Search NASA⌕ Search

SEARCH · Search NASA

Results for “multi objective”

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 163 records · Page 9
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Optimizing pressurized-water reactor equilibrium cycle using a novel loading pattern encoding and rule-based genetic crossover operators

This work presents an extended multi-batch approach applied in shuffling scheme optimization for equilibrium cycle for pressurized water reactors using Genetic Algorithms (GAs). A new ruled based GA crossover operator called Inherited Location and Batch (ILB) was introduced to enhance offsprings reproduction efficiency specialized for equilibrium cycle optimization problem. This approach was implemented within the Plant ReLoad Optimization (PRLO) framework and validated using a generic reactor model based on the AP1000 design, with core parameters calculated via the CASMO/SIMULATE software package. The ILB approach is then applied for both single and multi-objective problems in maximizing cycle length and core average exposure while minimizing the average enrichment of the 57 fresh fuel assemblies (FAs) per cycle. The optimal solutions are selected based on their dominance to the objectives from all feasible solutions. This research identified three optimal solutions satisfied safety constraints: The first solution minimizes feed enrichment costs with a cycle length of 338.8 days and core exposure of 25.39 MWd/MT; the second solution extends cycle length to 361.2 days, with the highest core exposure of 26.84 MWd/MT, using 3.75 wt% average fuel enrichment; the third solution balances both objectives with a cycle length of 349.6 days, core exposure of 25.82 MWd/MT with a slight enrichment increase compared to the first solution. Collectively, these findings underscore the efficiency and effectiveness of the proposed approach in achieving practical multi-objective optimal equilibrium cycle designs using GAs optimizer.

11 - NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2025

Surrogate-driven design optimization with uncertainty constraints in Monte Carlo simulations

In multi-objective design tasks, the computational cost increases rapidly when high-fidelity simulations are used to evaluate objective functions. Surrogate models help mitigate this cost by approximating the simulation output, simplifying the design process. However, under high uncertainty, surrogate models trained on noisy data can produce inaccurate predictions, as their performance depends heavily on the quality of training data. This study investigates the impact of data uncertainty on two multi-objective design problems modelled using Monte Carlo transport simulations: a neutron moderator and an ion-to-neutron converter. For each, a grid search was performed using five different tally uncertainty levels to generate training data for neural network surrogate models. These models were then optimized using NSGA-III. The recovered Pareto-fronts were analyzed across uncertainty levels: in the moderator problem, normalized hypervolume dropped from 0.886 at 1.0% uncertainty to 0.748 at 10% uncertainty, while in the converter problem it remained near 0.50 for all cases. Average simulation times were also compared to evaluate the trade-off between accuracy and computational cost. Results show that the influence of simulation uncertainty is strongly problem-dependent. In the neutron moderator case, higher uncertainties led to exaggerated objective sensitivities and distorted Pareto-fronts, reducing normalized hypervolume. In contrast, the ion-to-neutron converter task was less affected—low-fidelity simulations produced results similar to those from high-fidelity data. These findings suggest that a fixed-fidelity approach is not optimal. Surrogate models can recover the Pareto-front under noisy conditions, and multi-fidelity studies help identify suitable uncertainty levels for each problem to balance efficiency and accuracy.

07 ISOTOPE AND RADIATION SOURCES↗
DOE OSTI2020

Design Paradigm for Modular Multilevel Converter-Based Generator Rectifier Systems

Modular Multilevel Converters (MMC) are being widely considered for medium to high voltage DC applications. Designing such converters through multi-objective optimization is of interest because such an approach allows the trade-off between competing objectives (for example mass and loss) to be explicitly and quantitatively identified. In this work, an optimization based design paradigm for MMC based generator rectifier systems is proposed. Such development typically requires detailed component design and simulation models for the electric machine and converter which are computationally expensive. As an alternative, the proposed work utilizes an electric machine metamodel, inductor metamodel, and high-speed steady-state simulation model for the MMC to facilitate multi-objective optimization minimizing system metrics of interest while satisfying system constraints. A case study is undertaken to demonstrate the functionality of the proposed design paradigm.

