Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “model driven architecture”

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 397 records · Page 22

Summary Report of Working Group 2: Computation

The working group on computation addressed three physics areas: (i) plasma-based accelerators (laser-driven and beam-driven), (ii) high gradient structure-based accelerators, and (iii) electron beam sources and transport [1]. Highlights of the talks in these areas included new models of breakdown on the microscopic scale, new three-dimensional multipacting calculations with both finite difference and finite element codes, and detailed comparisons of new electron gun models with standard models such as PARMELA. The group also addressed two areas of advances in computation: (i) new algorithms, including simulation in a Lorentz-boosted frame that can reduce computation time orders of magnitude, and (ii) new hardware architectures, like graphics processing units and Cell processors that promise dramatic increases in computing power. Highlights of the talks in these areas included results from the first large-scale parallel finite element particle-in-cell code (PIC), many order-of-magnitude speedup of, and details of porting the VPIC code to the Roadrunner supercomputer. The working group featured two plenary talks, one by Brian Albright of Los Alamos National Laboratory on the performance of the VPIC code on the Roadrunner supercomputer, and one by David Bruhwiler of Tech-X Corporation on recent advances in computation for advanced accelerators. Highlights of the talk by Albright included the first one trillion particle simulations, a sustained performance of 0.3 petaflops, and an eight times speedup of science calculations, including back-scatter in laser-plasma interaction. Highlights of the talk by Bruhwiler included simulations of 10 GeV accelerator laser wakefield stages including external injection, new developments in electromagnetic simulations of electron guns using finite difference and finite element approaches.

43 PARTICLE ACCELERATORS↗

Impact of Technology and Mission Variations on Aircraft Designed for Advanced Air Mobiity

NASA is conducting investigations in Advanced Air Mobility (AAM) aircraft and operations, including the development of Urban Air Mobility (UAM) aircraft designs that can be used to focus and guide research activities in support of AAM. This report is an investigation of the impact of technology and mission variations on several of the NASA AAM concept aircraft: quadrotor, quiet single main rotor, side-by-side, and tiltrotor configurations, with turboshaft and electric propulsion variants for each. First, the mission and aircraft models of the baseline designs were reassessed and updated, including rotor geometry optimization, update of the rotor performance models, and disk loading optimization. For these eight designs, technology and mission excursions were performed. Relative to the calibration cases that can be considered examples of good design practice, the impact of the weight technology factors is significant. For the electric aircraft, there is a very large impact of battery specific energy (Wh/kg), and correspondingly a very large impact of mission range. The vision of Advanced Air Mobility is driven by missions that will enable new transportation capabilities. Hence it is appropriate to compare Concept Vehicles of different lift and propulsive architectures, all designed to accomplish the same UAM mission. It is also useful however to consider specific missions that can take advantage of the strengths of individual aircraft configurations. So alternate designs were also developed for the concept vehicles: for turboshaft aircraft, longer unrefueled range, including faster cruise speed for the tiltrotor; for electric aircraft, shorter range and more realistic battery weight.

Technology↗

Bi-fidelity modeling of uncertain and partially unknown systems using DeepONets

Recent advances in modeling large-scale, complex physical systems have shifted research focuses towards data-driven techniques. However, generating datasets by simulating complex systems can require significant computational resources. Similarly, acquiring experimental datasets can prove difficult. For these systems, often computationally inexpensive, but in general inaccurate models, known as the low-fidelity models, are available. Here in this paper, we propose a bi-fidelity modeling approach for complex physical systems, where we model the discrepancy between the true system's response and a low-fidelity response in the presence of a small training dataset from the true system's response using a deep operator network, a neural network architecture suitable for approximating nonlinear operators. We apply the approach to systems that have parametric uncertainty and are partially unknown. Three numerical examples are used to show the efficacy of the proposed approach to model uncertain and partially unknown physical systems.

