Search NASA⌕ Search

SEARCH · Search NASA

Results for “large graph”

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 235 records · Page 13
All sourcesNASA NTRSDOE OSTI
NASA NTRS1990

A VLSI decomposition of the deBruijn graph

A new Viterbi decoder for convolutional codes with constraint lengths up to 15, called the Big Viterbi Decoder, is under development for the Deep Space Network. It will be demonstrated by decoding data from the Galileo spacecraft, which has a rate 1/4, constraint-length 15 convolutional encoder on board. Here, the mathematical theory underlying the design of the very-large-scale-integrated (VLSI) chips that are being used to build this decoder is explained. The deBruijn graph B sub n describes the topology of a fully parallel, rate 1/v, constraint length n+2 Viterbi decoder, and it is shown that B sub n can be built by appropriately wiring together (i.e., connecting together with extra edges) many isomorphic copies of a fixed graph called a B sub n building block. The efficiency of such a building block is defined as the fraction of the edges in B sub n that are present in the copies of the building block. It is shown, among other things, that for any alpha less than 1, there exists a graph G which is a B sub n building block of efficiency greater than alpha for all sufficiently large n. These results are illustrated by describing a special hierarchical family of deBruijn building blocks, which has led to the design of the gate-array chips being used in the Big Viterbi Decoder.

Collins, O.↗
DOE OSTI2020

Quantum simulation of hyperbolic space with circuit quantum electrodynamics: From graphs to geometry

We show how quantum many-body systems on hyperbolic lattices with nearest-neighbor hopping and local interactions can be mapped onto quantum field theories in continuous negatively curved space. Here, the underlying lattices have recently been realized experimentally with superconducting resonators and therefore allow for a table-top quantum simulation of quantum physics in curved background. Our mapping provides a computational tool to determine observables of the discrete system even for large lattices, where exact diagonalization fails. As an application and proof of principle we quantitatively reproduce the ground state energy, spectral gap, and correlation functions of the noninteracting lattice system by means of analytic formulas on the Poincaré disk, and show how conformal symmetry emerges for large lattices. This sets the stage for studying interactions and disorder on hyperbolic graphs in the future. Importantly, our analysis reveals that even relatively small discrete hyperbolic lattices emulate the continuous geometry of negatively curved space, and thus can be used to experimentally resolve fundamental open problems at the interface of interacting many-body systems, quantum field theory in curved space, and quantum gravity.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2024

Graph-Based Representations and Applications to Process Simulation

Rapid and robust convergence of a process flowsheet is critical to enable large-scale simulations that address core scientific questions related to process design, optimization, and sustainability. However, due to the highly coupled and nonlinear nature of chemical processes, efficiently solving a flowsheet remains a challenge. In this work, we show that graph representations of the underlying physical phenomena in unit operations may help identify potential avenues to systematically reformulate the network of equations and enable more robust topology-based convergence of flowsheets. To this end, we developed graph abstractions of the governing equations of vapor-liquid and liquid-liquid equilibrium separation equipment. These graph abstractions consist of a mesh of interconnected variable nodes and equation nodes that are systematically generated through PhenomeNode, a new open-source library in Python developed in this study. We show that partitioning the graph into separate mass, energy, and equilibrium subgraphs can help decouple nonlinearities and guide decomposition algorithms. By employing the graph abstraction on an industrial separation process for separating glacial acetic acid from water, we implemented a new block decomposition scheme in BioSTEAM and demonstrated that this can accelerate convergence over a traditional sequential modular approach.

Distillation↗
DOE OSTI2026

Multitask graph neural networks for elastoplastic response prediction in dual-phase polycrystals

