SEARCH · Search NASA

Results for “surrogate model development”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 253 records · Page 14

All sources NASA NTRS DOE OSTI

DOE OSTI2026

Spatial Optimization of Multiscale Biorefinery Deployment for a Diversified Bioeconomy in the United States

Strategic biorefinery siting is critical for a diversified bioeconomy, yet industry, policy, and research often focus on either large-scale biofuel plants or smaller-scale specialty bioproduct facilities, with limited coordination across scales. We address this gap by modeling biorefinery deployment spanning a 28-fold difference in capacity. We developed an open-source, spatially explicit framework integrating techno-economic analysis with logistics and refinery cost surrogate models to evaluate multiscale miscanthus-derived biorefineries across the rainfed U.S. for the production of ethanol, succinic acid, lactic acid, potassium sorbate, and acrylic acid. Overall costs change little as feedstock density increases, while transport distances decrease by ∼30 to 67% (∼100 km) and siting flexibility improves. Specifically, a 5-fold feedstock density increase (2% to 10% of suitable land) reduces minimum selling prices by <10% (e.g., 0.27 USD·gal –1 for ethanol). This limited economic sensitivity suggests dense planting is not required for competitive deployment, particularly for smaller-scale facilities. Representing collection areas as irregular rather than circular expands the feasible space under low-density scenarios. While large-scale refineries anchor regional supply chains, smaller facilities retain spatial flexibility even when large refineries are established. These findings highlight the importance of spatial representation and multiscale coordination for robust, regionally tailored biomanufacturing networks to advance renewable carbon integration without extensive land conversion.

biorefinery siting↗

DOE OSTI2021

Probing intermediate temperature heat release in autoignition of C3-C4 iso-alcohol/gasoline blends

In this work, we report an experimental and modeling study on the blending effects of C3–C4 iso-alcohols, namely iso-propanol and iso-butanol, on the characteristics of intermediate temperature heat release (ITHR) for a research-grade gasoline (FACE-F) in a rapid compression machine operating at diluted/stoichiometric fuel loading, compressed pressure of 43 bar, and compressed temperatures from 700 to 970 K. Changes in ITHR behavior are characterized through ITHR extent and evolution at 0 to 30 vol% iso-alcohol blending levels. Experimental observations reveal the strong promoting and decelerating effects on ITHR extent and evolution, respectively, within the low-temperature regime for both iso-alcohols, with stronger effects seen for iso-propanol; within the intermediate-temperature regime, the influence of both iso-alcohols diminishes greatly. Comprehensive chemical kinetic modeling is undertaken using a recently developed gasoline surrogate/alcohol model in conjunction with a five-component gasoline surrogate (FGF-LLNL), with good agreement obtained with the experiments for all the blends. Sensitivity, rate of production and flux analyses highlight the importance of chemical kinetic interactions via both fuel-specific and non-fuel-specific reactions. Disabling all the chemical kinetic interactions between iso-alcohol and FGF-LLNL in carbonated/non-fuel-specific species leads to somewhat extended ITHR duration and reduced ignition reactivity. Direct comparison between FGF-LLNL/iso-propanol and FGF-LLNL/iso-butanol blends is also made, where iso-propanol and iso-butanol are found to influence ITHR characteristics via significantly different chemical kinetic interactions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2022

Computation of conventional and alternative jet fuel sensitivity to lean blowout

Large Eddy Simulations (LES) are performed to compute the sensitivity of a conventional (A-2) and an alternate bio-jet (C-1) fuel to Lean Blowout (LBO). A realistic aviation gas turbine engine combustor configuration is considered. Reliable experimental LBO data and OH* chemiluminescence data for the conventional and alternate jet fuel in the combustor configuration have recently become available. The present work utilizes a highly automated, on-the-fly meshing strategy, along with adaptive mesh refinement, to demonstrate the feasibility of capturing the realistic combustion processes. A Lagrangian framework, with initial conditions specified using measurements of spray statistics, is used to model the fuel spray. Newly developed compact reaction mechanisms based on fuel surrogates are validated for the A-2 and the C-1 fuels. The compact reaction mechanisms are implemented using a detailed finite rate chemistry solver. Spray statistics computed by the present LES simulations compare well with available measurements at stable flame conditions near the lean blowout limit. The computed shape of the stable flame as represented by line integrated OH concentrations compares well with the experimental OH* chemiluminescence data. Lean blowout is reached by gradually decreasing the fuel flow rate in the computations, similar to that in the experiments. The results of the LES simulations effectively capture the fuel composition effects and estimate the sensitivity of the LBO limits to the fuel type. The computed trends in LBO limits agree within engineering accuracy with the experimental results for conventional and alternative aviation fuels. The methodology for predicting the fuel composition effects on the lean blowout limits in a fully resolved realistic, complex combustor is established for the first time.