MATHEMATICS AND COMPUTING↗

Agentic artificial intelligence for multistage physics experiments at a large-scale user facility particle accelerator

We present a language-model-driven agentic artificial intelligence (AI) system to autonomously execute multistage physics experiments on a production synchrotron light source. Implemented at the Advanced Light Source particle accelerator, the system translates natural language user prompts into structured execution plans that combine archive data retrieval, control-system channel resolution, automated script generation, controlled machine interaction, and analysis. In a representative machine physics task, we show that preparation time was reduced by 2 orders of magnitude relative to manual scripting even for a system expert, while operator-standard safety constraints were strictly upheld. Core architectural features, plan-first orchestration, bounded tool access, and dynamic capability selection, enable transparent, auditable execution with fully reproducible artifacts. These results establish a blueprint for the safe integration of agentic AI into accelerator experiments and demanding machine physics studies, as well as routine operations, with direct portability across accelerators worldwide and, more broadly, to other large-scale scientific infrastructures.

Accelerator/storage ring control systems↗

Bayesian Physics Informed Spatio-Temporal Network for Streamflow Data Imputation

Reliable reconstruction of incomplete streamflow records is critical for improving hydrological forecasting, flood preparedness, and water resource management. However, large observational gaps and uncertainties in governing physical parameters limit the accuracy of traditional statistical and machinelearning imputation frameworks. To address these challenges, we develop a Bayesian Physics-Informed Spatio-Temporal Network (BPI-STNet) that jointly captures spatial and temporal dependencies while enforcing hydrologic consistency through embedded physical constraints. The framework integrates a GraphSAGE-LSTM architecture to model spatial connectivity across gauges and temporal flow dynamics, coupled with a Bayesian update mechanism to estimate uncertain parameters in a simplified water-balance framework. Unlike conventional physics-informed networks that rely on sampling-based posterior estimation, BPI-STNet derives an analytic solution to the inverse problem, allowing closed-form Bayesian updates of uncertain parameters Λ={α,β,k} using Gaussian priors and likelihoods. Applied to daily observations from the Susquehanna River Basin (1980-2022), BPI-STNet achieves substantial improvements over a purely data-driven RGNN baseline, which reduced RMSE by 23 % and MAE by 9 %, and achieving an average NSE values up to 0.96. The results demonstrate that coupling Bayesian inference with physics-informed learning yields physically consistent, uncertainty-aware reconstructions that preserve the temporal persistence and statistical distribution of observed flows. The proposed framework establishes a generalizable paradigm for data-sparse hydrologic systems where both data fidelity and physical interpretability are essential.

Krishnan Kutty Ambika, Anukesh [ORNL] (ORCID:00000↗

Beyond Melting: Amorphous Bonding for Joining and Consolidation

Crystallization may be the hidden constraint in thermoplastic composite manufacturing. It requires tightly controlled cooling, induces residual stresses through shrinkage, and introduces path-dependent behavior that complicates predictive modeling yet remains essential for structural performance. This work asks: can bonding be achieved without relying on melt-driven crystallization? To address this, thin (5–20 μm) polyetherimide (PEI) interlayers are pre-healed to slow-cooled polyaryletherketone (PAEK) in two contexts. The first, Thermabond®, is sub-melt joining of low melt-PAEK laminates. Results show that bond quality is governed primarily by processing (i.e., adequate healing and film handling) rather than modest changes in interlayer thickness. This concept is then extended to laminate-scale manufacturing through an architecture known as OATMEAL (Out-of-autoclave Amorphous/semicrystalline Thermoplastic Material for Energy-efficient Aerospace-grade Laminates). PEI is healed to carbon fiber reinforced polyetheretherketone (PEEK) at the prepreg and excess PEI is then ablated from the surface. Crystallinity is developed off-line during prepreg fabrication, while subsequent consolidation occurs below the melt temperature to preserve it. Cross-ply warpage experiments show that, contrary to intuition, repeated amorphous interfaces reduce global curvature by lowering the effective stress lock-in temperature and eliminating crystallization shrinkage from the lamina response. Correspondingly, laminate behavior is accurately predicted using classical laminate theory (CLT) with a single effective stress-free temperature, whereas conventional CF/PEEK requires accounting for crystallization-driven effects. By decoupling interfacial healing from crystallization, OATMEAL enables sub-melt consolidation, reduces energy consumption by up to 75%, and increases manufacturing throughput by fivefold. These results demonstrate that amorphous bonding is not only a joining strategy, but a pathway to more predictable and scalable thermoplastic composite manufacturing.

