Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Performance Modeling”

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 973 records · Page 54
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Utilizing ensemble learning for performance and power modeling and improvement of parallel cancer deep learning CANDLE benchmarks

Abstract Machine learning (ML) continues to grow in importance across nearly all domains in modeling to learn from data. Often a tradeoff exists between a model's ability to minimize bias and variance. In this article, we utilize ensemble learning to combine linear, nonlinear, and tree‐/rule‐based ML methods to cope with the bias‐variance tradeoff and result in more accurate models. We use the datasets collected for two parallel cancer deep learning CANDLE benchmarks, NT3 and P1B2, to build performance and power models based on hardware performance counters using single‐object and multiple‐objects ensemble learning to identify the most important counters for improvement on the Cray XC40 Theta at Argonne National Laboratory. Based on the insights from these models, we improve the performance and energy of P1B2 and NT3 by optimizing the deep learning environments TensorFlow, Keras, Horovod, and Python under the huge page size of 8 MB. Experimental results show that ensemble learning not only produces more accurate models but also provides more robust performance counter ranking. We achieve up to 61.15% performance improvement and up to 62.58% energy saving for P1B2 and up to 55.81% performance improvement and up to 52.60% energy saving for NT3 on up to 24,576 cores.

Wu, Xingfu↗
DOE OSTI2020

Verification of Reactor Activation Modelling with Physical Characterization Data and the Impact on Long Term Assessments - 20303

Whiteshell Laboratories, located in Manitoba, Canada, provided research facilities for the Canadian nuclear industry since the early 1960's. The facility was centered on Whiteshell Reactor 1 (WR-1), an organically cooled, heavy water moderated nuclear reactor. WR-1 was safely shut down in 1985, defueled and drained, and has been in a safe Storage with Surveillance (Deferred Decommissioning) state until present day. In support of an Environmental Assessment for proposed in-situ disposal of WR-1, detailed characterization information regarding the reactor core components (calandria, calandria tubes, and pressure tubes) was necessary to provide confidence that the long-term impacts of the facility can be assessed adequately. A simplified estimate of the total activation of core components was completed in 1992 using conservative assumptions, expected reactor configuration and computer models built with WIMS-CRNL, ONEDANT and ORIGEN-S. Due to the simplification, and age of the work, there was considerable doubt that the modelled inventory estimate of the core components was sufficient to support long-term performance modeling of the planned in-situ decommissioning. Additional data would be required to provide this confidence. Several options were considered for characterizing the WR-1 core components, but it was decided to collect only a limited number of samples in an effort to validate the 1992 simplified estimate as bounding. The limited number of samples reduced risks to workers, both radiological and non-radiological, and significantly reduced the costs required to confidently define and bound the radiological inventory of activation products in the core. The results of the sampling were compared to the simplified activation estimation. There was good agreement between the modelled and sampled results, with all major contributors to total activity showing similar relative quantities. The samples were generally lower than predicted by the models, with few exceptions. Notable exceptions included Nb-94, which were higher in the sample results than the model. This was determined to be caused by the use of Niobium containing steel alloys in the stainless steel pressure tube that were not accounted for in the models. The comparison of results provided confidence that the original inventory estimates produced through the simplified model are conservative and bounding, and therefore suitable for use in the long-term assessment of in situ disposal of the WR-1 reactor. This paper provides details of the simplified computer model results, their comparison to results of the sampling, and the changes adopted in the assessment of the in-situ decommissioning of WR-1 as a result. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2025

Leveraging Large Language Models for Real-World Data Evidence: A Framework for Automated Treatment Extraction and Data Harmonization

Background: The ability to comprehensively collect treatment information from cancer patient medical records would enable studies to evaluate real-world benefits and risks tied to specific treatments. Currently, it is difficult to system- atically collect high-quality treatment information because it is often stored in unstructured text. Manually extracting and standardizing drug and regimen data is time-intensive. Recent advances in large language models (LLMs) offer a potential solution for automated extraction of structured treatment information from clinical text. Objective: This study systematically evaluates the utility of four LLMs from the Llama family for automated extraction of oncology treatment information from clinical text. This information can guide researchers using cancer registry data to provide insights into cancer care and outcomes beyond clinical trials. Methods: Four instruction-tuned Llama models with varying parameter counts (1B, 3B, 8B, and 70B) were evaluated for their ability to extract treatment information from clinical documents. A unified oncology knowledge base integrating seven major public data sources was developed to standardize and normalize extracted entities—a critical step for harmonizing data from diverse sources. Extracted treatment data were compared against expert-annotated ground truth. Model performance was assessed using accuracy metrics (Precision, Recall, F1-Score) and opera- tional feasibility metrics, including processing speed and structural compliance of the output. Results: A strong positive correlation was observed between model size and extraction accuracy. F1-score improved from 0.609 for the 1B model to 0.710 (3B), 0.807 (8B), and 0.828 (70B). While larger models demonstrated superior accuracy and compliance, they incurred higher computational costs. The modest performance difference between 8B and 70B suggests diminishing returns with increasing model size. Conclusions: LLMs represent a viable technology for automating oncology treatment extraction. The 8B-parameter model emerged as a highly effective option, balancing high accuracy and computational efficiency. Selecting an appropriate LLM for deployment in cancer registries involves a trade-off between desired accuracy and available operational resources. Harmonizing extracted entities with the oncology knowledge base facilitates standardized integration into common data models, enhancing data quality for real-world evidence analyses.

