Scalability study of molecular dynamics simulation on Godson-T many-core architecture
Explore the source record for details and available documents.
charge transport↗
SEARCH · Engineering Papers
Results for “Scalable”
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.
Scalable aqueous-phase fabrication of reduced graphene oxide nanofiltration membranes by an integrated roll-to-roll (R2R) process
Economic fractionation of multicomponent biomass-derived streams is a key challenge in biobased fuels and chemicals production. Kraft black liquor (BL) is generated at ~1 billion tons/yr globally from biomass pulping and contains ~15wt% total solids including lignin, hemicellulose fragments, and inorganics. Membrane-based BL concentration is attractive but challenged by low solute rejections and poor stability in BL, which combines alkaline pH (~13), high dissolved solids content (15+ wt%) and high temperature (70–85°C). Here we discuss the scaled fabrication and characterization of reduced graphene oxide (rGO) nanofiltration membranes by slot die coating on a roll-to-roll (R2R) with integrated vacuum filtration. Here, we obtain high-quality 90×30 cm rGO membranes (referred to as “R2R-rGO membranes” in this work) with ~100 nm average thickness supported on porous poly(ethersulfone) (PES) sheets, and an effective area of 2700cm 2 . We present characterizations of their microstructure and uniformity as mapped over large areas by scanning electron microscopy (SEM), ellipsometry, adhesion test and X-ray photoelectron spectroscopy (XPS). The constructed ellipsometry model for the rGO membrane/PES substrate composite was successfully applied to 48 different locations of the R2R-rGO membranes. The R2R-rGO membranes showed >98% lignin rejection with a steady state flux of ~12 LMH at 50bar in a 15.7wt% total solids (TS) kraft BL stream at 72°C and 2–3 gal/min crossflow rates. Slot die coating on a R2R platform with integrated vacuum filtration can enable rapid fabrication of high-quality GO membranes at scale without organic solvents or volatile organic compounds.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
A scalable wave resource assessment methodology: Application to U.S. waters
Waves deliver large quantities of energy to populated coastlines around the world, and wave energy technology research and development has accelerated over the last two decades. Throughout this time national and regional resource assessments have utilized disparate methodologies, which can cause confusion and skepticism. Here, in this work, we describe a theoretical wave resource assessment methodology that addresses many of the major areas of inconsistency and debate. Applying this revised methodology to U.S. waters, we find the theoretical U.S. wave energy resource to be 3300 TWh/yr, with region totals of 2000 TWh/yr in Alaska, 510 TWh/yr along the U.S. west coast, 380 TWh/yr in Hawaii, 290 TWh/yr along the east coast, 69 TWh/yr in the Gulf of Mexico, and 17 TWh/yr in Puerto Rico and the U.S. Virgin Islands. We also find significant uncertainty in these estimates associated with the underlying model dataset, which emphasizes the importance of thorough model validation and calibration as well as quantifying uncertainty.
16 TIDAL AND WAVE POWER↗
Scalable algorithms for physics-informed neural and graph networks
Physics-informed machine learning (PIML) has emerged as a promising new approach for simulating complex physical and biological systems that are governed by complex multiscale processes for which some data are also available. In some instances, the objective is to discover part of the hidden physics from the available data, and PIML has been shown to be particularly effective for such problems for which conventional methods may fail. Unlike commercial machine learning where training of deep neural networks requires big data, in PIML big data are not available. Instead, we can train such networks from additional information obtained by employing the physical laws and evaluating them at random points in the space–time domain. Such PIML integrates multimodality and multifidelity data with mathematical models, and implements them using neural networks or graph networks. Here, we review some of the prevailing trends in embedding physics into machine learning, using physics-informed neural networks (PINNs) based primarily on feed-forward neural networks and automatic differentiation. For more complex systems or systems of systems and unstructured data, graph neural networks (GNNs) present some distinct advantages, and here we review how physics-informed learning can be accomplished with GNNs based on graph exterior calculus to construct differential operators; we refer to these architectures as physics-informed graph networks (PIGNs). We present representative examples for both forward and inverse problems and discuss what advances are needed to scale up PINNs, PIGNs and more broadly GNNs for large-scale engineering problems.
42 ENGINEERING↗
Scalable Deposition of High-Efficiency Perovskite Solar Cells by Spray-Coating
Not Available
14 SOLAR ENERGY↗
A scalable membrane electrode assembly architecture for efficient electrochemical conversion of CO2 to formic acid
Abstract The electrochemical reduction of carbon dioxide to formic acid is a promising pathway to improve CO 2 utilization and has potential applications as a hydrogen storage medium. In this work, a zero-gap membrane electrode assembly architecture is developed for the direct electrochemical synthesis of formic acid from carbon dioxide. The key technological advancement is a perforated cation exchange membrane, which, when utilized in a forward bias bipolar membrane configuration, allows formic acid generated at the membrane interface to exit through the anode flow field at concentrations up to 0.25 M. Having no additional interlayer components between the anode and cathode this concept is positioned to leverage currently available materials and stack designs ubiquitous in fuel cell and H 2 electrolysis, enabling a more rapid transition to scale and commercialization. The perforated cation exchange membrane configuration can achieve >75% Faradaic efficiency to formic acid at <2 V and 300 mA/cm 2 in a 25 cm 2 cell. More critically, a 55-hour stability test at 200 mA/cm 2 shows stable Faradaic efficiency and cell voltage. Technoeconomic analysis is utilized to illustrate a path towards achieving cost parity with current formic acid production methods.