solidification↗

Integrating LEO and GEO Observations: Toward Optimal Summertime Satellite Precipitation Retrieval

Abstract Reliable quantitative precipitation estimation with a rich spatiotemporal resolution is vital for understanding the Earth’s hydrological cycle. Precipitation estimation over land and coastal regions is necessary for addressing the high degree of spatial heterogeneity of water availability and demand, and for resolving the extremes that modulate and amplify hazards such as flooding and landslides. Advancements in computation power along with unique high spatiotemporal and spectral resolution data streams from passive meteorological sensors aboard geosynchronous Earth-orbiting (GEO) and low Earth-orbiting (LEO) satellites offer exciting opportunities to retrieve information about surface precipitation phenomena using data-driven machine learning techniques. In this study, the capabilities of U-Net–like architecture are investigated to map instantaneous, summertime surface precipitation intensity at the spatial resolution of 2 km. The calibrated brightness temperature products from the Global Precipitation Measurement (GPM) Microwave Imager (GMI) radiometer are combined with multispectral images (visible, near-infrared, and infrared bands) from the Advanced Baseline Imager (ABI) aboard the GOES-R satellites as main inputs to the U-Net–like precipitation algorithm. Total precipitable water and 2-m temperature from the Global Forecast System (GFS) model are also used as auxiliary inputs to the model. The results show that the U-Net–like algorithm can capture fine-scale patterns and intensity of surface precipitation at high spatial resolution over stratiform and convective precipitation regimes. The evaluations reveal the potential of extracting relevant, high spatial features over complex surface types such as mountainous regions and coastlines. The algorithm allows users to interpret the inputs’ importance and can serve as a starting point for further exploration of precipitation systems within the field of hydrometeorology.

Meteorology & Atmospheric Sciences↗

Machine Vision based Sample-Tube Localization for Mars Sample Return

A potential Mars Sample Return (MSR) architecture is being jointly studied by NASA and ESA. As currently envisioned, the MSR campaign consists of a series of 3 missions: sample cache, fetch and return to Earth. In this paper, we focus on the fetch part of the MSR, and more specifically the problem of autonomously detecting and localizing sample tubes deposited on the Martian surface. Towards this end, we study two machine-vision based approaches: First, a geometrydriven approach based on template matching that uses hardcoded filters and a 3D shape model of the tube; and second, a data-driven approach based on convolutional neural networks (CNNs) and learned features. Furthermore, we present a large benchmark dataset of sample-tube images, collected in representative outdoor environments and annotated with ground truth segmentation masks and locations. The dataset was acquired systematically across different terrain, illumination conditions and dust-coverage; and benchmarking was performed to study the feasibility of each approach, their relative strengths and weaknesses, and robustness in the presence of adverse environmental conditions.

