Search NASA⌕ Search

SEARCH · Search NASA

Results for “parallel computers clusters”

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 217 records · Page 12
All sourcesNASA NTRSDOE OSTI
NASA NTRS2013

Profiling and Improving I/O Performance of a Large-Scale Climate Scientific Application

Exascale computing systems are soon to emerge, which will pose great challenges on the huge gap between computing and I/O performance. Many large-scale scientific applications play an important role in our daily life. The huge amounts of data generated by such applications require highly parallel and efficient I/O management policies. In this paper, we adopt a mission-critical scientific application, GEOS-5, as a case to profile and analyze the communication and I/O issues that are preventing applications from fully utilizing the underlying parallel storage systems. Through in-detail architectural and experimental characterization, we observe that current legacy I/O schemes incur significant network communication overheads and are unable to fully parallelize the data access, thus degrading applications' I/O performance and scalability. To address these inefficiencies, we redesign its I/O framework along with a set of parallel I/O techniques to achieve high scalability and performance. Evaluation results on the NASA discover cluster show that our optimization of GEOS-5 with ADIOS has led to significant performance improvements compared to the original GEOS-5 implementation.

Exascale↗
NASA NTRS1995

Stereo-Based Region-Growing using String Matching

We present a novel stereo algorithm based on a coarse texture segmentation preprocessing phase. Matching is performed using a string comparison. Matching sub-strings correspond to matching sequences of textures. Inter-scanline clustering of matching sub-strings yields regions of matching texture. The shape of these regions yield information concerning object's height, width and azimuthal position relative to the camera pair. Hence, rather than the standard dense depth map, the output of this algorithm is a segmentation of objects in the scene. Such a format is useful for the integration of stereo with other sensor modalities on a mobile robotic platform. It is also useful for localization; the height and width of a detected object may be used for landmark recognition, while depth and relative azimuthal location determine pose. The algorithm does not rely on the monotonicity of order of image primitives. Occlusions, exposures, and foreshortening effects are not problematic. The algorithm can deal with certain types of transparencies. It is computationally efficient, and very amenable to parallel implementation. Further, the epipolar constraints may be relaxed to some small but significant degree. A version of the algorithm has been implemented and tested on various types of images. It performs best on random dot stereograms, on images with easily filtered backgrounds (as in synthetic images), and on real scenes with uncontrived backgrounds.

Mandelbaum, Robert↗
NASA NTRS2018

An Investigation of Parallel Programming Techniques Applied to Monte Carlo Simulations for Post-Flight Reconstruction of Spacecraft Trajectory

Parallelizing software to execute on multi-core central processing units (CPUs) and graphics processing units (GPUs) can be challenging. For some fields outside of Computer Science, this transition comes with new issues. For example, memory limitations can require modifications to code not initially developed to run on GPUs. This work applies the Open Multi-Processing (OpenMP) and Open Accelerators (OpenACC) directive-based parallelization strategies on a Monte Carlo simulation approach for trajectory reconstruction enabling it to run on multi-core CPUs and GPUs. Large matrix operations are the most common use of GPUs, which are not present in this algorithm; however, the natural parallelism of independent trajectories in Monte Carlo simulations is exploited. Benchmarking data are presented comparing execution times of the software for single-thread CPUs, multi-thread CPUs with OpenMP, and multi-thread GPUs using OpenACC. These data were collected using nodes with Intel® Xeon® E5-2670 (Sandy Bridge) CPUs enhanced with NVIDIA® Tesla® K40 GPUs on the Pleiades Supercomputer cluster at the National Aeronautics and Space Administration (NASA) Ames Research Center (ARC) and a local Intel® Xeon Phi™ node at NASA Langley Research Center (LaRC).