Microstructure-sensitive prediction of elastoplastic response remains a recurring bottleneck in multiscale damage and fatigue modeling, where large ensembles of statistically distinct polycrystals are required to quantify variability and extreme-value behavior. In this work, we develop a multitask graph neural network (GNN) surrogate that maps dual-phase ferrite–martensite polycrystal microstructures to Statistical Volume Element (SVE)-level elastoplastic Quantities of Interest (QoIs). Each SVE is represented as a grain-adjacency graph, with node features encoding phase, geometry, and crystallographic orientation, and edge features encoding relative misorientation. A message-passing graph convolution generates node embeddings, which are pooled into a graph representation and passed to a multitask regression head that jointly predicts 10 scalar QoIs and vector-valued stress–strain responses in orthogonal loading directions across multiple martensite volume fractions and SVE sizes. Results show high accuracy for scalar QoIs and strong agreement for full stress–strain trajectories, with population envelopes reproducing both median behavior and finite-SVE variability across compositions and partition scales. A unified model trained on pooled volume-fraction data preserves most within-regime accuracy relative to regime-specific models while also capturing the broader cross-regime variation reflected in the pooled test set. Distributional comparisons further demonstrate that the surrogate preserves heterogeneity under SVE partitioning, enabling statistically consistent block-wise random-field construction for mesoscale analyses. Overall, the proposed grain-graph surrogate provides a practical pathway to accelerate ensemble-based studies of SVE-level constitutive variability in dual-phase polycrystals.

Crystal plasticity↗
DOE OSTI2024

Geometry-complete diffusion for 3D molecule generation and optimization

Abstract Generative deep learning methods have recently been proposed for generating 3D molecules using equivariant graph neural networks (GNNs) within a denoising diffusion framework. However, such methods are unable to learn important geometric properties of 3D molecules, as they adopt molecule-agnostic and non-geometric GNNs as their 3D graph denoising networks, which notably hinders their ability to generate valid large 3D molecules. In this work, we address these gaps by introducing the Geometry-Complete Diffusion Model (GCDM) for 3D molecule generation, which outperforms existing 3D molecular diffusion models by significant margins across conditional and unconditional settings for the QM9 dataset and the larger GEOM-Drugs dataset, respectively. Importantly, we demonstrate that GCDM’s generative denoising process enables the model to generate a significant proportion of valid and energetically-stable large molecules at the scale of GEOM-Drugs, whereas previous methods fail to do so with the features they learn. Additionally, we show that extensions of GCDM can not only effectively design 3D molecules for specific protein pockets but can be repurposed to consistently optimize the geometry and chemical composition of existing 3D molecules for molecular stability and property specificity, demonstrating new versatility of molecular diffusion models. Code and data are freely available on GitHub .

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Sparse Convolutional Neural Networks for particle classification in ProtoDUNE-SP events

Deep Learning (DL) methods and Computer Vision are becoming important tools for event reconstruction in particle physics detectors. In this work, we report on the use of submanifold sparse convolutional neural networks (SparseNets) for the classification of track and shower hits from a DUNE prototype liquid-argon detector at CERN (ProtoDUNE-SP). By taking advantage of the three-dimensional nature of the problem we use a set of nine input features to classify sparse and locally dense hits associated to track or shower particles. The SparseNet has been trained on a test sample and shows promising results: efficiencies and purities greater than 90%. This has also been achieved with a considerable speedup and substantially less resource utilization with respect to other DL networks such as graph neural networks. This method offers great scalability advantages for future large neutrino detectors such as the planned DUNE experiment.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2026

Evaluating the Limits of QAOA Parameter Transfer at High-Rounds on Sparse Ising Models With Geometrically Local Cubic Terms

The emergent practical applicability of the Quantum Approximate Optimization Algorithm (QAOA) for approximate combinatorial optimization is a subject of considerable interest. One of the primary limitations of QAOA is the task of finding a set of good parameters, which is usually done using a variational optimization loop. Parameter transfer, or parameter concentration, is a phenomenon where QAOA angles trained on problem instances that are self-similar tend to perform well for other problem instances from that similar class. This suggests a potentially highly efficient and scalable non-variational learning method for QAOA angle finding. In this work, we systematically study QAOA parameter transferability from small problem sizes (16 and 27 decision variables) onto large problem instances (up to 156 qubits) for heavy-hex graph Ising models with geometrically local higher order terms using the Julia based QAOA simulation tool \texttt{JuliQAOA} to perform classical angle finding for up to $49$ QAOA layers ($p$). Parameter transfer of the fixed angles is validated using a combination of full statevector, Projected Entangled Pair States (PEPS), Matrix Product State (MPS), and LOWESA numerical simulations. We find that the QAOA parameter transfer from single instances applied to other (unseen) problem instances does not in general provide monotonically improving performance as a function of $p$ - there are many cases where the performance temporarily decreases as a function of $p$ - but despite this the transferred angles have a general trend of improved expectation value as the QAOA depth increases, in many cases converging close to the true ground-state energy of the $100+$ qubit instances. We also sample the hardware-compatible Ising models using the ensemble of transfer-learned QAOA parameters on several superconducting qubit IBM Quantum processors with 127, 133, and 156 qubits. We find continuous solution quality improvement of the hardware-compatible QAOA circuits run on the IBM NISQ processors up to $p=5$ on \texttt{ibm\_fez}, up to $p=9$ on \texttt{ibm\_torino}, and up to $p=10$ on \texttt{ibm\_pittsburgh}.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2024

