SEARCH · Engineering Papers

Results for “high performance and distributed computing”

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 829 records · Page 46

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Sub-microsecond Transformers for Jet Tagging on FPGAs

We present the first sub-microsecond transformer implementation on an FPGA achieving competitive performance for state-of-the-art high-energy physics benchmarks. Transformers have shown exceptional performance on multiple tasks in modern machine learning applications, including jet tagging at the CERN Large Hadron Collider (LHC). However, their computational complexity prohibits use in real-time applications, such as the hardware trigger system of the collider experiments up until now. In this work, we demonstrate the first application of transformers for jet tagging on FPGAs, achieving $\mathcal{O}(100)$ nanosecond latency with superior performance compared to alternative baseline models. We leverage high-granularity quantization and distributed arithmetic optimization to fit the entire transformer model on a single FPGA, achieving the required throughput and latency. Furthermore, we add multi-head attention and linear attention support to hls4ml, making our work accessible to the broader fast machine learning community. This work advances the next-generation trigger systems for the High Luminosity LHC, enabling the use of transformers for real-time applications in high-energy physics and beyond.

Laatu, Lauri [Imperial Coll., London]↗

DOE OSTI2023

Reinforcement Learning for Load-balanced Parallel Particle Tracing

We explore an online reinforcement learning (RL) paradigm to dynamically optimize parallel particle tracing performance in distributed-memory systems. Our method combines three novel components: (1) a work donation algorithm, (2) a high-order workload estimation model, and (3) a communication cost model. First, we design an RL-based work donation algorithm. Our algorithm monitors workloads of processes and creates RL agents to donate data blocks and particles from high-workload processes to low-workload processes to minimize program execution time. The agents learn the donation strategy on the fly based on reward and cost functions designed to consider processes' workload changes and data transfer costs of donation actions. Second, we propose a workload estimation model, helping RL agents estimate the workload distribution of processes in future computations. Third, we design a communication cost model that considers both block and particle data exchange costs, helping RL agents make effective decisions with minimized communication costs. We demonstrate that our algorithm adapts to different flow behaviors in large-scale fluid dynamics, ocean, and weather simulation data. Our algorithm improves parallel particle tracing performance in terms of parallel efficiency, load balance, and costs of I/O and communication for evaluations with up to 16,384 processors.

Distributed and parallel particle tracing↗

DOE OSTI2021

A Modular and Transferable Reinforcement Learning Framework for the Fleet Rebalancing Problem

Mobility on demand (MoD) systems show great promise in realizing flexible and efficient urban transportation. However, significant technical challenges arise from operational decision making associated with MoD vehicle dispatch and fleet rebalancing. For this reason, operators tend to employ simplified algorithms that have been demonstrated to work well in a particular setting. To help bridge the gap between novel and existing methods, we propose a modular framework for fleet rebalancing based on model-free reinforcement learning (RL) that can leverage an existing dispatch method to minimize system cost. In particular, by treating dispatch as part of the environment dynamics, a centralized agent can learn to intermittently direct the dispatcher to reposition free vehicles and mitigate against fleet imbalance. We formulate RL state and action spaces as distributions over a grid partitioning of the operating area, making the framework scalable and avoiding the complexities associated with multiagent RL. Numerical experiments, using real-world trip and network data, demonstrate that RL reduces waiting time by 28% to 38% for the same-day evaluation, 17% to 44% for cross-day evaluation, and 22% to 25% for cross-season evaluation compared with no rebalancing scenarios. This approach has several distinct advantages over baseline methods including: improved system cost; high degree of adaptability to the selected dispatch method; and the ability to perform scale-invariant transfer learning between problem instances with similar vehicle and request distributions.

33 ADVANCED PROPULSION SYSTEMS↗

DOE OSTI2026

Neural operator transformers capture bifurcating drift-wave turbulence in fusion plasma simulations

Self-consistent modeling of turbulence-driven transport is critical for optimizing confinement in magnetically confined fusion plasmas, such as tokamaks and stellarators. In particular, capturing the long-term co-evolution of turbulence, flow, and background plasma profiles remains computationally challenging. Direct numerical simulation of these multiscale, highly nonlinear processes is often demanding and impractical for real-time control or design optimization. To address this bottleneck, we investigate transformer-based neural operator partial differential equation surrogates for emulating the dynamics of drift-wave turbulence bifurcation mediated by zonal flows, using the modified Hasegawa–Wakatani (MHW) model as a prototypical system. We find that the finetuned neural operator model has excellent performance in capturing the multi-spatiotemporal-scales of MHW turbulence bifurcation and is robust to testing on rare and out-of-distribution dynamics. Specifically, we demonstrate that a single unified model accurately predicts both quasi-steady-state turbulence and a wide range of dynamical transition processes, such as nonlinear saturation, spontaneous suppression of turbulence, and the emergence of macroscopic zonal flows, over time horizons vastly exceeding the local turbulence correlation time. This computationally efficient approach establishes a strong foundation for fast, AI-based modeling of complex, multiscale phenomena in magnetized fusion plasmas.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2016

A generalized multi-resolution expansion for uncertainty propagation with application to cardiovascular modeling