Williams, R. Anthony↗
DOE OSTI2025

ROOT RNTuple and EOS: The Next Generation of Event Data I/O

For several years, the ROOT team is developing the new RNTuple I/O subsystem in preparation of the next generation of collider experiments. Both HL-LHC and DUNE are expected to start data taking by the end of this decade. They pose unprecedented challenges to event data I/O in terms of data rates, event sizes, and event complexity. At the same time, the I/O landscape is becoming more diverse. HPC cluster file systems and object stores, NVMe disk cache layers in analysis facilities, and S3 storage on cloud resources are mixing with traditional XRootD-managed spinning disk pools.The ROOT team will finalize a first production version of the RNTuple binary format by the end of 2024. After this point, ROOT will provide backward compatibility for RNTuple data. This contribution provides an overview of the RNTuple feature set, the related R&D activities and the long-term vision for RNTuple. We report on performance, interface design, tooling, robustness, integration with experiment frameworks, and validation results, as well as recent R&D on parallel reading and writing and exploitation of modern hardware and storage systems. We will give an outlook on possible future features after a first production release.Collaboratively, the IT and EP departments at CERN have launched a formal project within the Research and Computing sector to evaluate the novel data format for physics analysis data utilized in LHC experiments and other fields. This part of the project focuses on validating the scalability of the EOS storage backend during the transition from the over 25 years old TTree production format to the newly developed RNTuple format, using both replicated and erasure-coded storage profiles.

Blomer, Jakob [CERN]↗
DOE OSTI2026

FitCache: A Transparent Drop-In Framework for Multi-Tier Caching to Accelerate Distributed Deep Learning Workloads

Training in Deep learning (DL) remains highly compute- and data-intensive, with I/O becoming a critical bottleneck as models and datasets scale. Recent studies report that data loading can dominate training time, especially on large-scale HPC systems with shared parallel file systems (PFS). Existing caching approaches either rely on single-tier designs or require intrusive modifications to training pipelines, limiting their portability and effectiveness. In this work, we present FitCache, a transparent drop-in framework for multi-tier caching to accelerate distributed DL training by coordinating fast local memory (e.g., DRAM, Persistent Memory (PMem)) and NVMe as hierarchical caches atop PFS. Our design adapts to hardware diversity, i.e., if NVMe is missing, memory transparently acts as a caching tier, ensuring stable performance. FitCache transparently intercepts I/O requests and issues concurrent fetches across all tiers, returning data from the fastest responder without centralized metadata or static redirection paths. FitCache adapts to dynamic workloads and heterogeneous clusters while maintaining POSIX compatibility. Experiments on Frontier (2048 GPUs) and smaller research clusters show that FitCache reduces training time by up to 40% and per-batch I/O latency by up to 71.6% compared to Lustre Orion PFS, offering a drop-in solution for scalable DL training.

Hu, Guangxing [ORNL] (ORCID:0009000283203614)↗
NASA NTRS2001

CFD Simulation of Liquid Rocket Engine Injectors

Detailed design issues associated with liquid rocket engine injectors and combustion chamber operation require CFD methodology which simulates highly three-dimensional, turbulent, vaporizing, and combusting flows. The primary utility of such simulations involves predicting multi-dimensional effects caused by specific injector configurations. SECA, Inc. and Engineering Sciences, Inc. have been developing appropriate computational methodology for NASA/MSFC for the past decade. CFD tools and computers have improved dramatically during this time period; however, the physical submodels used in these analyses must still remain relatively simple in order to produce useful results. Simulations of clustered coaxial and impinger injector elements for hydrogen and hydrocarbon fuels, which account for real fluid properties, is the immediate goal of this research. The spray combustion codes are based on the FDNS CFD code' and are structured to represent homogeneous and heterogeneous spray combustion. The homogeneous spray model treats the flow as a continuum of multi-phase, multicomponent fluids which move without thermal or velocity lags between the phases. Two heterogeneous models were developed: (1) a volume-of-fluid (VOF) model which represents the liquid core of coaxial or impinger jets and their atomization and vaporization, and (2) a Blob model which represents the injected streams as a cloud of droplets the size of the injector orifice which subsequently exhibit particle interaction, vaporization, and combustion. All of these spray models are computationally intensive, but this is unavoidable to accurately account for the complex physics and combustion which is to be predicted, Work is currently in progress to parallelize these codes to improve their computational efficiency. These spray combustion codes were used to simulate the three test cases which are the subject of the 2nd International Workshop on-Rocket Combustion Modeling. Such test cases are considered by these investigators to be very valuable for code validation because combustion kinetics, turbulence models and atomization models based on low pressure experiments of hydrogen air combustion do not adequately verify analytical or CFD submodels which are necessary to simulate rocket engine combustion. We wish to emphasize that the simulations which we prepared for this meeting are meant to test the accuracy of the approximations used in our general purpose spray combustion models, rather than represent a definitive analysis of each of the experiments which were conducted. Our goal is to accurately predict local temperatures and mixture ratios in rocket engines; hence predicting individual experiments is used only for code validation. To replace the conventional JANNAF standard axisymmetric finite-rate (TDK) computer code 2 for performance prediction with CFD cases, such codes must posses two features. Firstly, they must be as easy to use and of comparable run times for conventional performance predictions. Secondly, they must provide more detailed predictions of the flowfields near the injector face. Specifically, they must accurately predict the convective mixing of injected liquid propellants in terms of the injector element configurations.