A Deep Multimodal Representation Learning Framework for Accurate Molecular Properties Prediction

Drug discovery is a complex and challenging process, requiring the optimization of candidate compounds to identify those with the potential to become safe and effective drugs. Predicting molecular properties is an indispensable step in the drug discovery pipeline. Traditionally, this process is costly and time-intensive, involving multiple rounds of experiments and clinical trials, rendering it impractical for every candidate compound. Deep learning techniques have emerged as a promising approach to drug discovery to reduce the cost and time required to identify novel drugs. However, prevalent research in deep learning models focused on predicting molecular properties has primarily fixated on single-modal models, which utilize a single modality of data, neglecting the potential benefits of combining different data modalities. To overcome this limitation, we introduce MRL-Mol: a deep \textbf{M}ultimodal \textbf{R}epresentation \textbf{L}earning framework for accurate \textbf{Mol}ecular properties prediction. MRL-Mol harnesses three data modalities: sequence, graph, and image, augmenting the depth of comprehension. Leveraging a large-scale unlabeled dataset~($\sim$1M unique molecules), we pretrain MRL-Mol to extract inter- and intra-modal information. Our study demonstrates the superior performance of MRL-Mol in predicting molecular properties across six benchmark datasets, including both classification and regression tasks. Notably, MRL-Mol outperforms other state-of-the-art molecular properties prediction models. These findings suggest that by combining information from multiple data modalities, MRL-Mol can comprehend molecules better than single-modal deep learning models and identify molecular properties with better accuracy.

Yang, Yuxin↗
NASA NTRS1987

Auroral zone field-aligned currents observed in the magnetotail and at intermediate altitudes - An ISEE perspective

The relationship between low-altitude auroral-zone field-aligned currents and the geomagnetic tail is investigated on the basis of a statistical analysis of ISEE and DE observations. The data are compiled in diagrams and graphs and characterized in detail. It is shown that the large-scale current systems of low-altitude regions 1 and 2 are only reduplicated in 27 percent of the observations at 2-7 earth radii (those associated with the onset of a substorm expansion), while the flow at the plasma-sheet boundary opposes the region 1 polarity (the highest-latitude near-midnight currents being directed earthward).