42 ENGINEERING↗
DOE OSTI2025

Economic Evaluation of Modernization Expenditures for Electric Utility Distribution Systems: A Guide for Utility Regulators

Cost-effectiveness evaluation of potential grid modernization solutions is integral to integrated distribution system planning (IDSP). This report synthesizes and updates prior cost-effectiveness approaches identified by the U.S. Department of Energy for grid modernization investments, incorporating multi-objective decision-making. The structure of the report follows the IDSP process, from identifying objectives and priorities to prioritizing grid solutions. The report details two methods for conducting initial cost-effectiveness screening of individual and interdependent grid components: (1) Lowest Reasonable Cost and (2) Benefit-Cost Analysis. Drawing on solutions that passed the initial cost-effectiveness screen, the utility prioritizes grid solutions based on policy objectives and priorities, regulatory compliance, operational efficiencies, and other factors to develop a portfolio of distribution solutions. The utility uses multi-objective decision-making to assess each proposed expenditure against each objective and respective metric. The utility applies a weighting factor, reflecting the priority ranking of the objective, to the total numerical “score” based on the contribution of the proposed solution to addressing each objective. Ultimately, the utility ranks each expenditure from highest to lowest priority using the final score for each solution as well as its cost. The report includes examples of emerging best practices for states and utilities and a cost-effectiveness process checklist.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2025

Hierarchical Gaussian process-based Bayesian optimization for materials discovery in high entropy alloy spaces

Bayesian optimization (BO) is a powerful and data-efficient method for iterative materials discovery and design, particularly valuable when prior knowledge is limited, underlying functional relationships are complex or unknown, and the cost of querying the materials space is significant. Traditional BO methodologies typically utilize conventional Gaussian Processes (cGPs) to model the relationships between material inputs and properties, as well as correlations within the input space. However, cGP-BO approaches often fall short in multi-objective optimization scenarios, where they are unable to fully exploit correlations between distinct material properties. Leveraging these correlations can significantly enhance the discovery process, as information about one property can inform and improve predictions about others. Here, this study addresses this limitation by employing advanced kernel structures to capture and model multi-dimensional property correlations through multi-task (MTGPs) or deep Gaussian Processes (DGPs), thus accelerating the discovery process. We demonstrate the effectiveness of MTGP-BO and DGP-BO in rapidly and robustly solving complex materials design challenges that occur within the context of complex multi-objective optimization over FCC FeCrNiCoCu high entropy alloy (HEA) spaces, where traditional cGP-BO approaches fail. Furthermore, we highlight how the differential costs associated with querying various material properties can be strategically leveraged to make the materials discovery process more cost-efficient.

36 MATERIALS SCIENCE↗
DOE OSTI2026

HARMONY: Large-Scale Architecture Search for Efficient Hybrid Language Models

As large language models scale to trillions of parameters, their computational and memory requirements present critical challenges for efficient training and deployment. While Mixture of Experts (MoE) architectures enable efficient scaling through sparse parameter activation, and state-space models like Mamba offer linear-time complexity, principled methods for combining these paradigms remain undeveloped. We introduce HARMONY (Hybrid Architecture Research for Mamba, Optimized with Neural efficiencY), a multi-objective evolutionary neural architecture search framework for discovering efficient hybrid language models that integrate Transformer attention mechanisms, Mixture-of-Experts routing, and Mamba state-space components. Through large-scale distributed search using 16,384 MI250X GPUs on the Frontier supercomputer, HARMONY explores a comprehensive design space encompassing six attention variants (MHA, MQA, GQA, MLA, SWA, and Mamba-2), variable MoE configurations with both routed and shared experts, and extensive Mamba hyperparameters. Our framework discovers heterogeneous architectures that balance training performance with computational efficiency through multi-objective optimization incorporating latency penalties and fitness-based selection. Analysis of discovered architectures reveals that optimal hybrid designs favor heterogeneous component mixing rather than homogeneous patterns, with Mamba-2 and Multi-Head Latent Attention (MLA) emerging as preferred mechanisms. Discovered architectures demonstrate superior training efficiency: our best configuration achieves a final perplexity of 1.0874 with 2.38B parameters while processing 4,320 tokens/second, outperforming significantly larger manually designed models. Full-scale evaluation shows HARMONY's top architectures achieve better loss trajectories than equivalently-sized models using state-of-the-art configurations including Mixtral, Jamba, and Samba. Additionally, we demonstrate 91% weak scaling efficiency when training discovered 36B-parameter models across 1,024 GPUs. HARMONY is released as an open framework with comprehensive tools for building and training hybrid models using expert-data-pipeline parallelism, democratizing access to automated architecture design for next-generation language models.