Detry, R.↗

Operational constraints and strategies for systems to effect the sustainable, solar-driven reduction of atmospheric CO 2

The operational constraints for a 6-electron/6-proton CO 2 reduction system that operates at the concentration of CO 2 in the current atmosphere (p CO 2 = 400 ppm) have been evaluated on a variety of scale lengths that span from laboratory scale to global scale. Due to the low concentration of CO 2 in the atmosphere, limitations due to mass transport of CO 2 from the tropopause have been evaluated through five different regions, each with different characteristic length scales: the troposphere; the atmospheric boundary layer (ABL); the canopy layer; a membrane layer; and an aqueous electrolyte layer. The resulting CO 2 conductances, and associated physical transport limitations, will set the ultimate limit on the efficiency and areal requirements of a sustainable solar-driven CO 2 reduction system regardless of the activity or selectivity of catalysts for reduction of CO 2 at the molecular level. At the electrolyte/electrode interface, the steady-state limiting current density and the concomitant voltage loss associated with the CO 2 concentration overpotential in a one-dimensional solar-driven CO 2 reduction cell have been assessed quantitatively using a mathematical model that accounts for diffusion, migration and convective transport, as well as for bulk electrochemical reactions in the electrolyte. At p CO 2 = 400 ppm, the low diffusion coefficient combined with the low solubility of CO 2 in aqueous solutions constrains the steady-state limiting current density to <0.1 mA cm -2 in a typical electrochemical cell with natural convection and employing electrolytes with a range of pH values. Hence, in such a system, the CO 2 capture area must be 100- to 1000-fold larger than the solar photon collection area to enable a >10% efficient solar-driven CO 2 reduction system (based on the solar collection area). This flux limitation is consistent with estimates of oceanic CO 2 uptake fluxes that have been developed in conjunction with carbon-cycle analyses for use in coupled atmosphere/ocean general circulation models. Two strategies to improve the feasibility of obtaining efficient and sustainable CO 2 transport to a cathode surface at p CO 2 = 400 ppm are described and modeled quantitatively. The first strategy employs yet unknown catalysts, analogous to carbonic anhydrases, that dramatically accelerate the chemically enhanced CO 2 transport in the aqueous electrolyte layer by enhancing the acid–base reactions in a bicarbonate buffer system. The rapid interconversion from bicarbonate to CO 2 in the presence of such catalysts near the cathode surface would in principle yield significant increases in the steady-state limiting current density and allow for >10% solar-fuel operation at the cell level. The second strategy employs a thin-layer cell architecture to improve the diffusive transport of CO 2 by use of an ultrathin polymeric membrane electrolyte. Rapid equilibration of CO 2 at the gas/electrolyte interface, and significantly enhanced diffusive fluxes of CO 2 in electrolytes, are required to increase the steady-state limiting current density of such a system. This latter approach however only is feasible for gaseous products, because liquid products would coat the electrode and therefore thicken the hydrodynamic boundary layer and accordingly reduce the diffusive CO 2 flux to the electrode surface. Regardless of whether the limitations due to mass transport to the electrode surface are overcome on the laboratory scale, at global scales the ultimate CO 2 flux limitations will be dictated by mass transport considerations related to transport of atmospheric CO 2 to the boundary plane of the solar-driven reactor system. The transport of CO 2 across the troposphere/ABL interface, the ABL/canopy layer interface, and the canopy layer/electrolyte interface have therefore been assessed in this work, to provide upper bounds on the ultimate limits for the solar-to-fuel (STF) conversion efficiency for systems that are intended to effect the reduction of atmospheric CO 2 in a sustainable fashion at global scale.

54 ENVIRONMENTAL SCIENCES↗

Reduced‐Order Modeling of Energetic Materials Using Physics‐Aware Recurrent Convolutional Neural Networks in a Latent Space (LatentPARC)