Adaptive mesh refinement↗

DOE OSTI2023

Towards a unified nonlocal, peridynamics framework for the coarse-graining of molecular dynamics data with fractures

Molecular dynamics (MD) has served as a powerful tool for designing materials with reduced reliance on laboratory testing. However, the use of MD directly to treat the deformation and failure of materials at the mesoscale is still largely beyond reach. In this work, we propose a learning framework to extract a peridynamics model as a mesoscale continuum surrogate from MD simulated material fracture data sets. Firstly, we develop a novel coarse-graining method, to automatically handle the material fracture and its corresponding discontinuities in the MD displacement data sets. Inspired by the weighted essentially non-oscillatory (WENO) scheme, the key idea lies at an adaptive procedure to automatically choose the locally smoothest stencil, then reconstruct the coarse-grained material displacement field as the piecewise smooth solutions containing discontinuities. Then, based on the coarse-grained MD data, a two-phase optimization-based learning approach is proposed to infer the optimal peridynamics model with damage criterion. In the first phase, we identify the optimal nonlocal kernel function from the data sets without material damage to capture the material stiffness properties. Then, in the second phase, the material damage criterion is learnt as a smoothed step function from the data with fractures. As a result, a peridynamics surrogate is obtained. As a continuum model, our peridynamics surrogate model can be employed in further prediction tasks with different grid resolutions from training, and hence allows for substantial reductions in computational cost compared with MD. We illustrate the efficacy of the proposed approach with several numerical tests for the dynamic crack propagation problem in a single-layer graphene. Our tests show that the proposed data-driven model is robust and generalizable, in the sense that it is capable of modeling the initialization and growth of fractures under discretization and loading settings that are different from the ones used during training.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Grassmannian diffusion maps based surrogate modeling via geometric harmonics

Abstract A novel surrogate model based on the Grassmannian diffusion maps (GDMaps) and utilizing geometric harmonics (GH) is developed for predicting the response of complex physical phenomena. The method utilizes GDMaps to obtain a low‐dimensional representation of the underlying behavior of physical/mathematical systems with respect to uncertain input parameters. Using this representation, GH, an out‐of‐sample extension technique, is employed to create a global map from the input parameter space to a Grassmannian diffusion manifold. GH is further employed to locally map points on the diffusion manifold onto the tangent space of a Grassmann manifold. The exponential map is then used to project the points in the tangent space onto the Grassmann manifold, where reconstruction of the full solution is performed. The performance of the proposed surrogate model is verified with three examples. The first problem is a toy example used to illustrate the technique. In the second example, errors associated with the various mappings are assessed by studying response predictions of the electric potential of a dielectric cylinder in a homogeneous electric field. The last example applies the method for uncertainty prediction in the strain field evolution in a model amorphous material using the shear transformation zone theory of plasticity.

