Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Graph Processing”

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 325 records · Page 18
All sourcesNASA NTRSDOE OSTI
DOE OSTI2018

Finding Maximum Cliques on the D-Wave Quantum Annealer

This work assesses the performance of the D-Wave 2X (DW) quantum annealer for finding a maximum clique in a graph, one of the most fundamental and important NP-hard problems. Because the size of the largest graphs DW can directly solve is quite small (usually around 45 vertices), we also consider decomposition algorithms intended for larger graphs and analyze their performance. For smaller graphs that fit DW, we provide formulations of the maximum clique problem as a quadratic unconstrained binary optimization (QUBO) problem, which is one of the two input types (together with the Ising model) acceptable by the machine, and compare several quantum implementations to current classical algorithms such as simulated annealing, Gurobi, and third-party clique finding heuristics. We further estimate the contributions of the quantum phase of the quantum annealer and the classical post-processing phase typically used to enhance each solution returned by DW. We demonstrate that on random graphs that fit DW, no quantum speedup can be observed compared with the classical algorithms. On the other hand, for instances specifically designed to fit well the DW qubit interconnection network, we observe substantial speed-ups in computing time over classical approaches.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Distributed-Memory Fast Maximal Independent Set

The Maximal Independent Set (MIS) graph problem arises in many applications such as computer vision, information theory, molecular biology, and process scheduling. The growing scale of MIS problems suggests the use of distributed-memory hardware as a cost-effective approach to providing necessary compute and memory resources. Luby proposed four randomized algorithms to solve the MIS problem. All those algorithms are designed focusing on shared-memory machines and are analyzed using the PRAM model. These algorithms do not have direct efficient distributed-memory implementations. In this paper, we extend two of Luby’s seminal MIS algorithms, “Luby(A)” and “Luby(B),” to distributed-memory execution, and we evaluate their performance. We compare our results with the “Filtered MIS” implementation in the Combinatorial BLAS library for two types of synthetic graph inputs.

distributed memory parallel↗
DOE OSTI2021

Understanding Discrete Fracture Networks Through Spectral Graph Theory

Discrete Fracture Network models (DFNs) are used to simulate fluid flow and particle transport through fracture networks in low permeability rock. Understanding these processes are essential in many subsurface applications, such as environmental restoration of contaminated fractured media, CO 2 sequestration, detection of low-level nuclear tests, and hydrocarbon extraction. Compared with other models, DFNs allow for incorporation of a wider range of network characteristics but have substantially greater computation cost. These networks can be represented with graphs, allowing the use of graph theory tools to study the networks. I used Python to simulate flow and transport on a range of DFNs and analyzed these networks using methods from network analysis and spectral graph theory. My purpose was to find ways to gain insight about flow and transport on DFNs using these graph representations, bypassing the computationally intensive meshing typically required. My work is still in progress, but I have discovered several interesting trends and patterns that I believe could be useful towards my goal. If I am able to bring these results to fruition, they will aid subsurface geologists in extracting flow and transport information about fracture networks more efficiently.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Language models for materials discovery and sustainability: Progress, challenges, and opportunities

Significant advancements have been made in one of the most critical branches of artificial intelligence: natural language processing (NLP). These advancements are exemplified by the remarkable success of OpenAI’s GPT-3.5/4 and the recent release of GPT-4.5, which have sparked a global surge of interest akin to an NLP gold rush. Here, in this article, we offer our perspective on the development and application of NLP and large language models (LLMs) in materials science. We begin by presenting an overview of recent advancements in NLP within the broader scientific landscape, with a particular focus on their relevance to materials science. Next, we examine how NLP can facilitate the understanding and design of novel materials and its potential integration with other methodologies. To highlight key challenges and opportunities, we delve into three specific topics: (i) the limitations of LLMs and their implications for materials science applications, (ii) the creation of a fully automated materials discovery pipeline, and (iii) the potential of GPT-like tools to synthesize existing knowledge and aid in the design of sustainable materials.

36 MATERIALS SCIENCE↗
DOE OSTI2018

A Heterogeneity-Aware Task Scheduler for Spark

Big data processing systems such as Spark are employed in an increasing number of diverse applications-such as machine learning, graph computation, and scientific computing-each with dynamic and different resource needs. These applications increasingly run on heterogeneous hardware, e.g., with out-of-core accelerators. However, big data platforms do not factor in the multi-dimensional heterogeneity of applications and hardware. This leads to a fundamental mismatch between the application and hardware characteristics, and the resource scheduling adopted in big data platforms. For example, Hadoop and Spark consider only data locality when assigning tasks to nodes, and typically disregard the hardware capabilities and suitability to specific application requirements. In this paper, we present RUPAM, a heterogeneity-aware task scheduling system for big data platforms, which considers both task-level resource characteristics and underlying hardware characteristics, as well as preserves data locality. RUPAM adopts a simple yet effective heuristic to decide the dominant scheduling factor (e.g., CPU, memory, or I/O), given a task in a particular stage. Our experiments show that RUPAM is able to improve the performance of representative applications by up to 62.3% compared to the standard Spark scheduler.