Physics-aware deep learning (PADL) has gained popularity for use in spatiotemporal dynamics simulations, such as those in computational modeling of energetic materials (EM). We show that the challenge PADL methods face while learning complex field evolution problems can be simplified and accelerated by decoupling it into two tasks: learning complex geometric features in evolving fields and modeling dynamics over these features in a lower-dimensional feature space. We build upon our previous work on physics-aware recurrent convolutional neural networks (PARC). PARC embeds knowledge of underlying physics into its neural network architecture for more robust and accurate prediction of evolving physical fields. PARC was shown to effectively learn complex nonlinear features such as the formation of hotspots and coupled shock fronts in various initiation scenarios of EMs, as a function of microstructures, serving effectively as a microstructure-aware burn model. Here, we further accelerate PARC and reduce its computational cost by projecting the original dynamics onto a lower-dimensional invariant manifold, or “latent space.” The projected latent representation encodes the complex geometry of evolving fields (e.g., temperature and pressure) in a set of data-driven features. The reduced dimension of this latent space allows us to learn the dynamics during the initiation of EM with a lighter and more efficient model. We observe a significant decrease in training and inference time while maintaining results comparable to PARC at inference. This work takes steps towards enabling rapid prediction of EM thermomechanics at larger scales and characterization of EM structure–property–performance linkages at a full application scale.

Mathematics and Computing↗

Integrating Cache Performance Modeling and Tuning Support in Parallelization Tools

With the resurgence of distributed shared memory (DSM) systems based on cache-coherent Non Uniform Memory Access (ccNUMA) architectures and increasing disparity between memory and processors speeds, data locality overheads are becoming the greatest bottlenecks in the way of realizing potential high performance of these systems. While parallelization tools and compilers facilitate the users in porting their sequential applications to a DSM system, a lot of time and effort is needed to tune the memory performance of these applications to achieve reasonable speedup. In this paper, we show that integrating cache performance modeling and tuning support within a parallelization environment can alleviate this problem. The Cache Performance Modeling and Prediction Tool (CPMP), employs trace-driven simulation techniques without the overhead of generating and managing detailed address traces. CPMP predicts the cache performance impact of source code level "what-if" modifications in a program to assist a user in the tuning process. CPMP is built on top of a customized version of the Computer Aided Parallelization Tools (CAPTools) environment. Finally, we demonstrate how CPMP can be applied to tune a real Computational Fluid Dynamics (CFD) application.

Waheed, Abdul↗

Applying the Dark Target Aerosol Algorithm with Advanced Himawari Imager Observations During the KORUS-AQ Field Campaign

For nearly 2 decades we have been quantitatively observing the Earth's aerosol system from space at one or two times of the day by applying the Dark Target family of algorithms to polar-orbiting satellite sensors, particularly MODIS and VIIRS. With the launch of the Advanced Himawari Imager (AHI) and the Advanced Baseline Imagers (ABIs) into geosynchronous orbits, we have the new ability to expand temporal coverage of the traditional aerosol optical depth (AOD) to resolve the diurnal signature of aerosol loading during daylight hours. The Korean–United States Air Quality (KORUS-AQ) campaign taking place in and around the Korean peninsula during May–June 2016 initiated a special processing of full-disk AHI observations that allowed us to make a preliminary adoption of Dark Target aerosol algorithms to the wavelengths and resolutions of AHI. Here,we describe the adaptation and show retrieval results from AHI for this 2-month period. The AHI-retrieved AOD is collocated in time and space with existing AErosol RObotic NETwork stations across Asia and with collocated Terra and Aqua MODIS retrievals. The new AHI AOD product matches AERONET, and the standard MODIS product does as well, and the agreement between AHI and MODIS retrieved AOD is excellent, as can be expected by maintaining consistency in algorithm architecture and most algorithm assumptions. Furthermore, we show that the new product approximates the AERONET-observed diurnal signature. Examining the diurnal patterns of the new AHI AOD product we find specific areas over land where the diurnal signal is spatially cohesive. For example, in Bangladesh the AOD in-creases by 0.50 from morning to evening, and in northeast China the AOD decreases by 0.25. However, over open ocean the observed diurnal cycle is driven by two artifacts, one associated with solar zenith angles greater than 70t hat may be caused by a radiative transfer model that does not properly represent the spherical Earth and the other artifact associated with the fringes of the 40 degree glint angle mask. This opportunity during KORUS-AQ provides encouragement to move towards an operational Dark Target algorithm for AHI. Future work will need to re-examine masking including snow mask, reevaluate assumed aerosol models for geosynchronous geometry, address the artifacts over the ocean, and investigate size parameter retrieval from the over-ocean algorithm.

