Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Batch”

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 1,117 records · Page 62
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Solution scattering at the Life Science X-ray Scattering (LiX) beamline

This work reports the instrumentation and software implementation at the Life Science X-ray Scattering (LiX) beamline at NSLS-II in support of biomolecular solution scattering. For automated static measurements, samples are stored in PCR tubes and grouped in 18-position sample holders. Unattended operations are enabled using a six-axis robot that exchanges sample holders between a storage box and a sample handler, transporting samples from the PCR tubes to the X-ray beam for scattering measurements. The storage box has a capacity of 20 sample holders. At full capacity, the measurements on all samples last for ∼9 h. For in-line size-exclusion chromatography, the beamline-control software coordinates with a commercial high-performance liquid chromatography (HPLC) system to measure multiple samples in batch mode. The beamline can switch between static and HPLC measurements instantaneously. In all measurements, the scattering data span a wide q -range of typically 0.006–3.2 Å −1 . Functionalities in the Python package py4xs have been developed to support automated data processing, including azimuthal averaging, merging data from multiple detectors, buffer scattering subtraction, data storage in HDF5 format and exporting the final data in a three-column text format that is acceptable by most data analysis tools. These functionalities have been integrated into graphical user interfaces that run in Jupyter notebooks, with hooks for external data analysis software.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

Advanced Health Information Technology Analytic Framework and Application to Hazard Detection

Health Information Technology (HIT) aims to improve healthcare outcomes by organizing and analyzing various health-related data. With data accumulating at a staggering rate, the importance of real-time analytics has been increasing dramatically, shifting the focus of informatics from batch processing to streaming analytics. HIT is also facing unprecedented challenges in adapting to this new requirement and leveraging advanced IT technologies. This paper introduces a HIT data and compute platform that supports multi-granularity real-time analytics from heterogeneous data sources. The paper first identifies functional requirements and proposes a framework that satisfies the requirements using state-of-the-art big data technologies including Apache Kafka, Spark Structured Streaming Engine, and Delta Lake. To demonstrate its capability to support data analytics in multiple time granularities analytics, a statistical process control-based hazard detection algorithm has been implemented on top of the framework to detect unexpected hazards from order cancellation data of the Department of US Veterans Affairs (VA) in near real-time.

Kumar, Mohit↗
DOE OSTI2022

Real-time Multi-granular Analytics Framework for HIT Systems

Streaming analytics is the process of ingesting and digesting live data from multiple data sources. In the healthcare domain, as the importance of extracting immediate insights while data are streaming into the system grows, the focus is shifting from batch processing to streaming analytics. With data increasing dramatically at high speeds, many informatics designs have been proposed to adapt healthcare domain into this new environment. In our previous work, we introduced a prototype of health informatics technology (HIT) framework that aims to address challenges in adopting state-of-the-art technologies to enable advanced healthcare analytic tasks in new streaming environments. We recently made major updates to the framework so that anomaly from multiple streaming data sources at different granularity levels can be detected in near real-time. In this paper, we detail the implementation and deployment of the framework in Kubernetes clusters and report its performances when tested on electronic health record (EHR) data of Veterans Affairs.

Park, Byung↗
DOE OSTI2024

An Active Learning-Based Streaming Pipeline for Reduced Data Training of Structure Finding Models in Neutron Diffractometry

Structure determination workloads in neutron diffractometry are computationally expensive and routinely require several hours to many days to determine the structure of a material from its neutron diffraction patterns. The potential for machine learning models trained on simulated neutron scattering patterns to significantly speed up these tasks have been reported recently. However, the amount of simulated data needed to train these models grows exponentially with the number of structural parameters to be predicted and poses a significant computational challenge. To overcome this challenge, we introduce a novel batch-mode active learning (AL) policy that uses uncertainty sampling to simulate training data drawn from a probability distribution that prefers labelled examples about which the model is least certain. We confirm its efficacy in training the same models with ∼ 75% less training data while improving the accuracy. We then discuss the design of an efficient stream-based training workflow that uses this AL policy and present a performance study on two heterogeneous platforms to demonstrate that, compared with a conventional training workflow, the streaming workflow delivers ∼ 20% shorter training time without any loss of accuracy.