Herron, Emily [ORNL] (ORCID:0000000273008172)↗
DOE OSTI2026

Technoeconomic Design Optimization for Fast Reactors. Part II: Impact of Technoeconomic Constraints on Optimal Design

There is a current drive toward optimizing reactors, particularly small/micro reactors to minimize cost and maximize performance. Previous work has investigated the development of technoeconomic workflows for the design optimization of pool-type fast reactors that aim to deploy into district energy grids. Initial scoping studies verified that the workflow was capable of capturing design trends throughout a variety of design configurations and problem formulations while remaining sufficiently flexible. In this paper, this methodology is applied to understand how cost functions and technoeconomic constraints can drive optimal reactor design. Specifically, the UPu10Zr-fueled fast reactor model from Part I is adapted to include changes in the fissile content limits, control rod worth limits, control rod drive cost, and assumed fuel form. In the case of constraint relaxation at fixed power (fissile content and control rod worth limits), cost sensitivities of 5% to 10% were uncovered. Multi-objective optimization at varying reactor power levels with individualized control rod drives for each assembly (as opposed to one operational and one safety drive) increased cost by $\$10$ to $\$25$ million and substantially altered the optimal core geometry, favoring geometries with substantially fewer control rod placements relative to baseline optimization. Finally, a multi-objective optimization was performed at varying power levels with the fuel form overhauled to metallic, high-assay low-enriched uranium–based U10Zr with more refined fuel cost models. In the case of uranium fueling, the costs increased by at least $50 million relative to the baseline case. Furthermore, economic fuel zoning and lower reactivity swing cores were recovered. Each case serves to demonstrate the value of applying technoeconomic workflows to initial reactor design scoping studies to better understand the trade-off for a proposed concept between different design options.

Argonne Reactor Computation (ARC) codes↗
DOE OSTI2025

Digital twin framework for PIP-II linac: AI-driven multi-scale modeling from ion source to 800 MeV

The PIP-II superconducting linac at Fermilab is designed to deliver multi-megawatt proton beams for neutrino physics and other high-intensity applications. To expedite commissioning and enhance operational reliability, we have developed an EPICS-based data flow framework that seamlessly integrates digital twins (DT) with physical twins (PT). These digital twins comprise high-fidelity beam dynamics models or data-driven surrogate models connected to their physical counterparts through real-time diagnostics and advanced machine-learning algorithms.Central to this framework is Linac_Gen, an accelerated simulation tool that incorporates convolutional neural networks, random forests, and genetic algorithms to provide up to a tenfold speedup in optimizing the accelerator geometry model. An EPICS translator layer ensures interoperability by efficiently mapping lattice parameters across diverse simulation platforms.Our EPICS-based framework supports multiple operational modes—monitoring, passive learning, closed-loop control, and online learning—covering the entire machine lifecycle. By leveraging HPC resources and multi-objective optimization techniques, the digital twin enables adaptive trajectory correction, real-time fault detection, and predictive modeling of beam stability. This comprehensive approach paves the way for robust, high-intensity operation and data-driven accelerator R&D at Fermilab.