Gupta, Pawan↗

A high-resolution large-eddy simulation framework for wildland fire predictions using TensorFlow

Background: Wildfires are becoming more severe, so we need improved tools to predict them over a wide range of conditions and scales. One approach towards this goal entails the use of coupled fire/atmosphere modelling tools. Although significant progress has been made in advancing their physical fidelity, existing tools have not taken full advantage of emerging programming paradigms and computing architectures to enable high-resolution wildfire simulations. Aims: The aim of this study was to present a new framework that enables landscape-scale wildfire simulations with physical representation of combustion at an affordable cost. Methods: We developed a coupled fire/atmosphere simulation framework using TensorFlow, which enables efficient and scalable computations on Tensor Processing Units. Key Results: Simulation results for a prescribed fire were compared with experimental data. Predicted fire behavior and statistical analysis for fire spread rate, scar area, and intermittency showed overall reasonable agreement. Scalability analysis was performed, showing close to linear scaling. Conclusions: While mesh refinement was shown to have less impact on global quantities, such as fire scar area and spread rate, it benefits predictions of intermittent fire behavior, buoyancy-driven dynamics, and small-scale turbulent motion. Implications: This new simulation framework is efficient in capturing both global quantities and unsteady dynamics of wildfires at high spatial resolutions.

54 ENVIRONMENTAL SCIENCES↗

Very High Efficiency Solar Cell Modules

The Very High Efficiency Solar Cell (VHESC) program is developing integrated optical system - PV modules for portable applications that operate at greater than 50% efficiency. We are integrating the optical design with the solar cell design, and have entered previously unoccupied design space. Our approach is driven by proven quantitative models for the solar cell design, the optical design, and the integration of these designs. Optical systems efficiency with an optical efficiency of 93% and solar cell device results under ideal dichroic splitting optics summing to 42.7 {+-} 2.5% are described.

14 SOLAR ENERGY↗

Knowledge-guided learning with curated prior genetic biomarkers for robust model interpretation

Abstract Motivation Knowledge-guided learning offers effective and robust model training strategies in data-scarce settings by incorporating established domain knowledge, thereby enhancing generalization, robustness, and interpretability. By contrast, conventional deep learning approaches rely purely on data-driven learning, which can limit robust model interpretability, particularly in high-dimensional settings with limited size samples. In computational biology, knowledge-guided learning has primarily leveraged network- and structural-based knowledge, leading to biologically interpretable representations and enhanced predictive performance compared to conventional approaches. However, curated biomarkers, one of the most accessible forms of biological knowledge, remain largely unexplored within knowledge-guided paradigms. Results In this study, we propose a model-agnostic training paradigm, Biomarker-driven Explainable Prior-guided Learning (BioExPL), that can be applied to any neural networks that incorporates curated prior knowledge. BioExPL enforces neural networks to reflect curated biomarker priors in their latent representations through a novel knowledge-alignment loss. BioExPL consistently demonstrated significantly improved predictive performance and enhanced model interpretability with minimized computational overhead in simulation studies and intensive experiments on multiple cancer datasets. BioExPL not only integrates prior curated knowledge into the model but also accurately identifies unknown associated signals additionally. BioExPL is model-agnostic and domain-independent, enabling its integration into diverse neural network architectures. Availability and implementation The open-source is publicly available at: https://github.com/datax-lab/BioExPL.