Computer Science↗
DOE OSTI2018

Resource Aware Intelligent Network Services (RAINS) Final Technical Report

The Resource Aware Intelligent Network Services (RAINS) project conducted research and developed technologies in the area of cyber infrastructure resource modeling and computation. The goal of this work was to provide a foundation to enable intelligent, software defined services which spanned the network AND the resources which connect to the network. A Multi-Resource Service Plane (MRSP) was defined, which allows resource owners/managers to locate and place themselves from a topology and service availability perspective within the dynamic networked cyberinfrastructure ecosystem. The MRSP enables the presentation of integrated topology views and computation results which can include resources across the spectrum of compute, storage, and networks. The RAINS project developed MSRP includes the following key components: i) Multi-Resource Service (MRS) Ontology/Multi-Resource Markup Language (MRML), ii) Resource Computation Engine (RCE), iii) Modular Driver Framework (to allow integration of a variety of external resources). The MRS/MRML is a general and extensible modeling framework that allows for resource owners to model, or describe, a wide variety of resource types. All resources are described using three categories of elements: Resources, Services, and Relationships between the elements. This modeling framework defines a common method for the transformation of cyber infrastructure resources into data in the form of MRML models. In order to realize this infrastructure datification, the RAINS project developed a model based computation system, i.e. “RAINS Computation Engine (RCE)”. The RCE has the ability to ingest, process, integrate, and compute based on automatically generated MRML models. The RCE interacts with the resources thru system drivers which are specific to the type of external network or resource controller. The RAINS project developed a modular and pluggable driver system which facilities a variety of resource controllers to automatically generate, maintain, and distribute MRML based resource descriptions. Once all of the resource topologies are absorbed by the RCE, a connected graph of the full distributed system topology is constructed, which forms the basis for computation and workflow processing. The RCE includes a Modular Computation Element (MCE) framework which allows for tailoring of the computation process to the specific set of resources under control, and the services desired. The input and output of an MCE are both model data based on MRS/MRML ontology and schema. Some of the RAINS project accomplishments include: Development of general and extensible multi-resource modeling framework; Design of a Resource Computation Engine (RCE) system which includes the following key capabilities; Absorb a variety of multi-resource model types and build integrated models; Novel architecture which uses model based communications across the full stack for all Flexible provision of abstract or intent based user facing interfaces; Workflow processing based on model descriptions; Release of the RCE as an open source software; Deployment of RCE in the University of Maryland/Mid-Atlantic Crossroad ScienceDMZ in prototype mode with a plan under way to transition to production; Deployment at the Argonne National Laboratory DTN Facility in prototype mode; Selection of RCE by the DOE SENSE (SDN for End-to-end Networked Science at the Exascale) project as the basis for their orchestration service.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

Social Sensor Analytics: Making Sense of Network Models in Social Media

Social networks can be thought of as noisy sensor networks mapping real world information to the web. Owing to the extensive body of literature in sensor network analysis, this work sought to apply several novel and traditional methods in sensor network analysis for the purposes of efficiently interrogating social media data streams from raw data. We carefully revisit our definition of a social media signal from previous work both in terms of time-varying features within the data and the networked nature of the medium. Further, we detail our analysis of global patterns in Twitter over the months of November 2013 and June 2014, detect and categorize events, and illustrate how these analyses can be used to inform graph-based models of Twitter, namely using a recent network influence model called PhySense: similar to PageRank but tuned to behavioral analysis by leveraging a sociologically inspired probabilistic model. We ultimately identify forms of information dissemination via analysis of time series and dynamic graph spectra and corroborate these findings through manual investigation of the data as a requisite step in modeling the diffusion process with PhySense. We hope to sufficiently characterize global behavior in a medium such as Twitter as a means of learning global model parameters one may use to predict or simulate behavior on a large scale. We have made our time series and dynamic graph analytical code available via a GitHub repository https://github.com/cpatdowling/salsa and our data are available upon request.

event detection↗
DOE OSTI2020

Automated Realistic Testbed Synthesis for Power System Communication Networks based on Graph Metrics

Communication networks are integral to modern power grid operations and are becoming increasingly critical as grid dynamics speed up and as more controls become closed-loop in form. Focusing on the interdependence between the physical grid and the communication system, we identify several key characteristics and typical graph properties based on analysis of a real communication system for a power grid. Moreover, an automated process was provided for synthetic testbed in the NS-3 simulator for the power system test case, and its network characteristics have been further derived for power system monitoring and control applications

