Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “scalable deep learning”

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 91 records · Page 5
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Learning local equivariant representations for large-scale atomistic dynamics

Abstract A simultaneously accurate and computationally efficient parametrization of the potential energy surface of molecules and materials is a long-standing goal in the natural sciences. While atom-centered message passing neural networks (MPNNs) have shown remarkable accuracy, their information propagation has limited the accessible length-scales. Local methods, conversely, scale to large simulations but have suffered from inferior accuracy. This work introduces Allegro, a strictly local equivariant deep neural network interatomic potential architecture that simultaneously exhibits excellent accuracy and scalability. Allegro represents a many-body potential using iterated tensor products of learned equivariant representations without atom-centered message passing. Allegro obtains improvements over state-of-the-art methods on QM9 and revMD17. A single tensor product layer outperforms existing deep MPNNs and transformers on QM9. Furthermore, Allegro displays remarkable generalization to out-of-distribution data. Molecular simulations using Allegro recover structural and kinetic properties of an amorphous electrolyte in excellent agreement with ab-initio simulations. Finally, we demonstrate parallelization with a simulation of 100 million atoms.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2025

Dataset for Leveraging CryoEM and AI-Driven Morphological Feature Analysis for Insights on Bacterial Structures

This repository hosts an AI-assisted image segmentation and analysis pipeline for Pantoea sp. YR343 cryo-electron microscopy (cryoEM) datasets. The workflow automates membrane thickness measurements, flagella detection, and field-of-view (FOV) screening from low-dose, high-resolution cryoEM micrographs eliminating the need for slow manual annotation. By integrating deep-learning based segmentation (YOLOv11) with quantitative post-processing, this toolkit provides a scalable and reproducible way to study bacterial morphology under hydrated, near-native conditions. The GitHub repository for AI-based tools for cryoEM bacteria ultrastructures can be found here: https://github.com/Sireesiru/Cryo-EM-Ultrastructures/tree/main

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

On the Stochastic Stability of Deep Markov Models

Deep Markov models (DMM) are generative models which are scalable and expressive generalization of Markov models for representation, learning, and inference problems. However, the fundamental stochastic stability guarantees of such models have not been thoroughly investigated. In this paper, we present a novel stability analysis method and provide sufficient conditions of DMM's stochastic stability. The proposed stability analysis is based on the contraction of probabilistic maps modeled by deep neural networks. We make connections between the spectral properties of neural network's weights and different types of used activation function on the stability and overall dynamic behavior of DMMs with Gaussian distributions. Based on the theory, we propose a few practical methods for designing constrained DMMs with guaranteed stability. We empirically substantiate our theoretical results via intuitive numerical experiments using the proposed stability constraints.

Drgona, Jan↗
DOE OSTI2023

Using a physics-informed neural network and fault zone acoustic monitoring to predict lab earthquakes

Abstract Predicting failure in solids has broad applications including earthquake prediction which remains an unattainable goal. However, recent machine learning work shows that laboratory earthquakes can be predicted using micro-failure events and temporal evolution of fault zone elastic properties. Remarkably, these results come from purely data-driven models trained with large datasets. Such data are equivalent to centuries of fault motion rendering application to tectonic faulting unclear. In addition, the underlying physics of such predictions is poorly understood. Here, we address scalability using a novel Physics-Informed Neural Network (PINN). Our model encodes fault physics in the deep learning loss function using time-lapse ultrasonic data. PINN models outperform data-driven models and significantly improve transfer learning for small training datasets and conditions outside those used in training. Our work suggests that PINN offers a promising path for machine learning-based failure prediction and, ultimately for improving our understanding of earthquake physics and prediction.

42 ENGINEERING↗
DOE OSTI2019

Phase Segmentation in Atom-Probe Tomography Using Deep Learning-Based Edge Detection

Abstract Atom-probe tomography (APT) facilitates nano- and atomic-scale characterization and analysis of microstructural features. Specifically, APT is well suited to study the interfacial properties of granular or heterophase systems. Traditionally, the identification of the interface between, for precipitate and matrix phases, in APT data has been obtained either by extracting iso-concentration surfaces based on a user-supplied concentration value or by manually perturbing the concentration value until the iso-concentration surface qualitatively matches the interface. These approaches are subjective, not scalable, and may lead to inconsistencies due to local composition inhomogeneities. We introduce a digital image segmentation approach based on deep neural networks that transfer learned knowledge from natural images to automatically segment the data obtained from APT into different phases. This approach not only provides an efficient way to segment the data and extract interfacial properties but does so without the need for expensive interface labeling for training the segmentation model. We consider here a system with a precipitate phase in a matrix and with three different interface modalities—layered, isolated, and interconnected—that are obtained for different relative geometries of the precipitate phase. We demonstrate the accuracy of our segmentation approach through qualitative visualization of the interfaces, as well as through quantitative comparisons with proximity histograms obtained by using more traditional approaches.

