explore the structure of networks from a spectral graph-theoretic perspective by analyzing the properties of the Laplacian matrix associated with the graph induced by a network. We will see how the eigenvalues of the graph Laplacian relate to the underlying network structure and dynamics and provides insight into a phenomenon frequently observed in real world networks - the emergence of collective behavior from purely local interactions seen in the coordinated motion of animals and phase transitions in biological networks, to name a few.

97 MATHEMATICS AND COMPUTING↗

Ridge-Valley Graphs: Combinatorial Ridge Detection Using Jacobi Sets

Ridges are one of the key feature of interest in areas such as computer vision and image processing. Even though a significant amount of research has been directed to defining and extracting ridges some fundamental challenges remain. For example, the most popular ridge definition (height ridge) is not invariant under monotonic transformations and its global structure is typically ignored during numerical computations. Furthermore, many existing algorithm are based on numerical heuristics and are rarely guaranteed to produce consistent results. In this paper a small change to the height ridge definition gives a definition that is consistent with the desired invariants. Nevertheless, we show that this definition results in similar structures compared to the most common traditional approach and that both formulations are equivalent for quadratic functions. Furthermore, this definition can be cast in the form of a degenerate Jacobi set, which allows insights into the global structure of ridges. In particular, we introduce the Ridge-Valley graph as the complete description of all ridges in an image. Analyzing this graph reveals similar invariants to those found for traditional ridge formulations, such as the fact that ridges do not merge. Finally, using the connection to Jacobi sets we describe a new combinatorial algorithm to extract the Ridge-Valley graph from sampled images that is guaranteed to produce a valid structure.

97 MATHEMATICS AND COMPUTING↗

Multiplicative Attribute Graph Model of Real-World Networks

Large scale real-world network data, such as social networks, Internet andWeb graphs, is ubiquitous in a variety of scientific domains. The study of such social and information networks commonly finds patterns and explain their emergence through tractable models. In most networks, especially in social networks, nodes also have a rich set of attributes (e.g., age, gender) associatedwith them. However, most of the existing network models focus only on modeling the network structure while ignoring the features of nodes in the network. Here we present a class of network models that we refer to as the Multiplicative Attribute Graphs (MAG), which naturally captures the interactions between the network structure and node attributes. We consider a model where each node has a vector of categorical features associated with it. The probability of an edge between a pair of nodes then depends on the product of individual attributeattribute similarities. The model yields itself to mathematical analysis as well as fit to real data. We derive thresholds for the connectivity, the emergence of the giant connected component, and show that the model gives rise to graphs with a constant diameter. Moreover, we analyze the degree distribution to show that the model can produce networks with either lognormal or power-law degree distribution depending on certain conditions.

97 MATHEMATICS AND COMPUTING↗

Encoding and Analyzing Aerial Imagery Using Geospatial Semantic Graphs

While collection capabilities have yielded an ever-increasing volume of aerial imagery, analytic techniques for identifying patterns in and extracting relevant information from this data have seriously lagged. The vast majority of imagery is never examined, due to a combination of the limited bandwidth of human analysts and limitations of existing analysis tools. In this report, we describe an alternative, novel approach to both encoding and analyzing aerial imagery, using the concept of a geospatial semantic graph. The advantages of our approach are twofold. First, intuitive templates can be easily specified in terms of the domain language in which an analyst converses. These templates can be used to automatically and efficiently search large graph databases, for specific patterns of interest. Second, unsupervised machine learning techniques can be applied to automatically identify patterns in the graph databases, exposing recurring motifs in imagery. We illustrate our approach using real-world data for Anne Arundel County, Maryland, and compare the performance of our approach to that of an expert human analyst.

97 MATHEMATICS AND COMPUTING↗

Synthetic Graph Generation for Data-Intensive HPC Benchmarking: Scalability, Analysis and Real-World Application

The benchmarking effort within the Extreme Scale Systems Center at Oak Ridge National Laboratory seeks to provide High Performance Computing benchmarks and test suites of interest to the DoD sponsor. The work described in this report is a part of the effort focusing on graph generation. A previously developed benchmark, SystemBurn, allows the emulation of a broad spectrum of application behavior profiles within a single framework. To complement this effort, similar capabilities are desired for graph-centric problems. This report described the in-depth analysis of the generated synthetic graphs' properties at a variety of scales using different generator implementations and examines their applicability to replicating real world datasets.