08 HYDROGEN↗
Hierarchical and scalable integration of nanostructures for energy and environmental applications: a review of processing, devices, and economic analyses
Hierarchical nanostructure, an engineered nanostructure with a higher assembly-level of constituents by low dimensional nano-building blocks, has received extensive attentions in the latest advancements in nanomaterial science and nanotechnology development. The ordered hierarchical nanostructures could offer merits such as more active sites, synergistic properties owing to their geometric complexity and building blocks variation, as well as the resultant multi-functional capabilities. In this review, we specifically focus on the integration approaches that may be feasible for large-scale production and manufacturing of engineered nanomaterials for various industrial applications, and the design criteria of hierarchical nanostructures targeted at various sustainable energy and environmental applications. In addition to chemical synthesis and integration methods, the development and application of hierarchical nanostructures are surveyed in energy conversion, generation, storage, utilization, and sensing devices. Meanwhile, a much-needed survey is also touched upon the techno-economic analyses of engineered nanomaterials, proposing the need of establishing various standard protocols on various factors and parameters that could be envisaged as a guide toward the viable transition to the practices in the future nano-industries.
77 NANOSCIENCE AND NANOTECHNOLOGY↗
SAGIPS: a physics-inspired scalable asynchronous generative inverse-problem solver
Abstract Solving large-scale inverse problems using deep-learning algorithms have become an essential part of modern research and industrial applications. The complexity of the underlying inverse problem may require the utilization of high performance computing systems which poses a challenge on the algorithmic design of the inverse problem solver. Most deep learning algorithms require, due to their design, custom parallelization techniques in order to be resource efficient while showing a reasonable convergence. In this paper we introduce a S calable A synchronous G enerative I nverse P roblem S olver (SAGIPS) on high-performance computing systems. We present a workflow that utilizes an asynchronous ring-allreduce algorithm to transfer the gradients of the generator network across multiple GPUs. Experiments with a scientific proxy application demonstrate that SAGIPS shows near linear weak scaling, together with a convergence quality that is comparable to traditional methods. The approach presented here allows leveraging Generative Adverserial Network across multiple GPUs, promising advancements in solving complex inverse problems at scale.
97 MATHEMATICS AND COMPUTING↗
HELICS: A Co-Simulation Framework for Scalable Multi-Domain Modeling and Analysis
Not Available
24 POWER TRANSMISSION AND DISTRIBUTION↗
Scalable Optimization Methods for Distribution Networks With High PV Integration
Here this paper proposes a suite of algorithms to determine the active- and reactive-power setpoints for photovoltaic (PV) inverters in distribution networks. The objective is to optimize the operation of the distribution feeder according to a variety of performance objectives and ensure voltage regulation. In general, these algorithms take a form of the widely studied ac optimal power flow (OPF) problem. For the envisioned application domain, nonlinear power-flow constraints render pertinent OPF problems nonconvex and computationally intensive for large systems. To address these concerns, we formulate a quadratic constrained quadratic program (QCQP) by leveraging a linear approximation of the algebraic power-flow equations. Furthermore, simplification from QCQP to a linearly constrained quadratic program is provided under certain conditions. The merits of the proposed approach are demonstrated with simulation results that utilize realistic PV-generation and load-profile data for illustrative distribution-system test feeders.
24 POWER TRANSMISSION AND DISTRIBUTION↗
Jumping Droplet Dynamics on Scalable Nanostructured Superhydrophobic Surfaces
Explore the source record for details and available documents.
charge transport↗
Scalable Extended Dynamic Mode Decomposition Using Random Kernel Approximation
The Koopman operator is a linear, infinite-dimensional operator that governs the dynamics of system observables; extended dynamic mode decomposition (EDMD) is a data-driven method for approximating the Koopman operator using functions (features) of the system state snapshots. This paper investigates an approach to EDMD in which the features used provide random approximations to a particular kernel function. The objective of this is computational economy for large data sets: EDMD is generally ill-suited for problems with large state dimension, and its dual kernel formulation (KDMD) is well-suited for such problems only if the number of data snapshots is relatively small. We discuss two specific methods for generating features: random Fourier features and the Nÿstrom method. The first method is a data-independent method for translation-invariant kernels only and involves random sampling in feature space; the second method is a data-dependent empirical method that may be used for any kernel and involves random sampling of data. We first discuss how these ideas may be applied in an EDMD context, as well as a means for adaptively adding random Fourier features. As a result, we demonstrate these methods on two example problems and conclude with an analysis of the relative benefits and drawbacks of each method.