47 OTHER INSTRUMENTATION↗
DOE OSTI2024

Astronomaly at scale: searching for anomalies amongst 4 million galaxies

ABSTRACT Modern astronomical surveys are producing data sets of unprecedented size and richness, increasing the potential for high-impact scientific discovery. This possibility, coupled with the challenge of exploring a large number of sources, has led to the development of novel machine-learning-based anomaly detection approaches, such as astronomaly. For the first time, we test the scalability of astronomaly by applying it to almost 4 million images of galaxies from the Dark Energy Camera Legacy Survey. We use a trained deep learning algorithm to learn useful representations of the images and pass these to the anomaly detection algorithm isolation forest, coupled with astronomaly’s active learning method, to discover interesting sources. We find that data selection criteria have a significant impact on the trade-off between finding rare sources such as strong lenses and introducing artefacts into the data set. We demonstrate that active learning is required to identify the most interesting sources and reduce artefacts, while anomaly detection methods alone are insufficient. Using astronomaly, we find 1635 anomalies among the top 2000 sources in the data set after applying active learning, including eight strong gravitational lens candidates, 1609 galaxy merger candidates, and 18 previously unidentified sources exhibiting highly unusual morphology. Our results show that by leveraging the human–machine interface, astronomaly is able to rapidly identify sources of scientific interest even in large data sets.

Astronomy & Astrophysics↗
DOE OSTI2024

MDLoader: A Hybrid Model-Driven Data Loader for Distributed Graph Neural Network Training

Scalable data management is essential for processing large scientific dataset on HPC platforms for distributed deep learning. In-memory distributed storage is preferred for its speed, enabling rapid, random, and frequent data access required by stochastic optimizers. Processes use one-sided or collective communication to fetch remote data, with optimal performance depending on (i) dataset characteristics, (ii) training scale, and (iii) interconnection network. Empirical analysis shows collective communication excels with larger mini-batch sizes and/or fewer processes, whereas one-sided communication outperforms at larger scales. We propose MDLoader, a hybrid in-memory data loader for distributed graph neural network training. MDLoader features a model-driven performance estimator that dynamically selects between one-sided and collective communication at the beginning of training using Tree of Parzen Estimators (TPE). Evaluations on NERSC Perlmutter and OLCF Summit show MDLoader outperforms single-backend loaders by up to 2.83 × and predicts the suitable communication method with 96.3% (Perlmutter) and 94.3% (Summit) success rate.

Bae, Jonghyun↗
DOE OSTI2023

DDStore: Distributed Data Store for Scalable Training of Graph Neural Networks on Large Atomistic Modeling Datasets

Graph neural networks (GNNs) are a class of Deep Learning models used in designing atomistic materials for effective screening of large chemical spaces. To ensure robust prediction, GNN models must be trained on large volumes of atomistic data on leadership class supercomputers. Even with the advent of modern architectures that consist of multiple storage layers that include node-local NVMe devices in addition to device memory for caching large datasets, extreme-scale model training faces I/O challenges at scale.We present DDStore, an in-memory distributed data store designed for GNN training on large-scale graph data. DDStore provides a hierarchical, distributed, data caching technique that combines data chunking, replication, low-latency random access, and high throughput communication. DDStore achieves near-linear scaling for training a GNN model using up to 1000 GPUs on the Summit and Perlmutter supercomputers, and reaches up to a 6.15x reduction in GNN training time compared to state-of-the-art methodologies.

Choi, Jong Youl↗
DOE OSTI2025

HAMscope: a snapshot Hyperspectral Autofluorescence Miniscope for real-time molecular imaging

We introduce HAMscope, a compact, snapshot hyperspectral autofluorescence miniscope that enables real-time, label-free molecular imaging in a wide range of biological systems. By integrating a thin polymer diffuser into a widefield miniscope, HAMscope spectrally encodes each frame and employs a probabilistic deep learning framework to reconstruct 30-channel hyperspectral stacks (452-703 nm) or directly infer molecular composition maps from single images. A scalable multi-pass U-Net architecture with transformer-based attention and per pixel uncertainty estimation enables high spatio-spectral fidelity (mean absolute error ∼0.0048) at video rates. While initially demonstrated in plant systems, including lignin, chlorophyll, and suberin imaging in intact poplar and cork tissues, the platform is readily adaptable to other applications such as neural activity mapping, metabolic profiling, and histopathology. We show that the system generalizes to out-of-distribution tissue types and supports direct molecular mapping without the need for spectral unmixing. HAMscope establishes a general framework for compact, uncertainty-aware spectral imaging that combines minimal optics with advanced deep learning, offering broad utility for real-time biochemical imaging across neuroscience, environmental monitoring, and biomedicine.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2019

