SEARCH · Search NASA
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.
This award was relinquished by the University of Alabama at Birmingham (UAB) shortly after transferred to Indiana University. No research activities were conducted at UAB under this award. No personnel were supported, no funds were expended, and no publications or other research outputs were produced at UAB.
The Partitioned Global Address Space (PGAS) model is well suited for executing irregular applications on cluster-based systems, due to its efficient support for short, one-sided messages. Separately, the actor model has been gaining popularity as a productive asynchronous message-passing approach for distributed objects in enterprise and cloud computing platforms, typically implemented in languages such as Erlang, Scala or Rust. To the best of our knowledge, there has been no past work on using the actor model to deliver both productivity and scalability to irregular PGAS applications with large number of small messages. In this paper, we introduce a new programming system for PGAS applications, in which point-to-point remote operations can be expressed as fine-grained asynchronous actor messages. In our approach, the programmer does not need to worry about programming complexities related to message aggregation and termination detection. Our approach can be viewed as extending the classical Bulk Synchronous Parallelism model with fine-grained asynchronous communications within a phase or superstep. Here, we believe that our approach offers a desirable point in the productivity-performance space for PGAS applications, with more scalable performance and higher productivity relative to past approaches. Specifically, for seven irregular mini-applications from the Bale Kernels and three graph kernels executed using 2048 cores in the NERSC Cori system, our approach shows geometric mean performance improvements of ≥ 20X relative to standard PGAS versions (UPC and OpenSHMEM) while maintaining comparable productivity to those versions.
The rapid growth of large-scale datasets in fields like biology and social networks has driven the need for advanced graph analytics techniques. Community detection, a fundamental task in graph analytics, identifies closely connected groups of nodes within a network, providing valuable insights across various disciplines. This study focuses on two classic community detection methods, the Louvain algorithm and Markov Clustering (MCL), and evaluates the performance of two prominent distributed community detection algorithms: HiPDPL-GPU, our prior implementation, and HipMCL. We conduct experiments on GPU-accelerated heterogeneous HPC systems, Summit and Frontier, to assess their performance under varying conditions. Our objective is to identify the strengths and weaknesses of these algorithms in terms of scalability, and quality of solutions. We evaluate these algorithms on a diverse set of 70+ networks spanning 13 domains, with sizes ranging up to 4.2 billion edges. Our results demonstrate that HiPDPL-GPU consistently outperforms HipMCL, especially for large-scale networks. HiPDPL-GPU achieves significantly faster runtimes (47x to 1439x), higher modularity scores, and improved scalability. These findings highlight HiPDPL-GPU as a promising solution for efficient and effective large-scale graph analytics in diverse application domains, and provide insights into the feasibility of using MCL-based approaches for certain application domains.
Parallel computing plays a pivotal role in the efficient processing of large-scale graphs. Complex network analysis stands as a capti- vating research frontier, holding promise across diverse scientific domains such as sociology, biology, online media, and recommenda- tion systems. In this era, Machine Learning (ML) and Deep Learning (DL) have emerged as indispensable tools, underpinning remarkable technological achievements. Within this dynamic landscape, my research revolves around advancing parallel algorithms tailored for large-scale graph operations. To achieve this, I harness the power of cutting-edge technologies including OpenMP, MPI, HIP, and CUDA, on the High-Performance Computing (HPC) platforms to unlock optimal performance. I also apply ML/DL techniques to HPC operational data, to streamline the monitoring and maintenance of supercomputers, alleviating the complexities associated with their upkeep and enhancing user support. My research echoes the syn- ergy between parallel computing, large-scale graph analysis, and ML/DL, improving computational efficiency and user experience.
Because large unstructured datasets is important for many science domains, distributed graph analytics is critical to many scientists. Unfortunately, obtaining scaling and performance for irregular communication is challenging because contemporary network interconnects are primarily designed to maximize bandwidths of fixed-neighborhoods large-message exchanges (e.g., stencils). Although there is no consensus on the “best” network topologies for irregular communication, unstructured graph-based interconnects can be more suitable. We analyze three popular graph workloads – clustering, pattern enumeration, and traversal — on comparable networks (in terms of resources and costs) constructed from Jellyfish Random Regular, Dragonfly and Fat tree topologies, varying the routing algorithms. Using packet-level simulations, we demonstrate up to 60% improvement in communication time with Jellyfish due to diversity of the short paths between arbitrary endpoints, which can reduce overall network stalls and congestion.