Farmer, Richard↗
DOE OSTI2026

Symbol alphabets in QCD and flag cluster algebras

The full 245-letter symbol alphabet for all planar massless two-loop six-point Feynman integrals was recently determined in arXiv:2412.19884 and arXiv:2501.01847. In a parallel mathematical development, it was shown in arXiv:2408.14956 that there is an embedding of the cluster algebra associated to the partial flag variety $\mathcal{Fl}$ $2,n-2;n$ , which describes the kinematics of n massless particles, into that of the Grassmannian Gr(n–2, 2n–4). In this paper we connect these developments by showing that most of the rational symbol letters can be expressed in terms of flag cluster variables, and that all of the algebraic symbol letters arise from infinite mutation sequences.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

ArborX 2.0

ArborX library tackles a problem of efficiently finding geometric objects that are close in space. Variations of this problem, such as finding the nearest neighbors of a point, or finding all objects within a certain distance, are inherent components of applications in many fields. The data may be large so that solving the problem efficiently may require significant computational resources, such as multiple processors or accelerators such as general purpose GPUs. ArborX' main advantage in its ability to solve large problems efficiently utilizing a combination of distributed and on-node parallelism. ArborX can be run efficiently on a wide variety of hardware, including GPUs from different vendors, which distinguishes it from other available libraries which typically choose only few of these. The other advantage is that it supports both types of user problems: spatial problems (useful for intersections and finding objects within certain distance), and nearest neighbor problems. ArborX also supports flexible interface in its interaction with a user. Particularly, it allows a user to call user's own function on a positive match, a functionality not rarely available in other libraries. ArborX implements construction and traversal algorithms using efficient tree structures, such as bounding volume hierarchy (BVH). At its core, ArborX uses linear BVH for its low construction cost and sufficient quality. ArborX implements both spatial and nearest-neighbor traversal algorithms. ArborX also provides several clustering algorithms (minimum spanning tree, DBSCAN, HDBSCAN*), interpolation using minimum least squares and ray tracing. ArborX is written using C++, and is parallelized using the message passing interface (MPI) for the distributed communication, and the Kokkos library for on-node parallelism. This approach allows ArborX to be run on a wide variety of hardware, from common laptops and desktops to supercomputers while using the same codebase.

Prokopenko, Andrey [Oak Ridge National Laboratory ↗
NASA NTRS1990

Adaptive pattern recognition by mini-max neural networks as a part of an intelligent processor