artificial intelligence↗
DOE OSTI2020

An Energy Calculator for Simple Commercial Buildings

According to the EIA, simple commercial buildings account for 97% of total commercial building stock. However, most simple commercial buildings for example small- to mid-sized offices, retail, schools and warehouses do not benefit from the data-driven decision-making capabilities of whole-building energy modeling. The high cost of custom modeling limits the use of energy modeling of simple buildings for new construction or retrofit measures. Lack of tools providing helpful information on interactive savings estimates creates difficulties in meeting aggressive decarbonization and energy efficiency goals for simple building designers and utility program managers. This paper reviews a beta phase Simple Building Calculator with the ability to generate relatively accurate and interactive modeling results based on a limited but robust set of inputs. It can evaluate whole-building or single measure savings in new or existing buildings, compare measure package choices, or provide simplified performance modeling for energy codes and utility incentives. The tool combines physical (annual whole building prototype simulation) and statistical modeling techniques to predict annual energy performance. It supports a variety of building characteristics for envelope, HVAC, and lighting with parameters ranging from vintage to max tech configurations, as well as support for single-zone and simple multi-zone HVAC systems. The Simple Building Calculator was designed to provide immediate feedback for otherwise computationally intensive tasks like measure comparison, development of multiple measure package combinations, or verification that measures meet efficiency targets—all with the goal of providing a tool for quick annual energy simulation of simple commercial buildings.

Hart, Reid↗
DOE OSTI2023

Classifying metal‐binding sites with neural networks

Abstract To advance our ability to predict impacts of the protein scaffold on catalysis, robust classification schemes to define features of proteins that will influence reactivity are needed. One of these features is a protein's metal‐binding ability, as metals are critical to catalytic conversion by metalloenzymes. As a step toward realizing this goal, we used convolutional neural networks (CNNs) to enable the classification of a metal cofactor binding pocket within a protein scaffold. CNNs enable images to be classified based on multiple levels of detail in the image, from edges and corners to entire objects, and can provide rapid classification. First, six CNN models were fine‐tuned to classify the 20 standard amino acids to choose a performant model for amino acid classification. This model was then trained in two parallel efforts: to classify a 2D image of the environment within a given radius of the central metal binding site, either an Fe ion or a [2Fe‐2S] cofactor, with the metal visible (effort 1) or the metal hidden (effort 2). We further used two sub‐classifications of the [2Fe‐2S] cofactor: (1) a standard [2Fe‐2S] cofactor and (2) a Rieske [2Fe‐2S] cofactor. The accuracy for the model correctly identifying all three defined features was >95%, despite our perception of the increased challenge of the metalloenzyme identification. This demonstrates that machine learning methodology to classify and distinguish similar metal‐binding sites, even in the absence of a visible cofactor, is indeed possible and offers an additional tool for metal‐binding site identification in proteins.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2026

Interpretable Models for Workflow Differentiation in High-Performance Scientific Networks

Scientific workflows in high-performance networks spawn hundreds of interdependent flows that must be managed collectively—yet existing network classifiers treat each flow in isolation, leading to fragmented QoS decisions and missed interflow patterns. We present a novel traffic classification solution that operates at the workflow level, distinguishing entire filetransfer operations from streaming analytics by capturing how concurrent flows interact and burst together. We introduce a workflow identification window (WIW) that ingests raw packet headers from parallel flows into unified tensors, preserving the spatial-temporal patterns that differentiate scientific workflows. This approach achieves 98.7% accuracy using CNN, LSTM, and hybrid architectures, while maintaining 84% accuracy on production traffic collected a week later—demonstrating robustness to temporal drift. By integrating SHAP and GradCAM explainability, we reveal that early-packet timing patterns and cross-flow correlations drive classification decisions, providing operators with interpretable insights. Our system enables coherent workflow-level QoS enforcement and dynamic bandwidth allocation in scientific networks, eliminating manual per-flow configuration while maintaining classification latency at millisecond level.