Graph neural networks (GNNs), an emerging class of machine learning models for graphs, have gained popularity for their superior performance in various graph analytical tasks. Mini-batch training is commonly used to train GNNs on large graphs, and data parallelism is the standard approach to scale mini-batch training across multiple GPUs. Data parallel approaches contain redundant work as subgraphs sampled by different GPUs contain significant overlap. To address this issue, we introduce a hybrid parallel mini-batch training paradigm called Split parallelism. Split parallelism avoids redundant work by splitting the sampling, loading, and training of each mini-batch across multiple GPUs. Split parallelism, however, introduces communication overheads that can be more than the savings from removing redundant work. We further present a lightweight partitioning algorithm that probabilistically minimizes these overheads. We implement spllit parllelism in GSplit and show that it outperforms state-of-the-art mini-batch training systems like DGL, Quiver, and P3.
Large-scale data analytics workflows ingest massive input data into various data structures, including graphs and key-value datastores. These data structures undergo multiple transformations and computations and are typically reused in incremental and iterative analytics workflows. Persisting in-memory views of these data structures enables reusing them beyond the scope of a single program run while avoiding repetitive raw data ingestion overheads. Memory-mapped I/O enables persisting in-memory data structures without data serialization and deserialization overheads. However, memory-mapped I/O lacks the key feature of persisting consistent snapshots of these data structures for incremental ingestion and processing. The obstacles to efficient virtual memory snapshots using memory-mapped I/O include background writebacks outside the application’s control, and the significantly high storage footprint of such snapshots. To address these limitations, we present Privateer, a memory and storage management tool that enables storage-efficient virtual memory snapshotting while also optimizing snapshot I/O performance. Here, we integrated Privateer into Metall, a state-of-the-art persistent memory allocator for C++, and the Lightning Memory-Mapped Database (LMDB), a widely-used key-value datastore in data analytics and machine learning. Privateer optimized application performance by 1.22× when storing data structure snapshots to node-local storage, and up to 16.7× when storing snapshots to a parallel file system. Privateer also optimizes storage efficiency of incremental data structure snapshots by up to 11× using data deduplication and compression.
The integration of the global Photovoltaic (PV) market with real time data-loggers has enabled large scale PV data analytical pipelines for power forecasting and long-term reliability assessment of PV fleets. Nevertheless, the performance of PV data analysis heavily depends on the quality of PV timeseries data. This paper proposes a novel Spatio-Temporal Denoising Graph Autoencoder (STD-GAE) framework to impute missing PV Power Data. STDGAE exploits temporal correlation, spatial coherence, and value dependencies from domain knowledge to recover missing data. It is empowered by two modules. (1) To cope with sparse yet various scenarios of missing data, STD-GAE incorporates a domain-knowledge aware data augmentation module that creates plausible variations of missing data patterns. This generalizes STD-GAE to robust imputation over different seasons and environment. (2) STD-GAE nontrivially integrates spatiotemporal graph convolution layers (to recover local missing data by observed “neighboring” PV plants) and denoising autoencoder (to recover corrupted data from augmented counterpart) to improve the accuracy of imputation accuracy at PV fleet level. We have evaluated our proposed model on two realworld PV datasets. Experimental results show that STD-GAE can achieve a gain of 43.14% in imputation accuracy and remains less sensitive to missing rate, different seasons, and missing scenarios, compared with state-of-the-art data imputation methods such as MIDA and LRTC-TNN.
Abstract Power systems are facing an increasing number of cyber incidents, potentially leading to damaging consequences to both physical and cyber aspects. However, the development of analytical methods for the study of large‐scale power infrastructures as cyber‐physical systems is still in its early stages. Drawing inspiration from machine‐learning techniques, the authors introduce a method inspired by the principles of graph embedding that is tailored for quantitative risk assessment and the exploration of possible mitigation strategies of large‐scale cyber‐physical power systems. The primary advantage of the graph embedding approach lies in its ability to generate numerous random walks on a graph, simulating potential access paths. Meanwhile, it enables capturing high‐dimensional structures in low‐dimensional spaces, facilitating advanced machine‐learning applications, and ensuring scalability and adaptability for comprehensive network analysis. By employing this graph embedding‐based approach, the authors present a structured and methodical framework for risk assessment in cyber‐physical systems. The proposed graph embedding‐based risk analysis framework aims to provide a more insightful perspective on cyber‐physical risk assessment and situation awareness for power systems. To validate and demonstrate its applicability, the method has been tested on two cyber‐physical power system models: the Western System Coordinating Council (WSCC) 9‐Bus System and the Illinois 200‐Bus System , thereby showing its advantages in enhancing the accuracy of risk analysis and comprehensiveness of situational awareness.