Baek, Beomsu [Department of Computer Science, Univ↗

Exploring Randomly Wired Neural Networks for Climate Model Emulation

Exploring the climate impacts of various anthropogenic emissions scenarios is key to making informed decisions for climate change mitigation and adaptation. State-of-the-art Earth system models can provide detailed insight into these impacts but have a large associated computational cost on a per-scenario basis. This large computational burden has driven recent interest in developing cheap machine learning models for the task of climate model emulation. In this paper, we explore the efficacy of randomly wired neural networks for this task. We describe how they can be constructed and compare them with their standard feedforward counterparts using the ClimateBench dataset. Specifically, we replace the serially connected dense layers in multilayer perceptrons, convolutional neural networks, and convolutional long short-term memory networks with randomly wired dense layers and assess the impact on model performance for models with 1 million and 10 million parameters. We find that models with less-complex architectures see the greatest performance improvement with the addition of random wiring (up to 30.4% for multilayer perceptrons). Furthermore, of 24 different model architecture, parameter count, and prediction task combinations, only one had a statistically significant performance deficit in randomly wired networks relative to their standard counterparts, with 14 cases showing statistically significant improvement. We also find no significant difference in prediction speed between networks with standard feedforward dense layers and those with randomly wired layers. These findings indicate that randomly wired neural networks may be suitable direct replacements for traditional dense layers in many standard models.

54 ENVIRONMENTAL SCIENCES↗

Shock propagation in aerogel and TPP foams for inertial fusion energy target design

Achieving practical inertial fusion energy (IFE) requires the development of target designs with well-characterized microstructure and compression response. We measured shock dynamics in low-density (17.5–500 mg/cm 3 ) aerogel and two-photon polymerization (TPP) foams using x-ray phase contrast imaging (XPCI) methods and the Velocity Interferometer System for Any Reflector. By analyzing shock front evolution, we examined how target type and density influence shock propagation and energy dissipation. Talbot-XPCI shows that aerogels support a smooth, bowed shock front due to their homogeneous nanometer-scale pore network. In contrast, TPP foams exhibit irregular, stepwise propagation driven by interactions with their periodic micrometer-scale lattice. Shock velocity follows a power-law relation: aerogels deviate from classical ρ −1/2 scaling due to pore-collapse dissipation, while TPP foams follow the trend with larger uncertainties from density variations. Comparisons with xRAGE simulations reveal systematic underestimation of shock speeds. These results provide the first experimental constraints on shock propagation in TPP foams over a wide density range and highlight the influence of internal structure on anisotropic shock behavior. Our findings support improved benchmarking of EOS and hydrodynamic models and inform the design of foam architectures that promote implosion symmetry in IFE capsules.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Exploring the Use of Novel Spatial Accelerators in Scientific Applications

Driven by the need to find alternative accelerators which can viably replace GPUs in next-generation Supercomputing systems, this paper proposes a methodology to enable agile application/hardware co-design. The application-first methodology provides the ability to come up with design of accelerators while working with real-world workloads, available accelerators, and system software. The iterative design process targets a set of kernels in a workload for performance estimates that can prune the design space for later phases of detailed architectural evaluations. To this effect, in this paper, a novel data-parallel device model is introduced that simulates the latency of performance-sensitive operations in an accelerator including data transfers and kernel computation using multi-core CPUs. The use of off-the-shelf simulators, such as pre-RTL simulator Aladdin or multiple tools available for exploring the design of deep neural network accelerators (e.g., Timeloop) is demonstrated for evaluation of various accelerator designs using applications with realistic inputs. Examples of multiple device configurations that are instantiable in a system are explored to evaluate the performance benefit of deploying novel accelerators. The proposed device is integrated with a programming model and system software to potentially explore the impacts of high-level programming languages/compilers and low-level effects such as task scheduling on multiple accelerators. We analyze our methodology for a set of applications that represent high-performance computing (HPC) and graph analytics. The applications include a computational chemistry kernel realized using tensor contractions, triangle counting, GraphSAGE and Breadth-first Search. These applications include kernels such as dense matrix-dense matrix multiplication, sparse matrix-spare matrix multiplication, and sparse matrix-dense vector multiplication. Our results indicate potential performance benefits and insights for system design by including accelerators that realize these kernels along-side general purpose accelerators.

AI, codesign, Accelerated Computing, Modeling and ↗