In this decade and progressing into 21st Century, NASA will have missions including Space Station and the Earth related Planet Sciences. To support these missions, a high degree of sophistication in machine automation and an increasing amount of data processing throughput rate are necessary. Meeting these challenges requires intelligent machines, designed to support the necessary automations in a remote space and hazardous environment. There are two approaches to designing these intelligent machines. One of these is the knowledge-based expert system approach, namely AI. The other is a non-rule approach based on parallel and distributed computing for adaptive fault-tolerances, namely Neural or Natural Intelligence (NI). The union of AI and NI is the solution to the problem stated above. The NI segment of this unit extracts features automatically by applying Cauchy simulated annealing to a mini-max cost energy function. The feature discovered by NI can then be passed to the AI system for future processing, and vice versa. This passing increases reliability, for AI can follow the NI formulated algorithm exactly, and can provide the context knowledge base as the constraints of neurocomputing. The mini-max cost function that solves the unknown feature can furthermore give us a top-down architectural design of neural networks by means of Taylor series expansion of the cost function. A typical mini-max cost function consists of the sample variance of each class in the numerator, and separation of the center of each class in the denominator. Thus, when the total cost energy is minimized, the conflicting goals of intraclass clustering and interclass segregation are achieved simultaneously.

Szu, Harold H.↗
NASA NTRS1983

Parallel architectures for iterative methods on adaptive, block structured grids

A parallel computer architecture well suited to the solution of partial differential equations in complicated geometries is proposed. Algorithms for partial differential equations contain a great deal of parallelism. But this parallelism can be difficult to exploit, particularly on complex problems. One approach to extraction of this parallelism is the use of special purpose architectures tuned to a given problem class. The architecture proposed here is tuned to boundary value problems on complex domains. An adaptive elliptic algorithm which maps effectively onto the proposed architecture is considered in detail. Two levels of parallelism are exploited by the proposed architecture. First, by making use of the freedom one has in grid generation, one can construct grids which are locally regular, permitting a one to one mapping of grids to systolic style processor arrays, at least over small regions. All local parallelism can be extracted by this approach. Second, though there may be a regular global structure to the grids constructed, there will be parallelism at this level. One approach to finding and exploiting this parallelism is to use an architecture having a number of processor clusters connected by a switching network. The use of such a network creates a highly flexible architecture which automatically configures to the problem being solved.

Gannon, D.↗
DOE OSTI2025

Using Apptainer in a Pilot-based Distributed Workload

GlideinWMS is a pilot and pressure-based workload manager for distributed scientific computing. Many experiments like CMS and Fermilab’s Neutrino experiments use it to provision elastic clusters for their analysis and simulations, split into close to a million concurrent jobs. Most user jobs require containers, and the pilots use Apptainer to set up the desired platform. For the pilots that run as regular batch jobs, Apptainer is safer, lighter, and easier to use than other containerization solutions. Many images used by the pilots are expanded SIF images distributed via the CernVM-FS: this combination is very efficient. At Fermilab, for example, we store on GitHub Dockerfiles that mimic the platform in the worker nodes of local clusters. GitHub workflows build and push the images to Docker Hub, and a service periodically pulls and converts them to the expanded SIF images in the CernVM-FS, so the scientists can find a familiar environment everywhere. Apptainer has also been used to run services inside the pilot jobs, like benchmarks that characterize the worker node being used, or a Triton Inference Server that allows sharing a GPU with all the jobs that run in parallel on a node.