Wang, Tianle [Brookhaven National Laboratory (BNL)↗
DOE OSTI2020

MARBLE: A Multi-GPU Aware Job Scheduler for Deep Learning on HPC Systems

Deep learning (DL) has become a key tool for solving complex scientific problems. However, managing the multi-dimensional large-scale data associated with DL, especially atop extant multiple graphics processing units (GPUs) in modern supercomputers poses significant challenges. Moreover, the latest high-performance computing (HPC) architectures bring different performance trends in training throughput compared to the existing studies. Existing DL optimizations such as larger batch size and GPU locality-aware scheduling have little effect on improving DL training throughput performance due to fast CPU-to-GPU connections. Additionally, DL training on multiple GPUs scales sublinearly. Thus, simply adding more GPUs to a system is ineffective. To this end, we design MARBLE, a first-of-its-kind job scheduler, which considers the non-linear scalability of GPUs at the intra-node level to schedule an appropriate number of GPUs per node for a job. By sharing the GPU resources on a node with multiple DL jobs, MARBLE avoids low GPU utilization in current multi-GPU DL training on HPC systems. Our comprehensive evaluation in the Summit supercomputer shows that MARBLE is able to improve DL training performance by up to 48.3% compared to the popular Platform Load Sharing Facility (LSF) scheduler. Compared to the state-of-the-art of DL scheduler, Optimus, MARBLE reduces the job completion time by up to 47%.

Han, Jingoo↗
DOE OSTI2022

Towards Low-Overhead Resilience for Data Parallel Deep Learning

Data parallel techniques have been widely adopted both in academia and industry as a tool to enable scalable training of deep learning models. At scale, DL training jobs can fail due to software or hardware bugs, may need to be preempted or terminated due to unexpected events, or may perform suboptimally because they were misconfigured. Under such circumstances, there is a need to recover and/or reconfigure data-parallel DL training jobs on-the-fly, while minimizing the impact on the accuracy of the DNN model and the runtime overhead. In this regard, state-of-art techniques adopted by the HPC community mostly rely on checkpoint-restart, which inevitably leads to loss of progress, thus increasing the runtime overhead. In this paper we explore alternative techniques that exploit the properties of modern deep learning frameworks (overlapping of gradient averaging and weight updates with local gradient computations through pipeline parallelism) to reduce the overhead of resilience/elasticity. To this end we introduce a failure simulation framework and two resilience strategies (immediate mini-batch rollback and lossy forward recovery), which we study compared with checkpoint-restart approaches in a variety of settings in order to understand the trade-offs between the accuracy loss of the DNN model and the runtime overhead.

data-parallel training↗
DOE OSTI2025

Reducing Communication Overhead in Federated Learning for Network Anomaly Detection with Adaptive Client Selection

Communication overhead in federated learning (FL) poses a significant challenge for network anomaly detection systems, where the myriad of client configurations and network conditions can severely impact system efficiency and detection accuracy. While existing approaches attempt to address this through individual optimization techniques, they often fail to maintain the delicate balance between reduced overhead and detection performance. This paper presents an adaptive FL framework that dynamically combines batch size optimization, client selection, and asynchronous updates to achieve efficient anomaly detection. Through extensive profiling and experimental analysis on two distinct datasets-UNSW-NBIS for general network traffic and ROAD for automotive networks-our framework reduces communication overhead by 97.6%; (from 700.0s to 16.8s) compared to synchronous baseline approaches while maintaining comparable detection accuracy (95.10%; vs. 95.12%;). Statistical validation using Mann-Whitney U test confirms significant improvements (p < 0.05) over existing FL approaches across both datasets, demonstrating the framework's adaptability to different network security contexts. Detailed profiling analysis reveals the efficiency gains through dramatic reductions in GPU operations and memory transfers while maintaining robust detection performance under varying client conditions.