59 BASIC BIOLOGICAL SCIENCES↗
Continuous Microreactor-Assisted Solution Deposition for Scalable Production of CdS Films
Solution deposition offers an attractive, low temperature option in the cost effective production of thin film solar cells. Continuous microreactor-assisted solution deposition (MASD) was used to produce nanocrystalline cadmium sulfide (CdS) films on fluorine doped tin oxide (FTO) coated glass substrates with excellent uniformity. We report a novel liquid coating technique using a ceramic rod to efficiently and uniformly apply reactive solution to large substrates (152 mm × 152 mm). This technique represents an inexpensive approach to utilize the MASD on the substrate for uniform growth of CdS films. Nano-crystalline CdS films have been produced from liquid phase at ~90°C, with average thicknesses of 70 nm to 230 nm and with a 5 to 12% thickness variation. The CdS films produced were characterized by UV-Vis spectroscopy, transmission electron microscopy, and X-Ray diffraction to demonstrate their suitability to thin-film solar technology.
scalable process↗
Scalable and Power Efficient Data Analytics for Hybrid Exascale Systems
This project developed a generic and optimized set of core data analytics functions. These functions organically consolidate a broad constellation of high performance analytical pipelines. As the architectures of emerging HPC systems become inherently heterogeneous, there is a need to design algorithms for data analysis kernels accelerated on hybrid multi-node, multi-core HPC architectures comprised of a mix of CPUs, GPUs, and SSDs. Furthermore, the power-aware trend drives the advances in our performance-energy tradeoff analysis framework which enables our data analysis kernels algorithms and software to be parameterized so that users can choose the right power-performance optimizations.
97 MATHEMATICS AND COMPUTING↗
Scalable Multilevel Uncertainty Quantification Concepts for Extreme-Scale Multiscale Problems
Our work in this project is aimed at making fundamental advances in multilevel and multiscale methods for problems with uncertainties related to flows in porous media. The main thrust of this research is to develop a systematic multiscale, multilevel parallel approaches for problems with uncertainties and use them in the uncertainty quantification. A key emphasis is on problems without an apparent scale separation in space, time, and uncertainty spaces. Multiscale solution methods are currently under active investigation. However, the develop- ment of multiscale methods for problems with uncertainties where the scales and uncertainties are tightly coupled is a research area that is less explored. During this project, we have made some significant contributions to the developments of multiscale methods, the uncertainty quantification for multiscale methods, multiscale model learning, uncertainty quantification in inverse problems, and applications.
97 MATHEMATICS AND COMPUTING↗
Qualitative Analysis, Structure-Preserving Discretizations and Scalable Solvers for Multiphysics Problems. Final Technical Report
Many problems in physics, engineering, meteorology, computer science, and other fields involve a certain level of coupling between different physical quantities. With the development of advanced computational modeling, simulation, and hardware, simulations for many important multiphysics/multiscale coupling effects are now more tractable. In this project, we have developed robust, efficient, and practical numerical algorithms for such simulation software that are capable of tackling these complex mathematical models in many fundamental ways such as stability, wellposedness, and consistency. We expect that such an integrated theoretical and algorithmic framework will have a substantial impact on numerous multiscale and multiphysics modeling and simulations.
71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
Scalable Fabrication of a Fibrous Amine-functionalized Matrix (FAM) Sorbent for Critical Mineral Recovery
We report a novel flat sheet Fibrous Amine-functionalized Matrix (FAM) sorbent platform designed for efficient and selective capture of CM from dilute solutions. The FAM sorbent features crosslinked amine microfilms coated onto/within a glass fiber matrix, providing fast mass transfer and excellent mechanical stability. Systematic batch and flow-through tests with FAM revealed rapid metal uptake kinetics and high capacity for representative species, achieving ~90 mg/g of Gallium, ~100 mg/g of Cobalt, and ~90 mg/g for Neodymium. Moreover, multiple eluents, including mineral acids and complexing agents, enabled highly effective desorption of adsorbed metals, demonstrating the feasibility of regenerating FAM sorbents. Importantly, tests with authentic coal ash leachate demonstrated strong selectivity toward U.S. Department of Energy (DOE)-listed CM and rare earth elements over abundant base cations, confirming the robustness of FAM in realistic complex solutions. The flat sheet geometry was amenable to scaling into durable spiral wound modules, highlighting the potential for future regeneration and reuse. This work establishes FAM sorbents as a promising platform for the recovery of CM from wastewaters, advancing both resource sustainability and environmental stewardship.
critical mineral recovery↗
Scalable foundation models for numerical simulations on HPC platforms
In recent years, foundation models (FMs) have begun to reshape numerical simulations on high-performance computing (HPC) platforms. These large, pre-trained AI models enable rapid predictions across a broad range of physical domains, including Earth system modeling, fluid dynamics, materials science, as well as complex multi-modal simulations in aerospace engineering and fusion research. By training on diverse datasets, FMs learn intricate relationships and underlying physical behavior while also enabling the quantification of uncertainty in their predictions. This capability allows simulations that once required days of numerical calculation to be completed in minutes (FM inference), supporting real-time design optimization, uncertainty-aware decision making, and more comprehensive exploration of complex scenarios.
AI↗