Elphic, R. C.↗
NASA NTRS2007

Exploring and Visualizing A-Train Instrument Data

The succession of US and international satellites that follow each other in close succession, known as the A-Train, affords an opportunity to atmospheric researchers that no single platform could provide: Increasing the number of observations at any given geographic location.. . a more complete "virtual science platform". However, vertically and horizontally, co-registering and regridding datasets from independently developed missions, Aqua, Calipso, Cloudsat, Parasol, and Aura, so that they can be inter-compared can be daunting to some, and may be repeated by many. Scientists will individually spend much of their time and resources acquiring A-Train datasets of interest residing at various locations, developing algorithms to match up and graph datasets along the A-Train track, and search through large amounts of data for areas and/or phenomena of interest. The aggregate amount of effort that can be expended on repeating pre-science tasks could climb into the tens of millions of dollars. The goal of the A-Train Data Depot (ATDD) is to enable free movement of remotely located A-Train data so that they are combined to create a consolidated vertical view of the Earth's Atmosphere along the A-Train tracks. The innovative approach of analyzing and visualizing atmospheric profiles along the platforms track (i.e., time) is accomplished by through the ATDDs Giovanni data analysis and visualization tool. Giovanni brings together data from Aqua (MODIS, AIRS, AMSR-E), Cloudsat (cloud profiling radar) and Calipso (CALIOP, IIR), as well as the Aura (OMI, MLS, HIRDLS, TES) to create a consolidated vertical view of the Earth's Atmosphere along the A-Train tracks. This easy to learn and use exploration tool will allow users to create vertical profiles of any desired A-Train dataset, for any given time of choice. This presentation shows the power of Giovanni by describing and illustrating how this tool facilitates and aids A-Train science and research. A web based display system Giovanni provides users with the capability of creating co-located profile images of temperature and humidity data from the MODIS, MLS and AIRS instruments for a user specified time and spatial area. In addition, Cloud and Aerosol profiles may also be displayed for the Cloudsat and Caliop instruments. The ability to modify horizontal and vertical axis range, data range and dynamic color range is also provided. Two dimensional strip plots of MODIS, AIRS, OM1 and POLDER parameters, co-located along the Cloudsat reference track, can also be plotted along with the Cloudsat cloud profiling data. Center swath pixels for the same parameters can also be shown as line plots overlaying the Cloudsat or Calipso profile images. Images and subsetted data produced in each analysis run may be downloaded. Users truly can explore and discover data specific to their needs prior to ever transferring data to their analysis tools.

Kempler, S.↗
DOE OSTI2025

Graph decomposition techniques for solving combinatorial optimization problems with variational quantum algorithms

The quantum approximate optimization algorithm (QAOA) has the potential to approximately solve complex combinatorial optimization problems in polynomial time. However, current noisy quantum devices cannot solve large problems due to hardware constraints. In this work, we develop an algorithm that decomposes the QAOA input problem graph into a smaller problem and solves MaxCut using QAOA on the reduced graph. The algorithm requires a subroutine that can be classical or quantum—in this work, we implement the algorithm twice on each graph. One implementation uses the classical solver Gurobi in the subroutine and the other uses QAOA. We solve these reduced problems with QAOA. On average, the reduced problems require only approximately 1/10 of the number of vertices than the original MaxCut instances. Furthermore, the average approximation ratio of the original MaxCut problems is 0.75, while the approximation ratios of the decomposed graphs are on average of 0.96 for both Gurobi and QAOA. With this decomposition, we are able to measure optimal solutions for ten 100-vertex graphs by running single-layer QAOA circuits on the Quantinuum trapped-ion quantum computer H1-1, sampling each circuit only 500 times. This approach is best suited for sparse, particularly k-regular graphs, as k-regular graphs on n vertices can be decomposed into a graph with at most $\frac{nk}{k+1}$ vertices in polynomial time. Further reductions can be obtained with a potential trade-off in computational time. In conclusion, while this paper applies the decomposition method to the MaxCut problem, it can be applied to more general classes of combinatorial optimization problems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Harnessing distributed GPU computing for generalizable graph convolutional networks in power grid reliability assessments

Although machine learning (ML) has emerged as a powerful tool for rapidly assessing grid contingencies, prior studies have largely considered a static grid topology in their analyses. This limits their application, since they need to be re-trained for every new topology. Here, this paper explores the development of generalizable graph convolutional network (GCN) models by pre-training them across a range of grid topologies and contingency types. We found that a GCN model with auto-regressive moving average (ARMA) layers with a line graph representation of the grid offered the best predictive performance in predicting voltage magnitudes (VM) and voltage angles (VA). We introduced the concept of phantom nodes to consider disparate grid topologies with a varying number of nodes and lines. For pre-training the GCN ARMA model across a variety of topologies, distributed graphics processing unit (GPU) computing afforded us significant training scalability. The predictive performance of this model on grid topologies that were part of the training data is substantially better than the direct current (DC) approximation. Although direct application of the pre-trained model to topologies that are not part of the grid is not particularly satisfactory, fine-tuning with small amounts of data from a specific topology of interest significantly improves predictive performance. In general, this paper highlights the feasibility of training large-scale GNN models to assess the reliability of power grids by considering a wide variety of grid topologies and contingency types. With the advent of foundational models in ML and the exponential increase in GPU computing clusters, generalizable ML models will significantly enhance how utilities manage power systems and make decisions in real-time or near-real-time.