Pathak, Abhishek [Fermilab]↗
DOE OSTI2023

OzDES Reverberation Mapping Program: Hβ lags from the 6-yr survey

ABSTRACT Reverberation mapping measurements have been used to constrain the relationship between the size of the broad-line region and luminosity of active galactic nuclei (AGN). This R–L relation is used to estimate single-epoch virial black hole masses, and has been proposed to use to standardize AGN to determine cosmological distances. We present reverberation measurements made with Hβ from the 6-yr Australian Dark Energy Survey (OzDES) Reverberation Mapping Program. We successfully recover reverberation lags for eight AGN at 0.12 < z < 0.71, probing higher redshifts than the bulk of Hβ measurements made to date. Our fit to the R–L relation has a slope of α = 0.41 ± 0.03 and an intrinsic scatter of σ = 0.23 ± 0.02 dex. The results from our multi-object spectroscopic survey are consistent with previous measurements made by dedicated source-by-source campaigns, and with the observed dependence on accretion rate. Future surveys, including LSST, TiDES, and SDSS-V, which will be revisiting some of our observed fields, will be able to build on the results of our first-generation multi-object reverberation mapping survey.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Web-Based Tools for Data-Informed Remedy Optimization: Software Theory and User Guide

This report documents the development and application of two web-based decision-support tools for pump-and-treat (P&T) groundwater remediation systems: PTOLEMY (Pump-and-Treat Optimized Location Evaluation to Maximize Yields) and OPTIMA (Optimization for Pump-and-Treat Implementation, Management, & Assessment). These tools enhance remedy design and management by leveraging advanced computational methods – specifically deep learning and multi-objective optimization – within a user-friendly platform. By integrating data-driven models with established hydrogeological knowledge, PTOLEMY and OPTIMA enable more efficient evaluation of well placement and operational strategies, helping site managers balance multiple remediation objectives under complex conditions. Both tools are implemented as modules within the SOCRATES (Suite Of Comprehensive Rapid Analysis Tools for Environmental Sites) web platform, which provides data access, visualization, and analytics to support remedy optimization across sites in the U.S. Department of Energy Office of Environmental Management complex. PTOLEMY is a rapid screening module designed to identify promising locations for new extraction wells. It employs a multi-channel three-dimensional convolutional neural network (MC3D-CNN) trained on high-fidelity simulation data to predict the relative performance (in terms of contaminant mass recovery) of potential well sites. Through an interactive web interface, PTOLEMY visualizes the probability of high performance across a site, highlighting areas where an extraction well is likely to yield above-threshold contaminant removal over a multi-year period. PTOLEMY’s map-based displays and exportable results support transparent communication of screening analyses. By focusing attention on the most favorable candidate locations, the tool augments traditional engineering judgment and physics-based modeling, providing a data informed basis for subsequent detailed evaluations. OPTIMA is a multi objective optimization module designed to find wellfield layouts and operating schedules that meet various cleanup goals. It quickly evaluates thousands of candidate setups – combinations of well locations, timing, and rates – and returns a small set of best trade-off options for comparison. At its core, OPTIMA uses a U-Net-based surrogate model – a deep-learning emulator of a groundwater flow and transport simulator – to dramatically accelerate scenario evaluations. Coupling this fast surrogate with the NSGA-II (Non-dominated Sorting Genetic Algorithm II) evolutionary algorithm, OPTIMA explores a wide decision space of well locations and schedules to identify Pareto-optimal solutions that trade off key objectives (e.g., minimizing cleanup time, maximizing contaminant mass removal, and minimizing plume extent). The tool outputs a family of optimal configurations and visualizes their trade-offs (Pareto frontiers of cleanup metrics and maps of optimized well placements). Site managers can use these results to understand the range of viable strategies and to select candidate designs for more detailed verification. OPTIMA is currently under active development and not yet fully released; this guide provides early documentation to support planning and gather user feedback.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Automated Design of Neuromorphic Networks for Scientific Applications at the Edge

Designing spiking neural networks for neuromorphic deployment is a non-trivial task. It is further complicated when there are resource constraints for the neuromorphic implementation, such as size or power constraints, that may be present in edge applications. In this work, we utilize a previously presented approach, EONS, to design spiking neural networks for a memristive neuromorphic implementation for scientific data applications. We specifically use a multi-objective approach in EONS to maximize network accuracy on the scientific data application task, but also to minimize network size and energy. We illustrate that EONS determines both the network structure and the parameters, removing the burden from the user on determining the appropriate spiking neural network structure, and we show that the resulting networks are very different from the layered structure of typical neural networks. Finally, we show that the multi-objective approach produces smaller, more energy efficient networks than the original EONS approach and produces comparable accuracy to a back-propagation style training approach.