Marfo, William [University of Texas at El Paso]↗
DOE OSTI2021

Virtual Log-Structured Storage for High-Performance Streaming

Over the past decade, given the higher number of data sources (e.g., Cloud applications, Internet of things) and critical business demands, Big Data transitioned from batch-oriented to real-time analytics. Stream storage systems, such as Apache Kafka, are well known for their increasing role in real-time Big Data analytics. For scalable stream data ingestion and processing, they logically split a data stream topic into multiple partitions. Stream storage systems keep multiple data stream copies to protect against data loss while implementing a stream partition as a replicated log. This architectural choice enables simplified development while trading cluster size with performance and the number of streams optimally managed. This paper introduces a shared virtual log-structured storage approach for improving the cluster throughput when multiple producers and consumers write and consume in parallel data streams. Stream partitions are associated with shared replicated virtual logs transparently to the user, effectively separating the implementation of stream partitioning (and data ordering) from data replication (and durability). We implement the virtual log technique in the KerA stream storage system. When comparing with Apache Kafka, KerA improves the cluster ingestion throughput by up to 4x when multiple producers write over hundreds of data streams.

consistent stream ordering↗
DOE OSTI2026

FitCache: A Transparent Drop-In Framework for Multi-Tier Caching to Accelerate Distributed Deep Learning Workloads

Training in Deep learning (DL) remains highly compute- and data-intensive, with I/O becoming a critical bottleneck as models and datasets scale. Recent studies report that data loading can dominate training time, especially on large-scale HPC systems with shared parallel file systems (PFS). Existing caching approaches either rely on single-tier designs or require intrusive modifications to training pipelines, limiting their portability and effectiveness. In this work, we present FitCache, a transparent drop-in framework for multi-tier caching to accelerate distributed DL training by coordinating fast local memory (e.g., DRAM, Persistent Memory (PMem)) and NVMe as hierarchical caches atop PFS. Our design adapts to hardware diversity, i.e., if NVMe is missing, memory transparently acts as a caching tier, ensuring stable performance. FitCache transparently intercepts I/O requests and issues concurrent fetches across all tiers, returning data from the fastest responder without centralized metadata or static redirection paths. FitCache adapts to dynamic workloads and heterogeneous clusters while maintaining POSIX compatibility. Experiments on Frontier (2048 GPUs) and smaller research clusters show that FitCache reduces training time by up to 40% and per-batch I/O latency by up to 71.6% compared to Lustre Orion PFS, offering a drop-in solution for scalable DL training.

Hu, Guangxing [ORNL] (ORCID:0009000283203614)↗
DOE OSTI2022

Worldwide Physics-Based Lifetime Prediction of c-Si Modules Due to Solder-Bond Failure

Lifetime prediction of the fielded c-Si solar modules due to location-specific weather conditions has been an important topic of photovoltaic research and the economic viability of solar energy. Data analytic techniques such as the performance ratio method, Statistical clear sky model, and Suns-Vmp methods quantify the degradation from measured data of a solar farm, however, the nonlinear time-dependence and correlated degradations make it difficult to use the empirical degradation rates for ultimate lifetime projection. In this article, we propose a complementary physics-based model to predict the solder bond failure caused by mechanical stress associated with the variations of the temperature. Integrating the worldwide weather information from NASA/NSRDB databases, the model predicts the location-specific output-power degradation and the lifetime of a module due to solder bond failure. The model parameters are calibrated against qualification tests involving thermal cycling of specific batches of modules from a specific technology/manufacturer. The results may be summarized as: 1) Modules installed at higher latitudes show a longer lifetime due to reduced damage accumulation. 2) The reduction of temperature fluctuation close to large bodies of water, such as seashores, increases solder bond lifetime significantly. 3) Relatively speaking, modules installed close to the Tropic of Cancer/Capricorn (23.5 degrees North/South) suffer from a higher solder bond damage and have a shorter lifetime, suggesting a conservative design. This model should serve as a building block of a comprehensive reliability framework that can predict the lifetime of a module that experiences simultaneous and correlated degradation mechanisms involving yellowing, corrosion, and potential-induced degradation.