24 - POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2021

An automated procedure built on MTEX for reconstructing deformation twin hierarchies from electron backscattered diffraction datasets of heavily twinned microstructures

Here we present a set of algorithms built on the MTEX and MATLAB graph toolboxes for automatic reconstruction of deformation twin hierarchies from Electron Backscatter Diffraction (EBSD) datasets with a focus on developing methods for heavily twinned microstructures (twin fractions >0.5). The algorithms address key issues arising at large strains, mainly: missing twin relationships, grouping of heavily deformed grain fragments into families of similar orientation originating from a single initial grain, identification of parent fragments for large twin volume fractions, and classification of families having twin relationships with multiple families. To facilitate the development of these algorithms, large-grained ultra-high purity α-Ti deformed in compression along two directions is investigated. Graphs are utilized to handle non-local geometric merging and to represent relationships throughout the reconstruction process. When determining if a grain fragment is from the undeformed microstructure, the combined metrics of the fragment's orientation volume fraction in the initial texture and the directed graph centrality measure of out-closeness (the number of nodes reached in a graph from a given node) are essential. To address automation in reconstructing the sequence of twinning and relating fragments originating from a single grain in the initial microstructure, the twin family tree is formulated as a minimum spanning tree emanating from the initial grain family. A scheme constructing the distances associated with twin relationship comprising the spanning tree is developed, and a novel quasi-directional Prim spanning tree algorithm is used to determine the twin family tree. The procedure is demonstrated to significantly improve the level of automation in reconstructing twin hierarchies in heavily twinned microstructure compared to other methodologies in literature. The procedure can readily be applied to analyses of twinning in metals, as well as provide an approach for routinely extracting twin statistics at larger deformation levels than previously possible. Significantly, the procedure is demonstrated to be capable of identifying third generation twinning in α-Ti microstructures.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Knowledge Graph of RB-Tnseq Data from Fitness Browser (KP-DP1)

Motivation: Predicting microbial gene fitness across environmental conditions remains a central challenge for predictive phenomics and autonomous experimentation. Fitness assays generate large volumes of genotype–phenotype measurements difficult to integrate with experimental metadata and biological function in a form that supports mechanistic reasoning. Knowledge graphs offer a semantic framework for unifying modalities and enabling context-aware inference. Results: We build GIMME (Graph Inference for Microbial Metabolism Exploration), a semantically grounded knowledge graph that unifies gene fitness measurements spanning 10 Pseudomonas species with experimental metadata and biological context. Media are decomposed into chemical components and experiments carry structured links to natural-language descriptions. The resulting graph supports two inference modes: (1) symbolic graph traversal to surface candidate gene–environment and gene–chemical associations, and (2) learned inference using heterogeneous graph neural networks that propagate information across neighborhoods. We formulate link regression over (gene, media, experiment) triplets, combining learned gene embeddings with pretrained LLM sourced text embeddings of node descriptions to predict gene fitness. We then augment a baseline MLP with an auxiliary message-passing encoder (GraphSAGE/GAT) that propagates information over gene–protein–function and media–chemical subgraphs, and fuse the two pathways with a gated residual connection. This approach produces strong agreement with held-out fitness measurements (GraphSAGE Pearson r 0.74) while also highlighting inference challenges in extreme-fitness regimes. We aggregate GAT edge-attention weights by relation type and layer to estimate which biological and environmental relations most influence fitness predictions. Conclusion: This work explores using knowledge graphs as “context graphs” for microbial phenotype prediction. They provide a rich substrate which enables explainable retrieval of supporting evidence, and provides a natural bridge to autonomous workflows that prioritize the next experiment.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRSDate not supplied

Information Extraction on an Earth Science Knowledge Graphs with Semantic Parsing

Knowledge graphs are an important tool, both for representing knowledge and for retrieving information. Fundamentally, they are semantic networks that represent entities and relationships in the form of nodes and edges. A large corpus of natural language text can bebroken down into discrete entities and relationships to form a useful knowledge graph. Existing research breaks down text into a subject, object, and verb relationship triple. Although this is a useful first step, it loses much of the original contextual information encoded within the text. Our process uses a novel 7-tuple approach, in which elements of sentences are programmatically parsed into seven categories: initiator, impacted, receiver, beneficiary, result, and context. In this presentation, we show a knowledge graph built using this 7-tupleprocessing of an Earth science corpus. We explain the techniques used to create the graph and analyze its information retrieval capability while assessing the accuracy and limitations of the results.