42 ENGINEERING↗

DOE OSTI2025

Multi-objective surrogate-assisted calibration of CPFEM models using macroscopic response and in situ EBSD measurements of grain reorientation trajectories

Crystal plasticity finite element method (CPFEM) models are widely used to simulate the deformation behaviour of polycrystalline materials, but their calibration is often limited by their high computational cost and the non-convexity of the optimisation landscape. Here, this study develops a multi-objective surrogate-assisted calibration workflow that couples a multi-objective genetic algorithm (MOGA) with an adaptively trained deep neural network (DNN) surrogate model to efficiently identify CPFEM parameters from experimental data. The workflow is demonstrated on three crystal plasticity (CP) formulations of increasing complexity — Voce hardening (VH), two-coefficient latent hardening (LH2), and six-coefficient latent hardening (LH6) — using in situ electron backscatter diffraction (EBSD) measurements of Alloy 617 under uniaxial tensile loading. The CPFEM models are calibrated against the experimentally observed stress–strain response and reorientation trajectories of eight grains, then validated against eight additional trajectories and overall texture evolution. Across the CP formulations, the macroscopic response was reproduced reliably, while differences emerged in the robustness and accuracy of the grain-scale predictions. Including grain reorientation trajectories in the multi-objective calibration improved texture evolution predictions and filtered out physically inconsistent parameter sets that can arise from calibrating against only the stress–strain data. The workflow also demonstrates good transferability of calibrated parameters from a low- to a high-fidelity microstructural model. These results provide practical guidance for integrating in situ microstructural data into CPFEM through efficient, repeatable, and physically meaningful multi-objective calibration.

Crystal plasticity finite element method↗

DOE OSTI2022

Prediction of grain structure after thermomechanical processing of U-10Mo alloy using sensitivity analysis and machine learning surrogate model

Abstract Hot rolling and annealing are critical intermediate steps for controlling microstructures and thickness variations when fabricating uranium alloyed with 10% molybdenum (U-10Mo), which is highly relevant to worldwide nuclear non-proliferation efforts. This work proposes a machine-learning surrogate model combined with sensitivity analysis to identify and predict U-10Mo microstructure development during thermomechanical processing. Over 200 simulations were collected using physics-based microstructure models covering a wide range of thermomechanical processing routes and initial alloy grain features. Based on the sensitivity analysis, we determined that an increase in rolling reduction percentage at each processing pass has the strongest effect in reducing the grain size. Multi-pass rolling and annealing can significantly improve recrystallization regardless of the reduction percentage. With a volume fraction below 2%, uranium carbide particles were found to have marginal effects on the average grain size and distribution. The proposed stratified stacking ensemble surrogate predicts the U-10Mo grain size with a mean square error four times smaller than a standard single deep neural network. At the same time, with a significant speedup (1000×) compared to the physics-based model, the machine learning surrogate shows good potential for U-10Mo fabrication process optimization.

36 MATERIALS SCIENCE↗

DOE OSTI2024

Mini-Canister Radiolysis Testing of ASNF Materials and Surrogates

An experimental irradiation campaign to investigate radiolysis behavior of ASNF was conducted using in situ gas monitoring of small, sealed stainless-steel vessels (mini-canisters) containing aluminum samples with adherent (oxy)hydroxide films under helium backfill. The samples were irradiated with gamma radiation from a Co-60 irradiator. The samples tested included aluminum plate assemblies with lab-grown (oxy)hydroxides as surrogates for fuel as well as an end cropping from an actual ASNF assembly retrieved from long-term wet storage. These experiments enabled investigation of the impacts of various fuel drying approaches on the radiolytic generation rate and measurement of the H 2 yield associated with a reactor exposed sample with reactor-formed (oxy)hydroxide. The resulting data can be incorporated into model development for ASNF in dry storage. This report presents the cumulative results from four surrogate assemblies tested after application of different preparation (drying) conditions as well as the ASNF cropping; some post-irradiation testing was included. The mini-canister results are compared to data from related experimental campaigns that also tested the impact of drying conditions using samples irradiated in glass ampoules and discusses implications of the combined data

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

DOE OSTI2022

Uncertainty Quantification for Dissimilar Material Joints Under Corrosion Environment