Schuman, Catherine↗
DOE OSTI2025

CALPHAD-based Bayesian optimization to accelerate alloy discovery for high-temperature applications

Two crucial properties influencing the performance of high-temperature alloys are coefficient of thermal expansion (CTE) and phase constitution. It is desirable to have alloys with low CTE, which reduces CTE mismatch with the surface oxide and the likelihood of oxide spallation. Reducing the amount of brittle intermetallic phases such as Sigma (σ) enhances alloy ductility and processability. Here, we propose a multi-objective Bayesian Optimization (BO) model to simultaneously minimize CTE (at an operational temperature of 1150 °C) and T σ (temperature when the Sigma phase completely dissolves in the metal matrix), properties which are obtained from high-throughput CALculation of PHAse Diagrams (CALPHAD). The model successfully identifies several alloys with CTE ≤ 2 × 10 –5 /K and T σ ≤ 500 °C by exploring just 7% of the nickel–chromium–cobalt–aluminum–iron (Ni–Cr–Co–Al–Fe) composition space. Such multi-objective alloy design frameworks can be used to inform additive manufacturing experiments and accelerate alloy discovery for high-temperature energy applications.

36 MATERIALS SCIENCE↗
DOE OSTI2024

A novel design optimization framework to sustain remanufacturability

The ever-increasing global carbon emissions have urged the need for environmentally conscious/sustainable product design, for which the design for remanufacturing (DfRem) is one potential approach. DfRem targets at designing products that have multiple life cycles, thus significantly reducing raw material usage, energy consumption, and carbon emissions. In this paper, we develop a three-stage framework that consists of (1) systematic design space exploration and a multi-objective optimization formulation to minimize the likelihood of failure causes (such as fatigue and wear) and environmental footprint, (2) topology optimization to further reduce material usage without significantly affecting the load-carrying capability of the product, and (3) post-topology optimization design verification to ensure the proposed design satisfies all design constraints. The environmental impact can be assessed at varying comprehensiveness levels (e.g., design and manufacturing phase, use phase) and in terms of carbon or GHG emission, energy use, and waste generation. Because the novel design framework predominantly adjusted the geometry, we focused on mass-based change and energy savings due to sustained remanufacturability. The multi-objective optimization formulation in the first step results in a Pareto optimal set of possible design solutions that the designer can use for the second step. Finally, we demonstrate the utility of this framework through a case study of an engine cylinder head subjected to thermo-mechanical loads, where we find that about 5% of the product mass can be conserved with only about a 3% increase in surface area that has a fatigue life less than 10,000 cycles.

42 ENGINEERING↗
DOE OSTI2026

From Modular ADMS to Plug-and-Play Ops: Distribution Grid Operations with Platform-Level Orchestration to Enable Ambitious App Hosting