A deep learning approach to anomaly detection in geological carbon sequestration sites using pressure measurements

Carbon capture and storage (CCS) has been extensively investigated as a potential engineering measure to reduce anthropogenic carbon emission to the atmosphere. Real-time monitoring of the safety and integrity of carbon storage reservoirs is a critical aspect of any commercial-scale CCS deployment. Pressure-based sensing is cost effective, suitable for real-time monitoring, and scalable to large monitoring networks. However, questions remain on how to best harness intelligent information from the high-frequency pressure monitoring sensors to support real-time decisions. Here this work presents a deep-learning-based framework for analyzing and detecting anomalies in pressure data streams by using a convolutional long short-term memory (ConvLSTM) neural network model, which allows for the fusion of both static and dynamic reservoir data. In ConvLSTM, the convolutional neural network (CNN) is used for spatial pattern mining and the LSTM is used for temporal pattern recognition. The performance of the ConvLSTM model for real-time anomaly detection is demonstrated using a set of pressure monitoring data collected from Cranfield, Mississippi, an active enhanced-oil-recovery field. The anomaly detection model is trained using bottom-hole pressure data acquired from the base experiment (without leak event) and then tested on pressure data collected during a series of controlled CO 2 release experiments (with artificially created leak events). Results show that the ConvLSTM neural network model successfully detected anomalies in the pressure time series obtained from the controlled release experiments. Inclusion of static information into the model further improves the robustness of ConvLSTM.

13 HYDRO ENERGY↗
DOE OSTI2022

Resilience Enhancements through Deep Learning Yields

This report documents the Resilience Enhancements through Deep Learning Yields (REDLY) project, a three-year effort to improve electrical grid resilience by developing scalable methods for system operators to protect the grid against threats leading to interrupted service or physical damage. The computational complexity and uncertain nature of current real-world contingency analysis presents significant barriers to automated, real-time monitoring. While there has been a significant push to explore the use of accurate, high-performance machine learning (ML) model surrogates to address this gap, their reliability is unclear when deployed in high-consequence applications such as power grid systems. Contemporary optimization techniques used to validate surrogate performance can exploit ML model prediction errors, which necessitates the verification of worst-case performance for the models.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2016

LLNL Partners with IBM on Brain-Like Computing Chip

Lawrence Livermore National Laboratory (LLNL) will receive a first-of-a-kind brain-inspired supercomputing platform for deep learning developed by IBM Research. Based on a breakthrough neurosynaptic computer chip called IBM TrueNorth, the scalable platform will process the equivalent of 16 million neurons and 4 billion synapses and consume the energy equivalent of a hearing aid battery – a mere 2.5 watts of power. The brain-like, neural network design of the IBM Neuromorphic System is able to infer complex cognitive tasks such as pattern recognition and integrated sensory processing far more efficiently than conventional chips.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Solving differential equations with unknown constitutive relations as recurrent neural networks

We solve a system of ordinary differential equations with an unknown functional form of a sink (reaction rate) term. We assume that the measurements (time series) of state variables are partially available, and use a recurrent neural network to “learn” the reaction rate from this data. This is achieved by including discretized ordinary differential equations as part of a recurrent neural network training problem. We extend TensorFlow’s recurrent neural network architecture to create a simple but scalable and effective solver for the unknown functions, and apply it to a fedbatch bioreactor simulation problem. Use of techniques from recent deep learning literature enables training of functions with behavior manifesting over thousands of time steps. Our networks are structurally similar to recurrent neural networks, but differ in purpose, and require modified training strategies.

deep learning↗
DOE OSTI2020

Concurrent Relaxation through Accelerated Deep Learning

CRADL captures performance metrics of machine learning algorithms operating on mesh data from multiphysics codes This proxy application is a tool to explore scalability of inference on HPC platforms, and also gather performance metrics for inference on new machine learning specific hardware. CRADL is designed to give users as fine a control as possible over an inference simulation. Users may select the number of cycles, amount of data, and batch size to pass to the accelerator of choice. Additionally the user may select a number of performance optimization libraries and flags. CRADL comes packaged with a repository of anonymized multi-physics simulation data, as well as a pretrained model for inference. The code allows a user to load their own pre-trained model and data if they wish. The code can operate in multiple parallelization schemes, with performance enhancing options such as half-precision libraries, PyTorch benchmarking, and pinned memory with non-blocking data transfers.