Abstract Self-Piercing Riveting (SPR) is one of the most commonly used methods for joining dissimilar materials in the automotive industry. These joints are popular due to their adaptability, high performance and short cycle time. However, since these joints involve two dissimilar materials, they are susceptible to galvanic corrosion in the presence of an electrolyte which is driven by the difference in the equilibrium potential of the metals. This can affect the safety and resilience of these joints. In this paper, we focus on galvanic corrosion in Al-Fe SPR joints. A Machine learning (ML) based surrogate model, which is based off of FE simulations, for statistical corrosion analysis is developed. This model enables the resilience and reliability analysis of SPR joints under corrosion environment. In this study, first a physics-based finite element (FE) corrosion model has been developed to simulate the galvanic corrosion between a Fe cathode and an Al anode of a SPR joint. This model takes into account the effect of the crystal microstructure of the Al anode and the precipitation of the corrosion product. Several geometric and environmental factors including crevice gap, roughness of anode, conductivity, pH and the temperature of the electrolyte that effect corrosion are investigated. A thorough Uncertainty Quantification (UQ) analysis is conducted for the overall corrosion behavior of the Fe-Al SPR joints using a novelistic Probabilistic Confidence-Based Adaptive Sampling (PCAS) technique. PCAS is used to train the surrogate model by identifying the critical sampling points and thus reducing the overall computational costs. It is found that the electrolyte temperature has the largest effects on the material loss and needs to be managed closely for better corrosion control. By understanding the corrosion performance and resultant uncertainty impact on SPR joints, the reliability and resilience of these joints can be improved.

36 MATERIALS SCIENCE↗

DOE OSTI2024

A physics-constrained deep learning surrogate model of the runaway electron avalanche growth rate

A surrogate model of the runaway electron avalanche growth rate in a magnetic fusion plasma is developed. This is accomplished by employing a physics-informed neural network (PINN) to learn the parametric solution of the adjoint to the relativistic Fokker–Planck equation. The resulting PINN is able to evaluate the runaway probability function across a broad range of parameters in the absence of any synthetic or experimental data. This surrogate of the adjoint relativistic Fokker–Planck equation is then used to infer the avalanche growth rate as a function of the electric field, synchrotron radiation and effective charge. Predictions of the avalanche PINN are compared against first principle calculations of the avalanche growth rate with excellent agreement observed across a broad range of parameters.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2022

Physics-Informed Neural Network Modeling of Li-Ion Batteries: Preprint

Li-ion batteries (LIB) are a promising solution to enable storage of intermittent energy sources due to their high energy density. However, LIBs are known to significantly degrade after about 1000 charge-discharge cycles. LIBs degrade following different degradation modes and at a rate that depends on the operating conditions (external temperature, load). To plan the installation of batteries, appropriate understanding and prediction capabilities of their lifecycle is needed. High-fidelity numerical models of LIBs such as the pseudo-two-dimensional (P2D) model have been shown to accurately represent the charge-discharge-cycle of an LIB given ac- curate choice of the physical parameters. Given the large number of P2D parameters, adjusting them using forward runs is intractable. This work describes the development of a physics-informed neural network (PINN) as a surrogate substitute of the P2D model that captures parameter dependence. The PINN is advantageous as it needs little to no data, and can naturally encode the dependence of every model parameter. Here, a specific training procedure is adopted to efficiently cover parameter space, handle model stiffness and enforce boundary conditions. The trained PINN is validated against numerical solutions of the P2D model, and its applicability to battery degradation modeling is discussed.

battery degradation↗

DOE OSTI2026

Heliostat Sizing Methodology for Solar Heat for Industrial Processes

This study presents a method to obtain a heliostat size that minimizes the levelized cost of a heliostat-based concentrating solar thermal system for industrial process heat (IPH) applications at operating temperatures from 565 to 1550 degrees Celsius. The method extends prior work by embedding a routine for system design that obtains near-optimal subsystem sizes, increasing the fidelity of drive cost functions, and adding an optical performance model. An illustrative business case is developed for Daggett, California, targeting specified annual thermal energy outputs of 50 to 400 GWhth. Optical performance is modeled using verified estimates from the literature. A surrogate heliostat cost model, derived from commercial heliostat designs and scaled for production volume, installation, and operations and maintenance costs, is used to develop cost functions. Results show that heliostat size strongly affects the levelized cost of heat (LCOH), producing a characteristic U-shaped trend with a robust near-optimal window of 8 - 12 m2; the heliostat size producing the lowest project cost in our study grows slightly as the project size increases, and is reduced as the operating temperature increases. The findings in this study are consistent with the general trend of smaller heliostats under deployment at existing projects for high-temperature industrial process heat and reflect the significant reduction in power electronics and other per-heliostat costs. The methodology we propose is general and can be tailored to revised cost curves as the technology continues to evolve.