97 MATHEMATICS AND COMPUTING↗

Geospatial-Temporal Semantic Graphs for Automated Wide-Area Search

We address the problem of wide-area search of overhead imagery. Given a time sequence of overhead images, we construct a geospatial-temporal semantic graph, which expresses the complex continuous information in the overhead images in a discrete searchable form, including explicit modeling of changes seen from one image to the next. We can then express desired search goals as a template graph, and search for matches using simple and efficient graph search algorithms. This produces a set of potential matches which provide cues for where to examine the imagery in detail, applying human expertise to determine which matches are correct. We include a match quality metric that scores the matches according to how well they match the stated search goal. This enables matches to be presented in sorted order with the best matches first, similar to the results returned by a web search engine. We present an evaluation of the method applied to several examples and data sets, and show that it can be used successfully for some problems. We also remark on several limitations of the method and note additional work needed to improve its scope and robustness. Approved for public release; further dissemination unlimited.

97 MATHEMATICS AND COMPUTING↗

Asynchronous Graphs (Final Report)

The most performant massive-scale distributed graph analytic engines were essentially unusable for anything but the analytic they were tuned to solve. The project team had previously developed HavoqGT and used this project to demonstrate it as a flexible graph framework, aiming to extend algorithmic capabilities, improve underlying data structures, and develop novel communication paradigms to enable new data analysis missions. Highlights of the technology developed include (a) a new distributed communication paradigm dubbed You’ve Got Mail (YGM), (b) a performant exact and fuzzy HPC graph pattern matching approach, and (c) a distributed biased random walk simulation capability. The project was successful on several measurable fronts, including advancing computer science (reflected by publishing 14 peer-reviewed research papers yielding 4 publication awards and over 16 conference presentations), contributing tangibly to LLNL recruiting rare talent, and obtaining follow-on funding. These technologies were transferred to several FY19-20 programmatic projects and will continue to be leveraged by new research projects for several years.

97 MATHEMATICS AND COMPUTING↗

Neuromorphic Graph Algorithms

Graph algorithms enable myriad large-scale applications including cybersecurity, social network analysis, resource allocation, and routing. The scalability of current graph algorithm implementations on conventional computing architectures are hampered by the demise of Moore’s law. We present a theoretical framework for designing and assessing the performance of graph algorithms executing in networks of spiking artificial neurons. Although spiking neural networks (SNNs) are capable of general-purpose computation, few algorithmic results with rigorous asymptotic performance analysis are known. SNNs are exceptionally well-motivated practically, as neuromorphic computing systems with 100 million spiking neurons are available, and systems with a billion neurons are anticipated in the next few years. Beyond massive parallelism and scalability, neuromorphic computing systems offer energy consumption orders of magnitude lower than conventional high-performance computing systems. We employ our framework to design and analyze new spiking algorithms for shortest path and dynamic programming problems. Our neuromorphic algorithms are message-passing algorithms relying critically on data movement for computation. For fair and rigorous comparison with conventional algorithms and architectures, which is challenging but paramount, we develop new models of data-movement in conventional computing architectures. This allows us to prove polynomial-factor advantages, even when we assume a SNN consisting of a simple grid-like network of neurons. To the best of our knowledge, this is one of the first examples of a rigorous asymptotic computational advantage for neuromorphic computing.

97 MATHEMATICS AND COMPUTING↗

Randomized Cholesky Preconditioning for Graph Partitioning Applications

Graph partitioning has emerged as an area of interest due to its use in various applications in computational research. One way to partition a graph is to solve for the eigenvectors of the corresponding graph Laplacian matrix. This project focuses on the eigensolver LOBPCG and the evaluation of a new preconditioner: Randomized Cholesky Factorization (rchol). This proconditioner was tested for its speed and accuracy against other well-known preconditioners for the method. After experiments were run on several known test matrices, rchol appears to be a better preconditioner for structured matrices. This research was sponsored by National Nuclear Security Administration Minority Serving Institutions Internship Program (NNSA-MSIIP) and completed at host facility Sandia National Laboratories. As such, after discussion of the research project itself, this report contains a brief reflection on experience gained as a result of participating in the NNSA-MSIIP.

97 MATHEMATICS AND COMPUTING↗

Graph Embeddings for CEBAF Operations: Progress and Future Plans

We describe research towards leveraging deep learning on graph representations of the injector beamline at the Continuous Electron Beam Accelerator Facility (CEBAF) in order to create a tool for improving the efficiency of beam tuning tasks. Specifically, we use graphs to represent the injector beamline at any arbitrary date and time and invoke a graph neural network to extract a low-dimensional, informative representation that can be visualized in two-dimensions. By analyzing years of operational data from the CEBAF archiver, good and bad regions of parameter space can be identified. The goal is to exercise this framework as a real-time tool to aid beam tuning, which represents the dominant source of machine downtime.