Carson Davis↗
DOE OSTI2026

Microstructure Validation of Graph Theory Model-Derived Cooling Rates in the Wire Arc Additive Manufacturing of ER70S-6 Steel

Wire arc additive manufacturing (WAAM) enables high-rate fabrication of large metallic components, but spatial variations in thermal history can lead to microstructural heterogeneity that requires efficient process models to evaluate. This study evaluates whether cooling rates extracted from a graph theory model (GTM)-based thermal simulation are consistent with the microstructural evolution observed in an ER70S-6 WAAM wall. Thermal histories from the model were analyzed at selected build heights, and cooling rates were extracted from the final thermal excursion through the austenite phase field. Microstructures at corresponding locations were characterized using electron backscatter diffraction (EBSD) to quantify grain size distributions, and pearlite interlamellar spacing was used as an additional indicator of cooling behavior. The modeled cooling rates were highest near the substrate and generally decreased with build height, consistent with the observed reduction in the fine grain fraction and the progressive shift in the grain size distribution as build height increased. Pearlite spacing trends also supported the modeled cooling rate variation. These results indicate that GTM-derived thermal histories can be post-processed into metallurgically meaningful cooling rate estimates for WAAM steel builds and linked to dataset specific empirical grain size distribution relationships for process–thermal history–microstructure assessment.

36 MATERIALS SCIENCE↗
NASA NTRS1993

Experience with parametric binary dissection

Parametric Binary Dissection (PBD) is a new algorithm that can be used for partitioning graphs embedded in 2- or 3-dimensional space. It partitions explicitly on the basis of nodes + (lambda)x(edges cut), where lambda is the ratio of time to communicate over an edge to the time to compute at a node. The new algorithm is faster than the original binary dissection algorithm and attempts to obtain better partitions than the older algorithm, which only takes nodes into account. The performance of parametric dissection with plain binary dissection on 3 large unstructured 3-d meshes obtained from computational fluid dynamics and on 2 random graphs were compared. It was showm that the new algorithm can usually yield partitions that are substantially superior, but that its performance is heavily dependent on the input data.

Bokhari, Shahid H.↗
DOE OSTI2024

Addressing GPU memory limitations for Graph Neural Networks in High-Energy Physics applications

Introduction Reconstructing low-level particle tracks in neutrino physics can address some of the most fundamental questions about the universe. However, processing petabytes of raw data using deep learning techniques poses a challenging problem in the field of High Energy Physics (HEP). In the Exa.TrkX Project, an illustrative HEP application, preprocessed simulation data is fed into a state-of-art Graph Neural Network (GNN) model, accelerated by GPUs. However, limited GPU memory often leads to Out-of-Memory (OOM) exceptions during training, due to the large size of models and datasets. This problem is exacerbated when deploying models on High-Performance Computing (HPC) systems designed for large-scale applications. Methods We observe a high workload imbalance issue during GNN model training caused by the irregular sizes of input graph samples in HEP datasets, contributing to OOM exceptions. We aim to scale GNNs on HPC systems, by prioritizing workload balance in graph inputs while maintaining model accuracy. Our paper introduces diverse balancing strategies aimed at decreasing the maximum GPU memory footprint and avoiding the OOM exception, across various datasets. Results Our experiments showcase memory reduction of up to 32.14% compared to the baseline. We also demonstrate the proposed strategies can avoid OOM in application. Additionally, we create a distributed multi-GPU implementation using these samplers to demonstrate the scalability of these techniques on the HEP dataset. Discussion By assessing the performance of these strategies as data loading samplers across multiple datasets, we can gauge their effectiveness in both single-GPU and distributed environments. Our experiments, conducted on datasets of varying sizes and across multiple GPUs, broaden the applicability of our work to various GNN applications that handle input datasets with irregular graph sizes.

Lee, Claire Songhyun↗