Zieb, KristoferJ.↗
DOE OSTI2022

PWR loading pattern optimization with reinforcement learning

The core loading pattern optimization problem belongs to the class of combinatorial optimization problem and has been studied since the dawn of commercial nuclear energy industry. It is characterized by multiple objectives and constraints, with a very high number of candidate patterns, which makes it impossible to solve explicitly. Stochastic optimization methodologies including Genetic Algorithms and Simulated Annealing are used by different nuclear utilities and vendors to perform fuel cycle reload design. Nevertheless, hand-designed solutions continue to be the prevalent method in the industry. To improve the state-of-the-art core reload patterns, we aim to create a method as scalable as possible, that agrees with the designer's goal of performance and safety. To help in this task Deep Reinforcement Learning (DRL), in particular Proximal Policy Optimization is leveraged. DRL has recently experienced a strong impetus from its successes applied to games, sometimes even reaching 'super-human' performances. This paper lays out the foundation of this method and proposes to study the behavior of several hyper-parameters that influence the DRL algorithm. The algorithm is highly dependent on multiple factors such as an exploration/exploitation trade-off that manifests through different parameters such as the number of loading patterns seen and the number of samples collected before a policy update, but also the shape of the objective function derived for the core design. Experimental results also demonstrate the effectiveness of the method in finding high-quality solutions from scratch within a reasonable amount of time. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2016

An analysis of image storage systems for scalable training of deep neural networks

This study presents a principled empirical evaluation of image storage systems for training deep neural networks. We employ the Caffe deep learning framework to train neural network models for three different data sets, MNIST, CIFAR-10, and ImageNet. While training the models, we evaluate five different options to retrieve training image data: (1) PNG-formatted image files on local file system; (2) pushing pixel arrays from image files into a single HDF5 file on local file system; (3) in-memory arrays to hold the pixel arrays in Python and C++; (4) loading the training data into LevelDB, a log-structured merge tree based key-value storage; and (5) loading the training data into LMDB, a B+tree based key-value storage. The experimental results quantitatively highlight the disadvantage of using normal image files on local file systems to train deep neural networks and demonstrate reliable performance with key-value storage based storage systems. When training a model on the ImageNet dataset, the image file option was more than 17 times slower than the key-value storage option. Along with measurements on training time, this study provides in-depth analysis on the cause of performance advantages/disadvantages of each back-end to train deep neural networks. We envision the provided measurements and analysis will shed light on the optimal way to architect systems for training neural networks in a scalable manner.

Lim, Seung-Hwan↗
DOE OSTI2016

The Livermore Brain: Massive Deep Learning Networks Enabled by High Performance Computing

The proliferation of inexpensive sensor technologies like the ubiquitous digital image sensors has resulted in the collection and sharing of vast amounts of unsorted and unexploited raw data. Companies and governments who are able to collect and make sense of large datasets to help them make better decisions more rapidly will have a competitive advantage in the information era. Machine Learning technologies play a critical role for automating the data understanding process; however, to be maximally effective, useful intermediate representations of the data are required. These representations or “features” are transformations of the raw data into a form where patterns are more easily recognized. Recent breakthroughs in Deep Learning have made it possible to learn these features from large amounts of labeled data. The focus of this project is to develop and extend Deep Learning algorithms for learning features from vast amounts of unlabeled data and to develop the HPC neural network training platform to support the training of massive network models. This LDRD project succeeded in developing new unsupervised feature learning algorithms for images and video and created a scalable neural network training toolkit for HPC. Additionally, this LDRD helped create the world’s largest freely-available image and video dataset supporting open multimedia research and used this dataset for training our deep neural networks. This research helped LLNL capture several work-for-others (WFO) projects, attract new talent, and establish collaborations with leading academic and commercial partners. Finally, this project demonstrated the successful training of the largest unsupervised image neural network using HPC resources and helped establish LLNL leadership at the intersection of Machine Learning and HPC research.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Fault-Tolerant Deep Learning Cache with Hash Ring for Load Balancing in HPC Systems

Large-scale DL on HPC systems like Frontier and Summit uses distributed node-local caching to address scalability and performance challenges. However, as these systems grow more complex, the risk of node failures increases, and current caching approaches lack fault tolerance, jeopardizing large-scale training jobs. We analyzed six months of SLURM job logs from Frontier and found that over 30% of jobs failed after an average of 75 minutes. To address this, we propose fault-tolerance strategies that recache data lost from failed nodes using a hash ring technique for balanced data recaching in the distributed node-local caching, reducing reliance on the PFS. Our extensive evaluations on Frontier showed that the hash ring-based recaching approach reduced training time by approximately 25% compared to the approach that redirects I/O to the PFS after node failures and demonstrated effective load balancing of training data across nodes.

Lee, Seoyeong↗