Mambelli, Marco [Fermilab] (ORCID:0000000294892681↗
NASA NTRS2006

Forecasting of Storm-Surge Floods Using ADCIRC and Optimized DEMs

Increasing the accuracy of storm-surge flood forecasts is essential for improving preparedness for hurricanes and other severe storms and, in particular, for optimizing evacuation scenarios. An interactive database, developed by WorldWinds, Inc., contains atlases of storm-surge flood levels for the Louisiana/Mississippi gulf coast region. These atlases were developed to improve forecasting of flooding along the coastline and estuaries and in adjacent inland areas. Storm-surge heights depend on a complex interaction of several factors, including: storm size, central minimum pressure, forward speed of motion, bottom topography near the point of landfall, astronomical tides, and, most importantly, maximum wind speed. The information in the atlases was generated in over 100 computational simulations, partly by use of a parallel-processing version of the ADvanced CIRCulation (ADCIRC) model. ADCIRC is a nonlinear computational model of hydrodynamics, developed by the U.S. Army Corps of Engineers and the US Navy, as a family of two- and three-dimensional finite-element-based codes. It affords a capability for simulating tidal circulation and storm-surge propagation over very large computational domains, while simultaneously providing high-resolution output in areas of complex shoreline and bathymetry. The ADCIRC finite-element grid for this project covered the Gulf of Mexico and contiguous basins, extending into the deep Atlantic Ocean with progressively higher resolution approaching the study area. The advantage of using ADCIRC over other storm-surge models, such as SLOSH, is that input conditions can include all or part of wind stress, tides, wave stress, and river discharge, which serve to make the model output more accurate. To keep the computational load manageable, this work was conducted using only the wind stress, calculated by using historical data from Hurricane Camille, as the input condition for the model. Hurricane storm-surge simulations were performed on an eight-node Linux computer cluster. Each node contained dual 2-GHz processors, 2GB of memory, and a 40GB hard drive. The digital elevation model (DEM) for this region was specified using a combination of Navy data (over water), NOAA data (for the coastline), and optimized Interferometric Synthetic Aperture Radar data (over land). This high-resolution topographical data of the Mississippi coastal region provided the ADCIRC model with improved input with which to calculate improved storm-surge forecasts.

Valenti, Elizabeth↗
NASA NTRS1994

MPIRUN: A Portable Loader for Multidisciplinary and Multi-Zonal Applications

Multidisciplinary and multi-zonal applications are an important class of applications in the area of Computational Aerosciences. In these codes, two or more distinct parallel programs or copies of a single program are utilized to model a single problem. To support such applications, it is common to use a programming model where a program is divided into several single program multiple data stream (SPMD) applications, each of which solves the equations for a single physical discipline or grid zone. These SPMD applications are then bound together to form a single multidisciplinary or multi-zonal program in which the constituent parts communicate via point-to-point message passing routines. One method for implementing the message passing portion of these codes is with the new Message Passing Interface (MPI) standard. Unfortunately, this standard only specifies the message passing portion of an application, but does not specify any portable mechanisms for loading an application. MPIRUN was developed to provide a portable means for loading MPI programs, and was specifically targeted at multidisciplinary and multi-zonal applications. Programs using MPIRUN for loading and MPI for message passing are then portable between all machines supported by MPIRUN. MPIRUN is currently implemented for the Intel iPSC/860, TMC CM5, IBM SP-1 and SP-2, Intel Paragon, and workstation clusters. Further, MPIRUN is designed to be simple enough to port easily to any system supporting MPI.

Fineberg, Samuel A.↗
NASA NTRS2016

DeepSAT: A Deep Learning Approach to Tree-Cover Delineation in 1-m NAIP Imagery for the Continental United States

High resolution tree cover classification maps are needed to increase the accuracy of current land ecosystem and climate model outputs. Limited studies are in place that demonstrates the state-of-the-art in deriving very high resolution (VHR) tree cover products. In addition, most methods heavily rely on commercial softwares that are difficult to scale given the region of study (e.g. continents to globe). Complexities in present approaches relate to (a) scalability of the algorithm, (b) large image data processing (compute and memory intensive), (c) computational cost, (d) massively parallel architecture, and (e) machine learning automation. In addition, VHR satellite datasets are of the order of terabytes and features extracted from these datasets are of the order of petabytes. In our present study, we have acquired the National Agriculture Imagery Program (NAIP) dataset for the Continental United States at a spatial resolution of 1-m. This data comes as image tiles (a total of quarter million image scenes with ~60 million pixels) and has a total size of ~65 terabytes for a single acquisition. Features extracted from the entire dataset would amount to ~8-10 petabytes. In our proposed approach, we have implemented a novel semi-automated machine learning algorithm rooted on the principles of "deep learning" to delineate the percentage of tree cover. Using the NASA Earth Exchange (NEX) initiative, we have developed an end-to-end architecture by integrating a segmentation module based on Statistical Region Merging, a classification algorithm using Deep Belief Network and a structured prediction algorithm using Conditional Random Fields to integrate the results from the segmentation and classification modules to create per-pixel class labels. The training process is scaled up using the power of GPUs and the prediction is scaled to quarter million NAIP tiles spanning the whole of Continental United States using the NEX HPC supercomputing cluster. An initial pilot over the state of California spanning a total of 11,095 NAIP tiles covering a total geographical area of 163,696 sq. miles has produced true positive rates of around 88 percent for fragmented forests and 74 percent for urban tree cover areas, with false positive rates lower than 2 percent for both landscapes.