The core function of the distribution grid is to provide electricity to consumers affordably, reliably, and securely. In pursuing these core objectives, distribution utilities are accountable to customers, regulators, and in some cases, shareholders. Other third parties such as aggregators and microgrids can also have a stake in the smooth operation of the grid. Each of these stakeholders has economic, business, and/or governance objectives that inform their expectations of the distribution grid. This multi-objective, multi-stakeholder environment creates tension that must be reconciled to successfully design and operate the distribution grid. Innovative companies are competing to bring high-tech solutions to electric utilities and their customers that address each of these objectives. Many developers of advanced distribution management systems (ADMS) and distributed energy resource management systems (DERMS) have adopted a modular architecture that allows grid operators to select functions and features according to their individual system needs. A modular platform also allows the solution provider to develop and integrate specific new product modules; however, the need to pursue multiple objectives with a fixed set of controllable devices makes integration expensive whether it is done at the product development stage or the deployment stage. This cost creates a significant barrier to adoption and can lengthen the product to market time of new solutions. To fundamentally address the complexity of system integration for distribution grid operations, the U.S. Department of Energy Office of Electricity has funded the GridAPPS-D project at PNNL, which streamlines integration by contributing to standards development, defining system architecture, applying advanced mathematics, and developing open-source software to demonstrate the concept of an open data-integration platform for distribution operations. The open data-integration platform concept enables system operators and solution providers to deploy ambitious, best-of-breed applications (or apps) without continually reengineering for integration. Ambitious apps developed by different solution providers will inevitably attempt to achieve different control objectives with the same set of controllable devices. If the open platform itself can resolve these conflicts in a way that achieves the best available outcomes for all apps, doesn’t restrict the ambitious design of apps, and ensures safe and secure operations, apps will be able to plug-and-play with the platform at the same time as other ambitious apps. In this paper, we describe a framework called App Deconfliction that empowers a platform to assign setpoints to controllable devices based on the values preferred by different apps (and even external stakeholder entities like customers or aggregators). The App Deconfliction framework is compatible with several methods for determining setpoint values. We present two methods based on game theory that provide a subtle built-in incentive structure for developers to adapt their apps to the fact that they will be operating in a moderated multi-app environment and to favor device setpoints that have the most effect on their objectives over those that have the least effect. Our simulation-based demonstrations have shown that game-theory-based deconfliction can lead to a 7% improvement in control space utilization compared to design-based methods.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2025

A Joint Search for the Electromagnetic Counterpart to the Gravitational-Wave Binary Black-Hole Merger Candidate S250328ae with the Dark Energy Camera and the Prime Focus Spectrograph

The first detection of an optical counterpart to a gravitational wave signal revealed that collaborative efforts between instruments with different specializations provide a unique opportunity to acquire impactful multi-messenger data. We present results of such a joint search with the Dark Energy Camera (DECam) and Prime Focus Spectrograph (PFS) for the optical counterpart of the LIGO-Virgo-KAGRA event S250328ae, a binary black hole merger candidate of high significance detected at a distance of 511$\pm$82 Mpc and localized within an area of 3 (15) square degrees at 50% (90%) confidence. We observed the 90% confidence area with DECam and identified 36 high-confidence transient candidates after image processing, candidate selection, and candidate vetting. We observed with PFS to obtain optical spectra of DECam candidates, Swift-XRT candidates, and potential host galaxies of S250328ae. In total, 3897 targets were observed by seven pointings covering ~50% of the 90% confidence area. After template fitting and visual inspection, we identified 12 SNe, 159 QSOs, 2975 galaxies, and 131 stars. With the joint observations of DECam and PFS, we found variability in 12 SNe, 139 QSOs, 37 galaxies, and 2 stars. We do not identify any confident optical counterparts, though the association is not ruled out for three variable candidates that are not observed by PFS and 6 QSO candidates without clear variability if the optical counterpart of S250328ae is faint. Despite the lack of confident optical counterparts, this paper serves as a framework for future collaborations between wide-field imagers and multi-object spectrographs to maximize multi-messenger analyses.

Zhang, Haibin [Natl. Astron. Observ. of Japan] (OR↗
DOE OSTI2021

A comprehensive techno-eco-assessment of CO 2 enhanced oil recovery projects using a machine-learning assisted workflow