14 SOLAR ENERGY↗

DOE OSTI2026

Heliostat sizing methodology for concentrating solar thermal industrial process heat projects

This study presents a method to obtain a heliostat size that minimizes the levelized cost of heat (LCOH) of a heliostat-based concentrating solar thermal system for applications of solar heating for industrial processes at operating temperatures from 565 to 1550°C. The method extends prior work by embedding a routine for system design that obtains near-optimal subsystem sizes, increasing the fidelity of drive cost functions, and adding an optical performance model to supplement the previously developed cost models, which we update to reflect current pricing trends. An illustrative business case is developed for Daggett, California, targeting specified annual thermal energy outputs of 50 to 400 GWh th . Optical performance is modeled using verified estimates from the literature. A surrogate heliostat cost model, derived from commercial heliostat designs and scaled for production volume, installation, and operations and maintenance costs, is used to develop cost functions. Results show that heliostat size strongly affects the LCOH, producing a characteristic U-shaped trend with a robust near-optimal window of 7-20 m 2 ; the heliostat size producing the lowest project cost in our study grows slightly as the project size increases, and is reduced as the operating temperature increases. The findings in this study are consistent with the general trend of smaller heliostats being deployed at existing projects for high-temperature industrial process heat and reflect the significant reduction in power electronics and other per-heliostat costs. The methodology we propose is general and can be tailored to revised cost curves as the technology continues to evolve.

14 SOLAR ENERGY↗

DOE OSTI2022

A computational pipeline to generate a synthetic dataset of metal ion sorption to oxides for AI/ML exploration

The charged mineral/electrolyte interfaces are ubiquitous in the surface and subsurface–including the surroundings of the geological disposal sites for radioactive waste. Therefore, understanding how ions interact with charged surfaces is critically important for predicting radionuclide mobility in the case of waste leakage. At present, the Surface Complexation Models (SCMs) are the most successful thermodynamic frameworks to describe ion retention by mineral surfaces. SCMs are interfacial speciation models that account for the effect of the electric field generated by charged surfaces on sorption equilibria. These models have been successfully used to analyze and interpret a broad range of experimental observations including potentiometric and electrokinetic titrations or spectroscopy. Unfortunately, many of the current procedures to solve and fit SCM to experimental data are not optimal, which leads to a non-transferable or non-unique description of interfacial electrostatics and consequently of the strength and extent of ion retention by mineral surfaces. Recent developments in Artificial Intelligence (AI) offer a new avenue to replace SCM solvers and fitting algorithms with trained AI surrogates. Unfortunately, there is a lack of a standardized dataset covering a wide range of SCM parameter values available for AI exploration and training–a gap filled by this study. Here, we described the computational pipeline to generate synthetic SCM data and discussed approaches to transform this dataset into AI-learnable input. First, we used this pipeline to generate a synthetic dataset of electrostatic properties for a broad range of the prototypical oxide/electrolyte interfaces. The next step is to extend this dataset to include complex radionuclide sorption and complexation, and finally, to provide trained AI architectures able to infer SCMs parameter values rapidly from experimental data. Here, we illustrated the AI-surrogate development using the ensemble learning algorithms, such as Random Forest and Gradient Boosting. These surrogate models allow a rapid prediction of the SCM model parameters, do not rely on an initial guess, and guarantee convergence in all cases.