Giannakou, Anna [LBL, Berkeley]↗
DOE OSTI2021

Understanding HPC Benchmark Performance on Intel Broadwell and Cascade Lake Processors

Hardware platforms in high performance computing are constantly getting more complex to handle even when considering multicore CPUs alone. Numerous features and configuration options in the hardware and the software environment that are relevant for performance are not even known to most application users or developers. Microbenchmarks, i.e., simple codes that fathom a particular aspect of the hardware, can help to shed light on such issues, but only if they are well understood and if the results can be reconciled with known facts or performance models. The insight gained from microbenchmarks may then be applied to real applications for performance analysis or optimization. In this paper we investigate two modern Intel x86 server CPU architectures in depth: Broadwell EP and Cascade Lake SP. We highlight relevant hardware configuration settings that can have a decisive impact on code performance and show how to properly measure on-chip and off-chip data transfer bandwidths. The new victim L3 cache of Cascade Lake and its advanced replacement policy receive due attention. Finally we use DGEMM, sparse matrix-vector multiplication, and the HPCG benchmark to make a connection to relevant application scenarios.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Evaluation of a preliminary regional Earth model through comparison of synthetic and observed waveform data

In this report, we document the process related to developing a regional geologic model of a 605 x 1334 km area centered around Utah and encompassing surrounding states. This model is developed to test the effect that composition of a model has on the generation of synthetic data with the intent of using this information to improve upon full waveform moment tensor inversions. We compare observed data from three seismic events and five stations to the synthetic data generated by a preliminary model derived from a geologic framework model (GFM) developed by the USGS. The synthetic data and observed data comparisons indicate that our preliminary model performs well at smaller offset distances in the northern and central sections of the model. However, the southern stations consistently display synthetic data P- and S-wave arrival times that do not match the observed data arrival times, indicating that the velocity structure of the southern part of the model especially is inaccurate.

58 GEOSCIENCES↗
DOE OSTI2023

Using Deep Learning to Develop a High Resolution Planetary Boundary Layer Model for Infrasound Propagation

Infrasound, with frequencies less than 20 Hz, is generated by both natural and anthropogenic sources. When one of these sources exerts a force on the atmosphere, infrasonic waves are generated. The propagation of these waves largely depends on temperature, wind speed, and wind direction. Previous work has used deep learning to accurately predict atmospheric specifications to altitudes of ~40 km. However, this model breaks down for local distances because it is too low resolution. Here we use a high-resolution meteorological dataset collected in Las Vegas, Nevada, USA to develop a deep learning model that can predict temperature, wind speed, and wind direction. Predictions are compared to ground truth observations to show that the model performs well at predicting temperature and wind direction but struggles with prediction wind speed. Model limitations and improvements are also discussed.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Leveraging Artificial Intelligence to Predict Novel Eutectic Alloys

The goal of this project was to train an artificial neural network (ANN) to predict the fractional composition and melting point of eutectic alloys using fundamental atomic properties as inputs. The fundamental properties considered include atomic number, atomic weight, atomic radius, valence electron concentration, electronegativity, and electron affinity. The project involved several phases, starting with data preparation, where phase diagram data was harvested from the ASM International database. Approximately 1300 binary eutectics were collected and cleaned to ensure relevance and accuracy. A regression model was selected for training, utilizing a rectified linear unit as the activation function. Various model configurations were evaluated for predictive accuracy, with validation techniques employed to ensure robustness. The model demonstrated predictive capabilities above random guessing and was able to achieve up to 11% accuracy under certain conditions. An ablative test identified atomic radius and valence electron concentration as critical inputs for model performance. Incorporating the melting point of atomic constituents improved accuracy significantly, although ultimately the model’s predictive capability still fell short of the 80% target. This report details the methodology, results, and implications of the research, contributing to the understanding of employing artificial intelligence to predict the phase transition behavior of eutectic alloys.