14 SOLAR ENERGY↗
DOE OSTI2024

Converged Computing: A Best of Both Worlds of High-Performance Computing and Cloud

Collaboration between Cloud and High Performance Computing (HPC) communities has accelerated in the last half decade. A common goal to run batch workloads combined with a desire for reproducibility, automation, and optimization has led to successful projects that range from container technologies to workload management and security. This span of current and future work defines a novel “Converged Computing” paradigm that aims to combine the best of both worlds, both from a technological and cultural standpoint. Furthermore, in this Special Issue, we review common themes in the space, showcasing current work and encouraging a continued effort toward innovative ideas that will enable the next generations of scientific discovery.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Adaptive Computing for Scale-Up Problems

Adaptive Computing is an application-agnostic outer loop framework to strategically deploy simulations and experiments to guide decision making for scale-up analysis. Resources are allocated over successive batches, which makes the allocation adaptive to some objective such as optimization or model training. The framework enables the characterization and management of uncertainties associated with predictive models of complex systems when scale-up questions lead to significant model extrapolation. A key advancement of this framework is its integration of multi-fidelity surrogate modeling, uncertainty management, and automated orchestration of various computing and experimentation resources into a single integrated software package. This enables efficient multi-fidelity modeling across multiple computing resources by incorporating real-world constraints such as relative queue times and throughput on individual machines into the multi-fidelity sampling decision. We discuss applications of this framework to problems in the renewable energy space, including biofuels production, material synthesis, perovskite crystal growth, and building electrical loads.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2022

Dynamic Parameter Estimation with Physics-based Neural Ordinary Differential Equations

Accurate estimation of dynamic parameters of gen-erators is crucial to building a reliable model for dynamical studies and reliable operation of the power system. This paper develops a physics-based neural ordinary differential equations (ODE) approach to learn the parameters of generator dynamic model using phasor measurement units (PMU) data. We design a physics-based neural network to represent the swing equations of the power system dynamics. A loss function is defined as the difference between dynamic simulation results from the physics-based neural networks and pseudo PMU measurements. The parameters of generator dynamic model are iteratively updated using the neural ODEs and the adjoint method. By exploiting the mini-batch scheme in neural ODE training, the parameter estimation performance is significantly improved. Numerical study results on a 3-machine 9-bus system show that the proposed algorithm outperforms state-of-the-art baseline method in both computation time and dynamic parameter estimation accuracy.

24 POWER TRANSMISSION AND DISTRIBUTION↗
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 OSTI2022

Online State Estimation for Time-Varying Systems

The paper investigates the problem of estimating the state of a time-varying system with a linear measurement model; in particular, the paper considers the case where the number of measurements available can be smaller than the number of states. In lieu of a batch linear least-squares (LS) approach well-suited for static networks, where a sufficient number of measurements could be collected to obtain a full-rank design matrix the paper proposes an online algorithm to estimate the possibly time-varying state by processing measurements as and when available. The design of the algorithm hinges on a generalized LS cost augmented with a proximal-point-type regularization. With the solution of the regularized LS problem available in closed-form, the online algorithm is written as a linear dynamical system where the state is updated based on the previous estimate and based on the new available measurements. Conditions under which the algorithmic steps are in fact a contractive mapping are shown, and bounds on the estimation error are derived for different noise models. Numerical simulations are provided to corroborate the analytical findings.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2021

Scale Up of High-Performance REBCO Tapes in a Pilot-Scale Advanced MOCVD Tool With In-Line 2D-XRD System