Imagery↗
NASA NTRS2001

Efficiency Improvements to the Displacement Based Multilevel Structural Optimization Algorithm

Multilevel Structural Optimization (MSO) continues to be an area of research interest in engineering optimization. In the present project, the weight optimization of beams and trusses using Displacement based Multilevel Structural Optimization (DMSO), a member of the MSO set of methodologies, is investigated. In the DMSO approach, the optimization task is subdivided into a single system and multiple subsystems level optimizations. The system level optimization minimizes the load unbalance resulting from the use of displacement functions to approximate the structural displacements. The function coefficients are then the design variables. Alternately, the system level optimization can be solved using the displacements themselves as design variables, as was shown in previous research. Both approaches ensure that the calculated loads match the applied loads. In the subsystems level, the weight of the structure is minimized using the element dimensions as design variables. The approach is expected to be very efficient for large structures, since parallel computing can be utilized in the different levels of the problem. In this paper, the method is applied to a one-dimensional beam and a large three-dimensional truss. The beam was tested to study possible simplifications to the system level optimization. In previous research, polynomials were used to approximate the global nodal displacements. The number of coefficients of the polynomials equally matched the number of degrees of freedom of the problem. Here it was desired to see if it is possible to only match a subset of the degrees of freedom in the system level. This would lead to a simplification of the system level, with a resulting increase in overall efficiency. However, the methods tested for this type of system level simplification did not yield positive results. The large truss was utilized to test further improvements in the efficiency of DMSO. In previous work, parallel processing was applied to the subsystems level, where the derivative verification feature of the optimizer NPSOL had been utilized in the optimizations. This resulted in large runtimes. In this paper, the optimizations were repeated without using the derivative verification, and the results are compared to those from the previous work. Also, the optimizations were run on both, a network of SUN workstations using the MPICH implementation of the Message Passing Interface (MPI) and on the faster Beowulf cluster at ICASE, NASA Langley Research Center, using the LAM implementation of UP]. The results on both systems were consistent and showed that it is not necessary to verify the derivatives and that this gives a large increase in efficiency of the DMSO algorithm.

Plunkett, C. L.↗
NASA NTRS2008

Experimental Measurement of RCS Jet Interaction Effects on a Capsule Entry Vehicle

An investigation was made in NASA Langley Research Center s 31-Inch Mach 10 Tunnel to determine the effects of reaction-control system (RCS) jet interactions on the aft-body of a capsule entry vehicle. The test focused on demonstrating and improving advanced measurement techniques that would aid in the rapid measurement and visualization of jet interaction effects for the Orion Crew Exploration Vehicle while providing data useful for developing engineering models or validation of computational tools used to assess actual flight environments. Measurements included global surface imaging with pressure and temperature sensitive paints and three-dimensional flow visualization with a scanning planar laser induced fluorescence technique. The wind tunnel model was fabricated with interchangeable parts for two different aft-body configurations. The first, an Apollo-like configuration, was used to focus primarily on the forward facing roll and yaw jet interactions which are known to have significant aft-body heating augmentation. The second, an early Orion Crew Module configuration (4-cluster jets), was tested blowing only out of the most windward yaw jet, which was expected to have the maximum heating augmentation for that configuration. Jet chamber pressures and tunnel flow conditions were chosen to approximate early Apollo wind tunnel test conditions. Maximum heating augmentation values measured for the Apollo-like configuration (>10 for forward facing roll jet and 4 for yaw jet) using temperature sensitive paint were shown to be similar to earlier experimental results (Jones and Hunt, 1965) using a phase change paint technique, but were acquired with much higher surface resolution. Heating results for the windward yaw jet on the Orion configuration had similar augmentation levels, but affected much less surface area. Numerical modeling for the Apollo-like yaw jet configuration with laminar flow and uniform jet outflow conditions showed similar heating patterns, qualitatively, but also showed significant variation with jet exit divergence angle, with as much as 25 percent variation in heat flux intensity for a 10 degree divergence angle versus parallel outflow. These results along with the fabrication methods and advanced measurement techniques developed will be used in the next phase of testing and evaluation for the updated Orion RCS configuration.