Tennant, C.↗

0BGRaman: Graph Network based Simulator for Forecasting Molecular Polarizability

This report presents the work performed under the GRaman project, sponsored by the PCSD LDRD Seed program. The project aimed at accelerating ab initio molecular dynamics simulation using Graph Networks. The Graph Network framework is a ML framework that has been successfully employed to simulate the dynamics of several physical systems: including water splashing in a container and flags moving with the wind. In this effort, we performed a data collection campaign for 3 different molecules of interest. We have built tools for preprocessing the trajectories obtained by simulating Raman Spectroscopy with NWChem and translating them into a suitable format for training. We have developed a training algorithm to train the Graph Network based simulators based on our data and developed a simulator that produces trajectories in the same NWChem format. While the tool has improved with each iteration of development and subsequent experiments, the current state of the tool does not allow to directly incorporate the technology within the NWChem framework because the trajectories produced by the tool are not yet accurate enough. However, the technology has proved to have good potential and it is certainly worth further research and development.

97 MATHEMATICS AND COMPUTING↗

Replace Human Intelligence with Fast and Smart Geometric Reasoning and Graph Neural Network to Accelerate Next Gen ModSim Workflows

We present an agent-guided approach to CAD geometry decomposition that automates hex/hybrid meshing with graph neural networks (GNNs) to accelerate next-generation ModSim workflows. Our end-to-end pipeline (i) reduces 3D boundary-representation (B-Rep) models to a 2D chordal axis skeleton (CAT) and then to a 1D bipartite graph of surface and curve nodes, (ii) assigns per node labels as Cubit® WebCut actions, (iii) trains a multi-action GNN under supervised learning, and (iv) predicts five surface-node and three curve-node actions on out-of-distribution test geometries. Each graph node carries geometric, topological, and meshing attributes drawn from the B-Rep “skin” and CAT “skeleton,” with two-way mappings across 3D↔2D↔1D representations to maintain traceability back to 3D CAD. The supervised learning model exhibits stable convergence of the binary cross-entropy loss and achieves 98.7% accuracy on unseen lattice models. To operationalize decision-making, we rank predicted commands by geometric significance and prototyped the agent-guided workflow through the Cubit® Meshing PowerTool GUI. As a stretch goal, we explore reinforcement learning (RL) to reduce or remove label requirements and to learn policies for action sequences that maximize total reward (e.g., size of hex-meshable regions and resulting hex mesh quality). When all-hex meshing is not feasible, the agent assists in producing hybrid meshes—prioritizing hex in critical regions and transitioning to tetrahedral elements (tets) elsewhere—maintaining fidelity while ensuring robustness. The overarching objective is to replace manual, heuristics-based decomposition with data-driven, reproducible automation, cutting meshing turnaround time by orders of magnitude. We anticipate direct impact on simulation workflows through intelligent, scalable decomposition of complex CAD models into hex-meshable subdomains.

97 MATHEMATICS AND COMPUTING↗

AI-Ready Semantic Infrastructure for CEBAF: From CED to PALS Knowledge Graphs

JLab and PNNL are jointly developing an AI-ready data ecosystem that exposes the Continuous Electron Beam Acceleration Facility’s (CEBAF’s) operational configuration, lattice description, and control-system channels to agentic optimization frameworks through a standards-based semantic layer. The effort integrates the existing facility-specific CEBAF Element Database (CED) with extensions of the emerging facility-agnostic Particle Accelerator Lattice Standard (PALS) to produce a knowledge graph (KG) containing coherent, machine-interpretable views of devices, signals, and regions. With this KG, CEBAF’s setpoints, readbacks, and device hierarchies become queryable using a uniform declarative graph query language (e.g., Neo4j Cypher), providing intents and inspectable semantics suitable for agentic control. The resulting graph-backed interfaces will allow autonomous agents to retrieve authoritative machine configurations, reason over device- and signal-level relationships, and execute tuning and diagnostic workflows without bespoke CEBAF-specific logic, thereby delivering a scalable pathway from operational data to trustworthy agentic accelerator tuning frameworks.