We have reported the development of an Advanced Metal-Organic Chemical Vapor Deposition (AMOCVD) method featuring ohmic heating of the substrate for REBCO film growth, direct tape temperature monitoring, and laminar precursor flow. This A-MOCVD method has been used to fabricate 4 – 5 m thick film REBCO tapes with record high performance: critical currents exceeding 8700 A/12 mm at 30 K, 3 T; and engineering current density of 5200 A/mm 2 at 4.2 K, 15 T. Recently, we constructed a pilot-scale reel-to-reel A-MOCVD system to scale up the technology to long tapes. Preliminary batches of tapes exhibit consistent performance and good uniformity along the length. An in-line 2D X-ray Diffraction (XRD) system has been integrated into the pilot A-MOCVD tool to monitor the REBCO texture and composition, RE 2 O 3 content, and the dimensions of BaMO 3 (BMO, M=Zr,Hf) nanorods that act as artificial pinning centers. Specifically, the streaking angle between REBCO (103) and BZO (101) has been found to correlate well with (Ba+M)/Cu of the film, lift factor in critical current over a range of temperatures and magnetic fields, and the size of the BMO nanorods. Furthermore, the in-line 2D-XRD is expected to serve as a valuable quality measuring tool supporting realtime feedback control for the process to yield uniform and consistent manufacturing of high performance REBCO tapes by Advanced MOCVD.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2024

Fast Neutron Irradiation of a Multichannel JFET-Based Optical Encoder

Modern electrical components are susceptible to damage from high levels of radiation and extreme temperatures found near reactors in terrestrial nuclear power plants and in aerospace applications. Radiation-hardened electronics are being developed, largely for the aerospace industry, but they sometimes rely on application-specific, small-batch semiconductor fabrication processes. These processes tend to be prohibitively expensive to develop and maintain outside major industrial facilities or governmental agencies. Recently, commercially available, nonradiation-rated junction-gate field-effect transistors (JFETs) were shown to maintain their functionality at gamma doses exceeding 1 MGy, suggesting that nonrated, commercially available electrical components could be used to develop systems that are tolerant to ionizing radiation. However, gamma ray survival is not indicative of neutron dose survival, and few studies characterize JFETs under neutron irradiation. To address this knowledge gap, a JFET-based analog multiplexer and optical pulsewidth modulation (PWM) encoder was developed and irradiated using a 252 Cf source to 1.6×10 13 n/cm 2 . The multiplexed optical encoder (MOE) system maintained functionality throughout testing and showed little evidence of radiation effects. These results indicate that circuitry tolerant to fast neutron damage can be developed using low-cost, nonradiation-rated, commercially available JFETs, which could provide a lower production cost alternative to specialized semiconductor processes when designing and building electronics better able to survive neutron irradiation.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2020

The Case for Strong Scaling in Deep Learning: Training Large 3D CNNs with Hybrid Parallelism

Here, we present scalable hybrid-parallel algorithms for training large-scale 3D convolutional neural networks. Deep learning-based emerging scientific workflows often require model training with large, high-dimensional samples, which can make training much more costly and even infeasible due to excessive memory usage. We solve these challenges by extensively applying hybrid parallelism throughout the end-to-end training pipeline, including both computations and I/O. Our hybrid-parallel algorithm extends the standard data parallelism with spatial parallelism, which partitions a single sample in the spatial domain, realizing strong scaling beyond the mini-batch dimension with a larger aggregated memory capacity. We evaluate our proposed training algorithms with two challenging 3D CNNs, CosmoFlow and 3D U-Net. Our comprehensive performance studies show that good weak and strong scaling can be achieved for both networks using up to 2K GPUs. More importantly, we enable training of CosmoFlow with much larger samples than previously possible, realizing an order-of-magnitude improvement in prediction accuracy.

97 MATHEMATICS AND COMPUTING↗