Buck, Gregory M.↗
NASA NTRS2019

Accelerated Simulation of Air Pollution Using NVIDIA RAPIDS

Atmospheric chemistry models are a central tool to study and forecast the impact of air pollution on the environment, vegetation, and human health. However, the numerical simulation of chemical kinetics is computationally expensive due to the stiffness of the system of ordinary differential equations that describes atmospheric chemistry. Here we present an alternative approach to the computation of atmospheric chemistry based on machine learning. Our training data set is produced using the NASA Goddard Earth Observing System (GEOS) model with GEOS-Chem chemistry, run on the NASA Center for Climate Simulation (NCCS) Discover supercomputing cluster on 384 Intel Xeon Haswell cores. This model spends more than 50% of total run time on solving atmospheric chemistry. The data set contains as input features the air pollution concentrations before solving the differential equations, together with some key physical parameters such as temperature and sun intensity. As target variables we define the air pollution concentrations after solving the differential equations. Using Dask-cuDF and Dask-XGBoost on the NVIDIA RAPIDS platform on 8 Tesla V100 GPUs, we generate from this training set gradient boosted decision tree models that can reproduce the simulation of chemical kinetics. We do this on the NCCS Advanced Data Analytics Platform (ADAPT) science cloud environment. Our application takes full advantage of recent advances in Dask-XGBoost, such as multi-node and multi-GPU scaling for distributed training with large data sets. The increase in training data size enabled by this is critical to capture the full range of chemical environments encountered across the globe and all annual seasons.The boosted tree models offer good predictability and show many of the features of the full chemistry reference simulation. Further improvements can be achieved through mass balance considerations and by accounting for error correlations. We incorporate the boosted tree models into the GEOS reference model using XGBoost's C API. This enables a seamless integration of the GPU trained models into GEOS-Chem, which is written in Fortran and optimized for use in a massively parallel CPU environment. We show the benefits of this approach and discuss the potential speedup of this machine learning accelerated atmospheric chemistry model.

Keller, Christoph A.↗
DOE OSTI2026

Spatially Accelerated Winding Numbers for Curved Geometry

The generalized winding number (GWN) is a scalar field that supports robust containment queries on curved geometry, including non-watertight, overlapping, and nested boundary representations. While queries can be easily parallelized over samples, direct evaluation on parametric curves and surfaces remains costly for large and complex models. Fast, state-of-the-art GWN approaches leverage a spatial index to approximate the GWN, typically coupled with a Taylor expansion which approximates the GWN contribution for far clusters of geometric primitives. However, such methods operate only on discrete inputs such as triangle meshes and point clouds, and would introduce containment errors near boundaries if applied to curved input. We extend support for fast GWN evaluation over arbitrary collections of NURBS curves in 2D and trimmed NURBS patches in 3D via a Bounding Volume Hierarchy that stores efficiently precomputed moment data in the hierarchy nodes. When querying the hierarchy, approximations for far clusters are used alongside direct evaluation for nearby NURBS primitives, achieving sub-linear complexity while preserving the geometric features in the vicinity of the query point. Central to our performance improvements is an adaptive subdivision strategy for NURBS primitives during a preprocessing phase, creating better spatial partitions while retaining the same accuracy for containment decisions as a direct evaluation. We demonstrate the performance and accuracy of our approach across a large collection of 2D and 3D datasets.

Computer science↗