Graph Analysis, Synthetic Testbed, communication n↗
DOE OSTI2025

RE-INTEGRATE EMT Simulation Tool: Input Data Processing Layer for Bulk Power System

This paper introduces an advanced input data processing layer for EMT simulations of large-scale bulk power systems. The paper proposes two versions of the RE-INTEGRATE EMT simulation tool, RE-INTEGRATE Gen-0 and RE-INTEGRATE Gen-1, which are developed to enhance simulation generalizability, scalability, and accuracy. The framework leverages a generic class design for components to incorporate linear equations, which are generated by discretizing the Differential-Algebraic Equations (DAEs) that represent the dynamics of the components. In addition, the framework employs a parsing algorithm that parses a power system’s raw and dyr files to generate a connectivity graph which is then traversed to form the overall system’s dynamics. The proposed input data processing layer is used to simulate the IEEE 39-bus test system. The obtained results demonstrate the framework’s capability to achieve simulation scalability and accuracy. Further, the results indicate that EMT simulations performed using the proposed automations can effectively handle complex grid configurations.

Mishra, Rahul [ORNL] (ORCID:0000000328205932)↗
DOE OSTI2016

Macro Supply Chain Lifecycle Decision Analytics

This report summarizes a two-year LDRD project that investigated the problem of representing complex supply chains, identifying the worst risks and evaluating mitigation options. Our team developed a framework that includes a representation for business processes, risk assessment questions, risk indicators and methods for analyzing and summarizing the data. In our approach, the Process Matrix represents an overall supply chain for an end product in a high-level, tabular form. It connects the various touch-points of a business process including people, external vendors, tools, storage locations and transportation services while capturing the flow of both physical and intellectual artifacts. We believe these direct connections are exactly the things that a process owner can typically control. These material flows (both physical and intellectual) are also represented in a graph. This enables us to use graph-oriented analysis such as fault tree analysis and attack graph generation. Our approach is top-down, which helps users to get a more holistic understanding for a given amount of resources. Understanding the provenance of materials is difficult and it is easy to exhaust the analysts' resources. Rather than a tool to do vendor analysis or product comparison, our framework enables an enterprise-level analysis. The risk assessment questionnaires have a varying levels of detail and cover various aspects of the supply chain such as process steps, artifacts, suppliers, etc. and connections between these aspects such as artifact-storage, artifact-supplier, etc. Each question is further associated with one of seven risk indicators which can be used to summarize the risk. These risk indicators can also be weighted to reflect a user's concerns. We have successfully applied our framework to several use cases in various stages of its development and provided valuable insights to our partners, but it can also be applied to other complex systems outside of the supply chain security problem.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
DOE OSTI2021

Scalable statistical inference of photometric redshift via data subsampling

Handling big data has largely been a major bottleneck in traditional statistical models. Consequently, when accurate point prediction is the primary target, machine learning models are often preferred over their statistical counterparts for bigger problems. But full probabilistic statistical models often outperform other models in quantifying uncertainties associated with model predictions. We develop a data-driven statistical modeling framework that combines the uncertainties from an ensemble of statistical models learned on smaller subsets of data carefully chosen to account for imbalances in the input space. We demonstrate this method on a photometric redshift estimation problem in cosmology, which seeks to infer a distribution of the redshift—the stretching effect in observing the light of far-away galaxies—given multivariate color information observed for an object in the sky. Our proposed method performs balanced partitioning, graph-based data subsampling across the partitions, and training of an ensemble of Gaussian process models.

data subsampling↗
DOE OSTI2025

Scalable training of trustworthy and energy-efficient predictive graph foundation models for atomistic materials modeling: a case study with HydraGNN

We present our work on developing and training scalable, trustworthy, and energy-efficient predictive graph foundation models (GFMs) using HydraGNN, a multi-headed graph convolutional neural network architecture. HydraGNN expands the boundaries of graph neural network (GNN) computations in both training scale and data diversity. It abstracts over message passing algorithms, allowing both reproduction of and comparison across algorithmic innovations that define nearest-neighbor convolution in GNNs. This work discusses a series of optimizations that have allowed scaling up the GFMs training to tens of thousands of GPUs on datasets consisting of hundreds of millions of graphs. Our GFMs use multitask learning (MTL) to simultaneously learn graph-level and node-level properties of atomistic structures, such as energy and atomic forces. Using over 154 million atomistic structures for training, we illustrate the performance of our approach along with the lessons learned on two state-of-the-art US Department of Energy (US-DOE) supercomputers, namely the Perlmutter petascale system at the National Energy Research Scientific Computing Center and the Frontier exascale system at Oak Ridge Leadership Computing Facility. The HydraGNN architecture enables the GFM to achieve near-linear strong scaling performance using more than 2000 GPUs on Perlmutter and 16,000 GPUs on Frontier.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Disordered topological graphs enhancing nonlinear phenomena