Carbon dioxide enhanced oil recovery (CO 2 -EOR) projects not only extract residual oil but also sequestrate CO 2 in the depleted reservoirs. Here, this study develops a machine-learning-based workflow to co-optimize the hydrocarbon recovery, CO 2 sequestration volume and project net present value (NPV) simultaneously. Considering the trade-off relationships among the objective functions, support vector regression with Gaussian kernel (Gaussian- SVR) proxies are coupled with multi-objective particle swarm optimization (PSO) protocol and generate Pareto optimal solutions. Taking advantage of the high computational efficacy of the proxy model, economic uncertainties introduced by tax credits, capital costs and oil price are investigated by this study. The results indicate that the tax incentive policy (Section 45Q) plays a vital role in enhancing the economic returns of CO 2 -EOR projects, especially under the depression of crude oil market. The proposed workflow has been successfully implemented to optimize a water alternative CO 2 (CO 2 -WAG) injection project in a depleted oil sand in the US. The optimization results yield an incremental oil production of 15.8 MM STB and 1.37 MM metric tons of CO 2 storage in a 20-year development strategy, with the highest project NPV to be 205.6 MM US dollars.

03 NATURAL GAS↗
DOE OSTI2023

An automated and portable method for selecting an optimal GPU frequency

Power consumption poses a significant challenge in current and emerging graphics processing unit (GPU) enabled high-performance computing systems. In modern GPUs, dynamic voltage frequency scaling (DVFS) appears to be a reliable control to regulate power consumption and performance. However, the DVFS design space is large - hence, brute-force approaches are infeasible to select the optimal frequency. Furthermore, no single frequency can be universally optimal for applications with varying computational intensities. Thus, the application's complexity and the availability of a wide range of frequency settings are a challenge in selecting the optimal frequency configuration for a given GPU workload. To that end, this paper proposes a systematic approach that consists of three steps. The feature characterization study identifies the fine-grain GPU utilization metrics that influence the power consumption and execution time of a given workload. To understand the performance, power, and energy consumption behaviors of a workload across GPU's DVFS design space, we derived analytical power and performance models using the identified fine-grain features. Here, it is shown that the same set of GPU utilization metrics can estimate both the power consumption and execution time while being agnostic of changes to frequency and input sizes. Applying a power control with the single objective of reducing power may cause performance degradation, leading to more energy consumption. A multi-objective approach is proposed to select the optimal GPU DVFS configuration for a workload that reduces power consumption with negligible degradation in performance. The evaluation was conducted using SPEC ACCEL benchmarks and three real applications - NAMD LAMMPS, and LSTM on NVIDIA GV100, GA100, and AMD MI210 GPUs. On average, real applications showed 29.6% energy savings with a performance loss of 5.2% on GA100 and 22.6% energy savings with a performance loss of 4.7% on GV100. Moreover, the proposed models are portable to real applications, GPU architectures, and vendors, and require metric collection at only the default frequency rather than all supported DVFS configurations. Additionally, we conducted a comparison between our models and the GPU assembly instructions (PTX)-based static models. The results revealed a significant reduction in the average error rates, with a decrease from 19.7% to 3.1% for power models and from 29.4% to 5.2% for performance models.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Practical CO2—WAG Field Operational Designs Using Hybrid Numerical-Machine-Learning Approaches

Machine-learning technologies have exhibited robust competences in solving many petroleum engineering problems. The accurate predictivity and fast computational speed enable a large volume of time-consuming engineering processes such as history-matching and field development optimization. The Southwest Regional Partnership on Carbon Sequestration (SWP) project desires rigorous history-matching and multi-objective optimization processes, which fits the superiorities of the machine-learning approaches. Although the machine-learning proxy models are trained and validated before imposing to solve practical problems, the error margin would essentially introduce uncertainties to the results. In this paper, a hybrid numerical machine-learning workflow solving various optimization problems is presented. By coupling the expert machine-learning proxies with a global optimizer, the workflow successfully solves the history-matching and CO2 water alternative gas (WAG) design problem with low computational overheads. The history-matching work considers the heterogeneities of multiphase relative characteristics, and the CO2-WAG injection design takes multiple techno-economic objective functions into accounts. This work trained an expert response surface, a support vector machine, and a multi-layer neural network as proxy models to effectively learn the high-dimensional nonlinear data structure. The proposed workflow suggests revisiting the high-fidelity numerical simulator for validation purposes. The experience gained from this work would provide valuable guiding insights to similar CO2 enhanced oil recovery (EOR) projects.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