Unlike circuit parameter and sizing optimizations, the automated design of analog circuit topologies poses significant challenges for learning-based approaches. One challenge arises from the combinatorial growth of the topology space with circuit size, which limits the topology optimization efficiency. Moreover, traditional circuit evaluation methods are time-consuming, while the presence of data discontinuity in the topology space makes the accurate prediction of circuit performance exceptionally difficult for unseen topologies. To tackle these challenges, we design a novel Graph-Transformer-based Network (GTN) as the surrogate model for circuit evaluation, offering a substantial acceleration in the speed of circuit topology optimization without sacrificing performance. Our GTN model architecture is designed to embed voltage changes in circuit loops and current flows in connected devices, enabling accurate performance predictions for circuits with unseen topologies. To address the cold start problem when scaling GTN to large-scale circuits, we further introduce a curriculum learning strategy that progressively trains GTN from small-scale to large-scale circuits. This approach enables the model to first learn fundamental physical principles from simpler topologies and gradually adapt to complex configurations, effectively bridging the circuit complexity gap and improving prediction accuracy. Taking the power converter circuit design as an experimental task, our GTN model significantly outperforms an analytical approach and baseline methods directly utilizing graph neural networks. Furthermore, GTN achieves less than 5% relative error and 196× speed-up compared with high-fidelity simulation. Notably, our GTN surrogate model empowers an automatic circuit design framework to discover circuits of comparable quality to those identified through high-fidelity simulation while reducing the time required by up to 98.2%. With curriculum learning, the enhanced GTN achieves a 51% improvement for performance prediction of large-scale circuits compared to the GTN model without this strategy. These advancements establish GTN as a scalable framework for automated analog circuit design across varying circuit complexity levels.
While standardized ontologies like the Brick schema address data heterogeneity in Building Automation Systems (BAS), accessing this semantic data remains a challenge as domain experts often lack the expertise to formulate complex SPARQL queries. To bridge this gap, we present BrickQA, a Large Language Model (LLM)-based framework that translates natural language into executable SPARQL queries through structured query decomposition, dynamic schema exploration, and inline validation. BrickQA utilizes an iterative reasoning agent to actively navigate graph topology through dynamic exploration actions without requiring exhaustive context injection or model fine-tuning. This approach effectively mitigates hallucinations, particularly in large-scale building knowledge graphs. Empirical evaluation on BuildingQA, a standardized benchmark, demonstrates that BrickQA significantly outperforms ReAct baselines, delivering a 0.291–0.355 absolute F1 improvement while achieving 3 × –12.7 × higher token cost-efficiency. Beyond these metrics, the framework maintains structural fidelity across heterogeneous buildings and remains resilient to ambiguous queries without requiring site-specific fine-tuning. Furthermore, a case study on operational analytics validates the framework’s capability to handle temporal and aggregation constraints, effectively transforming abstract semantic models into actionable facility management insights.1
Abstract Bridging the gap between genetic variations, environmental determinants, and phenotypic outcomes is critical for supporting clinical diagnosis and understanding mechanisms of diseases. It requires integrating open data at a global scale. The Monarch Initiative advances these goals by developing open ontologies, semantic data models, and knowledge graphs for translational research. The Monarch App is an integrated platform combining data about genes, phenotypes, and diseases across species. Monarch's APIs enable access to carefully curated datasets and advanced analysis tools that support the understanding and diagnosis of disease for diverse applications such as variant prioritization, deep phenotyping, and patient profile-matching. We have migrated our system into a scalable, cloud-based infrastructure; simplified Monarch's data ingestion and knowledge graph integration systems; enhanced data mapping and integration standards; and developed a new user interface with novel search and graph navigation features. Furthermore, we advanced Monarch's analytic tools by developing a customized plugin for OpenAI’s ChatGPT to increase the reliability of its responses about phenotypic data, allowing us to interrogate the knowledge in the Monarch graph using state-of-the-art Large Language Models. The resources of the Monarch Initiative can be found at monarchinitiative.org and its corresponding code repository at github.com/monarch-initiative/monarch-app.
Efficient prediction of sampling-intensive thermodynamic properties is needed to evaluate material performance and permit high-throughput materials modeling for a diverse array of technology applications. To alleviate the prohibitive computational expense of high-throughput configurational sampling with density functional theory (DFT), surrogate modeling strategies like cluster expansion are many orders of magnitude more efficient but can be difficult to construct in systems with high compositional complexity. We therefore employ minimal-complexity graph neural network models that accurately predict and can even extrapolate to out-of-train distribution formation energies of DFT-relaxed structures from an ideal (unrelaxed) crystallographic representation. This enables the large-scale sampling necessary for various thermodynamic property predictions that may otherwise be intractable and can be achieved with small training data sets. Two exemplars, optimizing the thermodynamic stability of low-density high-entropy alloys and modulating the plateau pressure of hydrogen in metal alloys, demonstrate the power of this approach, which can be extended to a variety of materials discovery and modeling problems.