36 MATERIALS SCIENCE↗
DOE OSTI2025

SIDDA: SInkhorn Dynamic Domain Adaptation

Modern neural networks (NNs) often do not generalize well in the presence of a "covariate shift"; that is, in situations where the training and test data distributions differ, but the conditional distribution of classification labels remains unchanged. In such cases, NN generalization can be reduced to a problem of learning more domain-invariant features. Domain adaptation (DA) methods include a range of techniques aimed at achieving this; however, these methods have struggled with the need for extensive hyperparameter tuning, which then incurs significant computational costs. In this work, we introduce SIDDA, an out-of-the-box DA training algorithm built upon the Sinkhorn divergence, that can achieve effective domain alignment with minimal hyperparameter tuning and computational overhead. We demonstrate the efficacy of our method on multiple simulated and real datasets of varying complexity, including simple shapes, handwritten digits, and real astronomical observations. SIDDA is compatible with a variety of NN architectures, and it works particularly well in improving classification accuracy and model calibration when paired with equivariant neural networks (ENNs). We find that SIDDA enhances the generalization capabilities of NNs, achieving up to a ≈40% improvement in classification accuracy on unlabeled target data. We also study the efficacy of DA on ENNs with respect to the varying group orders of the dihedral group DN, and find that the model performance improves as the degree of equivariance increases. Finally, we find that SIDDA enhances model calibration on both source and target data--achieving over an order of magnitude improvement in the ECE and Brier score. SIDDA's versatility, combined with its automated approach to domain alignment, has the potential to advance multi-dataset studies by enabling the development of highly generalizable models.

Pandya, Sneh [Northeastern U.]↗
NASA NTRS1998

Evaluation of the 29-km Eta Model: Objective Verification at Three Selected Stations - Part 1

This paper describes an objective verification of the National Centers for Environmental Prediction (NCEP) 29-km eta model from May 1996 through January 1998. The evaluation was designed to assess the model's surface and upper-air point forecast accuracy at three selected locations during separate warm (May - August) and cool (October - January) season periods. In order to enhance sample sizes available for statistical calculations, the objective verification includes two consecutive warm and cool season periods. Systematic model deficiencies comprise the larger portion of the total error in most of the surface forecast variables that were evaluated. The error characteristics for both surface and upper-air forecasts vary widely by parameter, season, and station location. At upper levels, a few characteristic biases are identified. Overall however, the upper-level errors are more nonsystematic in nature and could be explained partly by observational measurement uncertainty. With a few exceptions, the upper-air results also indicate that 24-h model error growth is not statistically significant. In February and August 1997, NCEP implemented upgrades to the eta model's physical parameterizations that were designed to change some of the model's error characteristics near the surface. The results shown in this paper indicate that these upgrades led to identifiable and statistically significant changes in forecast accuracy for selected surface parameters. While some of the changes were expected, others were not consistent with the intent of the model updates and further emphasize the need for ongoing sensitivity studies and localized statistical verification efforts. Objective verification of point forecasts is a stringent measure of model performance, but when used alone, is not enough to quantify the overall value that model guidance may add to the forecast process. Therefore, results from a subjective verification of the meso-eta model over the Florida peninsula are discussed in the companion paper by Manobianco and Nutter. Overall verification results presented here and in part two should establish a reasonable benchmark from which model users and developers may pursue the ongoing eta model verification strategies in the future.

Nutter, Paul A.↗
NASA NTRS1998

HPF Implementation of NPB2.3

We present an HPF implementation of BT, SP, LU, FT, CG and MG of NP132.3-serial benchmark set. The implementation is based on HPF performance model of the benchmark specific primitive operations with distributed arrays. We present profiling and performance data on SGI Origin 2000 and compare the results with NPB2.3. We discuss an HPF limitation in organizing pipelined computations which results in memory and communication overhead in BT and SP.

Frumkin, Michael↗
NASA NTRS2005

An information maximization model of eye movements

We propose a sequential information maximization model as a general strategy for programming eye movements. The model reconstructs high-resolution visual information from a sequence of fixations, taking into account the fall-off in resolution from the fovea to the periphery. From this framework we get a simple rule for predicting fixation sequences: after each fixation, fixate next at the location that minimizes uncertainty (maximizes information) about the stimulus. By comparing our model performance to human eye movement data and to predictions from a saliency and random model, we demonstrate that our model is best at predicting fixation locations. Modeling additional biological constraints will improve the prediction of fixation sequences. Our results suggest that information maximization is a useful principle for programming eye movements.

NASA Discipline Neuroscience↗
NASA NTRS2010

Experimental Validation of FE/BEM Dynamic Strain Model Under Diffuse Acoustic Field Loading

Structural finite element (FE) models naturally output displacement or acceleration response data. However, they can also be used to compute stress, internal forces, and strain response. When coupled with a boundary element model (BEM) of the fluid surrounding the structure, a fully coupled analysis can be performed. Modeling a diffuse acoustic field in the BEM fluid provides an excitation like that found when the structure is placed in a reverberation chamber. Fully coupling the structural FE model to the acoustic BEM model provides a means to predict not only the acceleration response of the panel to diffuse field loading, but also the ability to predict the dynamic stress and strain response. This type of model has been available with current predictive tools, but experimental validation of the prediction of dynamic stress or strain is difficult to find. An aluminum panel was instrumented with accelerometers and strain gages and hung in a reverberation room and subjected to a diffuse acoustic field. This paper presents the comparison of the experimental and predicted results.

Tsoi, W. Ben↗