Li, Chunhui↗

DOE OSTI2025

Quantum chemically calculated Abraham parameters for quantifying and predicting polymer hydrophobicity

The leakage and accumulation of plastic in the environment is a significant and growing problem with numerous detrimental impacts and has led to a push toward the design and development of more environmentally benign materials. To this end, we have developed a quantum chemistry-based model for predicting the mobility of polymer materials from molecular structure. Hydrophobicity is used as a surrogate for mobility given that hydrophobic interactions drive much of the partitioning of contaminants in and out of various environmentally relevant compartments. To model polymer hydrophobicity, we adjusted a previously developed Quantum Chemically Calculated Abraham Parameter model to calculate Abraham parameters of small molecules from molecular structure information. The resulting model predicted the octanol-water partition coefficient (K OW ) of polymer repeating units with a root mean square error (RMSE) of 0.48 (log scale). Additionally, the hydrophobicity of high molecular weight polymer materials was captured through solubility parameters and Nile red staining experiments from the literature and predicted with RMSEs of 1.21 (J/cc) 0.5 and 3.42 nm, respectively. Finally, to test the environmental applicability of the model, the relative adsorption capacity of three polymers was predicted and used to unify sorption isotherms across multiple sorbates and polymer sorbents.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2021

Composite binder, processing, and particle size effects on mechanical properties of non-hazardous high explosive surrogates

Plastic-bonded explosives (PBX) are polymer matrix composites that are highly loaded with micron-scale explosive crystals. The thermal and mechanical properties of these materials must be well-understood so that appropriate engineering models can be developed, but testing such properties under many scenarios can be hazardous and expensive. A non-hazardous PBX surrogate (“mock”) may be used in place of a PBX in many experiments if the thermomechanical properties are well-matched. Our previous research has shown that 5-iodo-2'-deoxyuridine (IDOX) is a promising crystalline filler material for the explosive cyclotetramethylene-tetranitramine (HMX). In this study, we investigate how the binder composition, variations in the IDOX particle size distribution, and pressing temperature affect the compressive strength and elastic modulus. We observe that while the binder system used in the composite dominates the mechanical response, the particle size distribution and pressing conditions do have a measurable effect on the material compliance and strength.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

DOE OSTI2024

Creep and Creep Fracture Modeling with Surrogate Creep Models and the Extended Finite Element Method

Alloy components in advanced nuclear reactors will be subjected to environmental conditions that could include high temperatures, irradiation, and exposure to corrosive salts. These conditions could lead to the formation of crack-like defects, which could grow over time in a mechanism known as creep crack growth (CCG). Predicting growth rates of these defects is important for assessing the safe operating life of advanced reactors. This project documents progress toward developing and testing next-generation data-driven constitutive models for deformation creep. It also documents the application of the extended finite element method in conjunction with surrogate creep models to predict CCG parameters under a variety of conditions. These important incremental developments contribute to the longer-term objective of developing microstructure-aware constitutive models that can be used for predicting creep deformation and CCG at the component scale with improved accuracy.

36 MATERIALS SCIENCE↗

DOE OSTI2025

wa-hls4ml and lui-gnn: A benchmark and GNN-based surrogate model for hls4ml resource and latency estimation

As machine learning (ML) increasingly serves as a tool for addressing real-time challenges in scientific applications, the development of advanced tooling has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as model synthesis, are now becoming limiting factors in the rapid iteration of designs. To reduce these emerging constraints, multiple efforts are being launched toward designing an ML-based surrogate model that estimates resource usage of synthesized accelerator architectures. This model would reduce the design iteration time, especially when designing within a set of given hardware constraints. This approach shows considerable potential, but as it stands, the effort is early and would benefit from coordination and standardization to assist future work as it emerges. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of more than 100,000 fully connected neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. In addition to the resource utilization and latency data provided, the dataset includes generated artifacts and log files for many of the synthesized neural networks, in order to support future research in ML-based code generation. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, as well as the average performance across a subset of the dataset. We measure the performance of a given predictor model through multiple metrics, including $R^2$ score and SMAPE on regression tasks, as well as inference time to further characterize the estimator under test. Additionally, we introduce the latency/utilization inference graph neural network (lui-gnn), a surrogate model that uses a graph neural network to represent input architectures in the form of a directed graph. This graph representation allows for a diverse set of model architectures to all be effectively handled by a surrogate model. We present the architecture and performance of the model, as evaluated by the new proposed benchmark, including SMAPE, $R^2$ score, and inference times, and find that lui-gnn generally predicts latency and utilization for the 75\% quantile within several percent of the synthesized resources on the synthetic test dataset, indicating that this approach of estimating resource and latency via a surrogate models has promise and warrants further research.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