Automated three-dimensional molecular construction from two-dimensional graph representations is critical to high-throughput discovery eIorts. Software capabilities in this area have accelerated research across fields ranging from protein design and drug discovery to transition metal catalyst development. When Architector was first introduced, it uniquely enabled high-throughput, chemically relevant three-dimensional construction of f-element complexes. Since its introduction, Architector has been applied in large-scale computational campaigns, targeted studies in critical mineral extraction, and artificial intelligence-driven discovery eIorts.
Molecular conformer generation (MCG) is an important task in cheminformatics and drug discovery. The ability to efficiently generate low-energy 3D structures can avoid expensive quantum mechanical simulations, leading to accelerated virtual screenings and enhanced structural exploration. Several generative models have been developed for MCG, but many struggle to consistently produce high-quality conformers for meaningful downstream applications. To address these issues, we introduce CoarsenConf, which coarse-grains molecular graphs based on torsional angles and integrates them into an SE(3)-equivariant hierarchical variational autoencoder. Through equivariant coarse-graining, we aggregate the fine-grained atomic coordinates of subgraphs connected via rotatable bonds, creating a variable-length coarse-grained latent representation. Our model uses a novel aggregated attention mechanism to restore fine-grained coordinates from the coarse-grained latent representation, enabling efficient generation of accurate conformers. Furthermore, we evaluate the chemical and biochemical quality of our generated conformers on multiple downstream applications, including property prediction and large-scale oracle-based protein docking. Overall, CoarsenConf generates more accurate conformer ensembles compared to prior generative models.
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 .
Nuclear power plants collect large amounts of equipment reliability data elements that contain information on the statuses of component, assets, and systems. All these data elements precisely record asset and system performance and health throughout the lifecycle of those assets and systems. However, several challenges have proved to be roadblocks to this process. While some of these challenges are technical in nature (i.e., data are often distributed over several physical servers or databases), others are conceptual in nature (i.e., data elements come in different formats, numeric or textual), and measured values have different scales (e.g., vibration spectra and oil temperature). This paper directly focuses on the integration of numeric and textual data elements in order to assist plant system engineers in analyzing equipment reliability data. This task begins with preprocessing the data by extracting knowledge from textual data via natural language processing methods and quantifying system, asset, and component health based on numeric data. We then employed model-based system engineering (MBSE) models of systems and assets to identify their architecture and functional (i.e., cause and effect) relations. Data elements were then associated with a single MBSE graph element, based on their nature. This bonding of MBSE models and data elements constitutes a first-of-its-kind knowledge graph of a nuclear power plants system, with data elements being organized in a structured manner that enables system engineers to identify cause-effect trends in data elements and carry out appropriate actions in response.
Atomic structure prediction and associated property calculations are the bedrock of chemical physics. Since high-fidelity ab initio modeling techniques for computing the structure and properties can be prohibitively expensive, this motivates the development of machine-learning (ML) models that make these predictions more efficiently. Training graph neural networks over large atomistic databases introduces unique computational challenges such as the need to process millions of small graphs with variable size and support communication patterns that are distinct from learning over large graphs such as social networks. We demonstrate a novel hardware-software co-design approach to scale up the training of atomistic graph neural networks (GNN) for structure and property prediction. First, to eliminate redundant computation and memory associated with alternative padding techniques and to improve throughput via minimizing communication, we formulate the effective coalescing of the batches of variable-size atomistic graphs as the bin packing problem and introduce a hardware-agnostic algorithm to pack these batches. In addition, we propose hardware-specific optimizations including a planner and vectorization for the gather-scatter operations targeted for Graphcore’s Intelligence Processing Unit (IPU), as well as model-specific optimizations such as merged communication collectives and optimized softplus. Putting these all together, we demonstrate the effectiveness of the proposed co-design approach by providing an implementation of a well-established atomistic GNN on the Graphcore IPUs. We evaluate the training performance on multiple atomistic graph databases with varying degrees of graph counts, sizes and sparsity. Here, we demonstrate that such a co-design approach can reduce the training time of atomistic GNNs and can improve the performance by up to 1.5× compared to the baseline implementation of the model on the IPUs. Additionally, we compare our IPU implementation with a Nvidia GPU-based implementation and show that our atomistic GNN implementation on the IPUs can run 1.8× faster on average compared to the execution time on the GPUs.