Zhang, He [Thomas Jefferson National Accelerator F↗

Improving the Performance of NEML2 with Modern Graph Compilation Backends

NEML2 vectorizes constitutive-model evaluation for large-scale multiphysics simulation, using PyTorch as its tensor backend so that a batch of material-point updates runs on CPU or GPU through a single implementation. In the two prior reports in this series it was a C++-native library, deployed through TorchScript tracing and just-in-time (JIT) compilation; it has since been rewritten from the ground up into a Python-native library deployed through Ahead-of-Time Inductor (AOTInductor), a modern PyTorch graph-compilation backend. The rewrite is driven by a persistent tension, not a language preference: NEML2 composes constitutive models at runtime from a registry of small, independently-authored pieces, and that flexibility is difficult to reconcile with the compile-time knowledge an efficient GPU kernel needs. This report documents the rewrite and the investment that accompanied it: the AOTInductor export pipeline that turns a Python-authored model into a portable, Python-free compiled artifact loadable from pure C++; the eager and compiled runtimes and the new implicit solver layer built on them; a head-to-head benchmark of legacy JIT against AOTInductor; the physics-model catalog and its worked examples; the developer tooling; and the corresponding overhaul of MOOSE’s NEML2 integration that lets MOOSE consume it. A central objective is to examine whether modern PyTorch graph-compilation backends are effective for MOOSE GPU integration. The benchmark answers directly: AOTInductor outperforms legacy JIT on every GPU scenario measured, by 1.0–4.5×. Modern graph-compilation backends are effective for MOOSE GPU integration, and AOTInductor specifically – not compilation in the abstract – is why.

Hu, Gary (Tianchen) [Argonne National Laboratory (↗

Algebraic connectivity and graph robustness.

Recent papers have used Fiedler's definition of algebraic connectivity to show that network robustness, as measured by node-connectivity and edge-connectivity, can be increased by increasing the algebraic connectivity of the network. By the definition of algebraic connectivity, the second smallest eigenvalue of the graph Laplacian is a lower bound on the node-connectivity. In this paper we show that for circular random lattice graphs and mesh graphs algebraic connectivity is a conservative lower bound, and that increases in algebraic connectivity actually correspond to a decrease in node-connectivity. This means that the networks are actually less robust with respect to node-connectivity as the algebraic connectivity increases. However, an increase in algebraic connectivity seems to correlate well with a decrease in the characteristic path length of these networks - which would result in quicker communication through the network. Applications of these results are then discussed for perimeter security.

97 MATHEMATICS AND COMPUTING↗

A Unification Framework for Euclidean and Hyperbolic Graph Neural Networks

Hyperbolic neural networks have recently gained significant attention due to their promising results on several graph problems including node classification and link prediction. The primary reason for this success is the effectiveness of hyperbolic space in capturing the inherent hierarchy of graph datasets. However, they are limited in terms of generalization, scalability, and have inferior performance when applied to non-hierarchical datasets. In this paper, we take a completely different perspective for modeling hyperbolic networks and answer the following question: is an Euclidean model able to approximate a function or behavior in the hyperbolic space? Extending the universal approximation theory developed for Euclidean models, We draw an analogy from the hyperbolic components to the Euclidean counterparts and conclude that, in order to capture hierarchical features, it is possible to generalize hyperbolic models to be a special case of Euclidean models with the proposed Pseudo-Poincaré technique. We applied our non-linear hyperbolic normalization to the current state-of-the-art homogeneous and multi-relational graph networks and demonstrate significant improvements in performance compared to both Euclidean and hyperbolic counterparts. The primary impact of this work lies in its ability to capture hierarchical features in the Euclidean space, and thus, can replace hyperbolic networks without any loss in performance metrics while simultaneously leveraging the power of Euclidean networks such as interpretability and efficient execution of various model components.

Khatir, Mehrdad↗

Graph Neural Networks for Charged Particle Tracking on FPGAs

The determination of charged particle trajectories in collisions at the CERN Large Hadron Collider (LHC) is an important but challenging problem, especially in the high interaction density conditions expected during the future high-luminosity phase of the LHC (HL-LHC). Graph neural networks (GNNs) are a type of geometric deep learning algorithm that has successfully been applied to this task by embedding tracker data as a graph—nodes represent hits, while edges represent possible track segments—and classifying the edges as true or fake track segments. However, their study in hardware- or software-based trigger applications has been limited due to their large computational cost. In this paper, we introduce an automated translation workflow, integrated into a broader tool called hls4ml , for converting GNNs into firmware for field-programmable gate arrays (FPGAs). We use this translation tool to implement GNNs for charged particle tracking, trained using the TrackML challenge dataset, on FPGAs with designs targeting different graph sizes, task complexites, and latency/throughput requirements. This work could enable the inclusion of charged particle tracking GNNs at the trigger level for HL-LHC experiments.

Elabd, Abdelrahman↗