Complex networks play a fundamental role in understanding phenomena from the collective behavior of spins, neural networks, and power grids to the spread of diseases. Topological phenomena in such networks have recently been exploited to preserve the response of systems in the presence of disorder. We propose and demonstrate topological structurally disordered systems with a modal structure that enhances nonlinear phenomena in the topological channels by inhibiting the ultrafast leakage of energy from edge modes to bulk modes. We present the construction of the graph and show that its dynamics enhances the topologically protected photon pair generation rate by an order of magnitude. Disordered nonlinear topological graphs will enable advanced quantum interconnects, efficient nonlinear sources, and light-based information processing for artificial intelligence.

42 ENGINEERING↗
DOE OSTI2016

URBAN-NET: A Network-based Infrastructure Monitoring and Analysis System for Emergency Management and Public Safety

Abstract Critical Infrastructures (CIs) such as energy, water, and transportation are complex networks that are crucial for sustaining day-to-day commodity flows vital to national security, economic stability, and public safety. The nature of these CIs is such that failures caused by an extreme weather event or a man-made incident can trigger widespread cascading failures, sending ripple effects at regional or even national scales. To minimize such effects, it is critical for emergency responders to identify existing or potential vulnerabilities within CIs during such stressor events in a systematic and quantifiable manner and take appropriate mitigating actions. We present here a novel critical infrastructure monitoring and analysis system named URBAN-NET. The system includes a software stack and tools for monitoring CIs, pre-processing data, interconnecting multiple CI datasets as a heterogeneous network, identifying vulnerabilities through graph-based topological analysis, and predicting consequences based on what-if simulations along with visualization. As a proof-of-concept, we present several case studies to show the capabilities of our system. We also discuss remaining challenges and future work.

Lee, Sangkeun↗
DOE OSTI2017

Formalization of the functional analysis methodology to improve NPP I and C system design process - 244

Functional analysis is a universal methodology, which is able to support NPP and I and C design process. The paper proposes an approach to combining the functional analysis process and I and C design process into a single I and C functional design process. Operation of a system is considered as a set of functions providing not only the achievement of goal, but also keeping material and energy balances. A notation for graphical representation of semantic network (hierarchy of functions) describing a system is suggested. This network includes four levels, namely functional goals, abstract functions level, process functions level, and equipment level. The elements of the hierarchy are connected by three kinds of relationships, namely: parent-child, main-support, and cause-consequence. Then the semantic network (hierarchy) of functions is converted into a functional network which looks as a directed graph. This graph is used for development of control algorithm and for building of instrumentation and control functions hierarchy. Farther analysis of process functions reveals a set of criteria indicating real-time status and effectiveness of functions. The criteria are used for allocation of instruments and actuators. The process of functional analysis and design is illustrated in application to a simple heat generation system. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
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↗
DOE OSTI2022

A graph neural network (GNN) approach to basin-scale river network learning: the role of physics-based connectivity and data fusion

Abstract. Rivers and river habitats around the world are under sustained pressure from human activities and the changing global environment. Our ability to quantify and manage the river states in a timely manner is critical for protecting the public safety and natural resources. In recent years, vector-based river network models have enabled modeling of large river basins at increasingly fine resolutions, but are computationally demanding. This work presents a multistage, physics-guided, graph neural network (GNN) approach for basin-scale river network learning and streamflow forecasting. During training, we train a GNN model to approximate outputs of a high-resolution vector-based river network model; we then fine-tune the pretrained GNN model with streamflow observations. We further apply a graph-based, data-fusion step to correct prediction biases. The GNN-based framework is first demonstrated over a snow-dominated watershed in the western United States. A series of experiments are performed to test different training and imputation strategies. Results show that the trained GNN model can effectively serve as a surrogate of the process-based model with high accuracy, with median Kling–Gupta efficiency (KGE) greater than 0.97. Application of the graph-based data fusion further reduces mismatch between the GNN model and observations, with as much as 50 % KGE improvement over some cross-validation gages. To improve scalability, a graph-coarsening procedure is introduced and is demonstrated over a much larger basin. Results show that graph coarsening achieves comparable prediction skills at only a fraction of training cost, thus providing important insights into the degree of physical realism needed for developing large-scale GNN-based river network models.

54 ENVIRONMENTAL SCIENCES↗