Computational models are used in a variety of fields to improve our understanding of complex physical phenomena. Recently, the realism of model predictions has been greatly enhanced by transitioning from deterministic to stochastic frameworks, where the effects of the intrinsic variability in parameters, loads, constitutive properties, model geometry and other quantities can be more naturally included. A general stochastic system may be characterized by a large number of arbitrarily distributed and correlated random inputs, and a limited support response with sharp gradients or event discontinuities. This motivates continued research into novel adaptive algorithms for uncertainty propagation, particularly those handling high dimensional, arbitrarily distributed random inputs and non-smooth stochastic responses. Here, we generalize a previously proposed multi-resolution approach to uncertainty propagation to develop a method that improves computational efficiency, can handle arbitrarily distributed random inputs and non-smooth stochastic responses, and naturally facilitates adaptivity, i.e., the expansion coefficients encode information on solution refinement. Our approach relies on partitioning the stochastic space into elements that are subdivided along a single dimension, or, in other words, progressive refinements exhibiting a binary tree representation. We also show how these binary refinements are particularly effective in avoiding the exponential increase in the multi-resolution basis cardinality and significantly reduce the regression complexity for moderate to high dimensional random inputs. The performance of the approach is demonstrated through previously proposed uncertainty propagation benchmarks and stochastic multi-scale finite element simulations in cardiovascular flow.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2026

Computational Thermal Hydraulics of a High-Performance Low-Enriched-Uranium Annular Target for HFIR Irradiation

Molybdenum-99 has historically been generated via isolation from fissioned highly enriched uranium (HEU) targets. Here, this isotope is in high demand due to its daily use across the world in radiopharmaceutical medical procedures. The primary objective of this work was to design and analyze an experimental target assembly containing one low-enriched-uranium (LEU) annular target for irradiation at the High Flux Isotope Reactor (HFIR). Efforts included incorporating spatially dependent energy sources from neutron and gamma interactions, quantifying thermal contact conductance at material interfaces, performing grid-independent studies, comparing turbulence models, and simulating various steady-state and transient scenarios relevant for irradiation qualification and eventual insertion. These models provide velocity, pressure, and temperature distributions in both space and time. Such results enable the selection of an appropriate irradiation location, fission rate density, and flow-limiting orifice size and demonstrate compliance with HFIR safety requirements such that insertion into the reactor can be approved. This analysis shows that across all scenarios, wetted surface temperatures remain below the coolant saturation temperature with no net vapor formation in the coolant. In every scenario, all components stay below 30% of the aluminum 6061 melting temperature. Computational fluid dynamics and system-level models predict peak target temperatures that agree within 4%, though the predicted axial location of the peak differs by about 10% of the heated length due to differences in flow development length. These results de-risk the irradiation of LEU (annular targets) and strengthen a domestic, HEU-independent 99 Mo supply by providing important fuel performance data to form the foundation for a robust licensing basis.

Molybdenum-99↗

DOE OSTI2021

A machine learning approach to galaxy properties: joint redshift–stellar mass probability distributions with Random Forest

We demonstrate that highly accurate joint redshift–stellar mass probability distribution functions (PDFs) can be obtained using the Random Forest (RF) machine learning (ML) algorithm, even with few photometric bands available. As an example, we use the Dark Energy Survey (DES), combined with the COSMOS2015 catalogue for redshifts and stellar masses. We build two ML models: one containing deep photometry in the griz bands, and the second reflecting the photometric scatter present in the main DES survey, with carefully constructed representative training data in each case. We validate our joint PDFs for 10 699 test galaxies by utilizing the copula probability integral transform and the Kendall distribution function, and their univariate counterparts to validate the marginals. Benchmarked against a basic set-up of the template-fitting code bagpipes, our ML-based method outperforms template fitting on all of our predefined performance metrics. In addition to accuracy, the RF is extremely fast, able to compute joint PDFs for a million galaxies in just under 6 min with consumer computer hardware. Such speed enables PDFs to be derived in real time within analysis codes, solving potential storage issues. As part of this work we have developed galpro 1, a highly intuitive and efficient python package to rapidly generate multivariate PDFs on-the-fly. galpro is documented and available for researchers to use in their cosmology and galaxy evolution studies.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2019

CosmoFlow: Using Deep Learning to Learn the Universe at Scale

Deep learning is a promising tool to determine the physical model that describes our universe. To handle the considerable computational cost of this problem, we present CosmoFlow: a highly scalable deep learning application built on top of the TensorFlow framework. CosmoFlow uses efficient implementations of 3D convolution and pooling primitives, together with improvements in threading for many element-wise operations, to improve training performance on Intel ® Xeon Phi™ processors. We also utilize the Cray PE Machine Learning Plugin for efficient scaling to multiple nodes. We demonstrate fully synchronous data-parallel training on 8192 nodes of Cori with 77% parallel efficiency, achieving 3.5 Pflop/s sustained performance. To our knowledge, this is the first large-scale science application of the TensorFlow framework at supercomputer scale with fully-synchronous training. Here, these enhancements enable us to process large 3D dark matter distribution and predict the cosmological parameters Ω M , σ 8 and n s with unprecedented accuracy.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

DOE OSTI2024

Uncertainty quantification for molecular property predictions with graph neural architecture search

Graph Neural Networks (GNNs) have emerged as a prominent class of data-driven methods for molecular property prediction. However, a key limitation of typical GNN models is their inability to quantify uncertainties in the predictions. This capability is crucial for ensuring the trustworthy use and deployment of models in downstream tasks. To that end, we introduce AutoGNNUQ, an automated uncertainty quantification (UQ) approach for molecular property prediction. AutoGNNUQ leverages architecture search to generate an ensemble of high-performing GNNs, enabling the estimation of predictive uncertainties. Our approach employs variance decomposition to separate data (aleatoric) and model (epistemic) uncertainties, providing valuable insights for reducing them. In our computational experiments, we demonstrate that AutoGNNUQ outperforms existing UQ methods in terms of both prediction accuracy and UQ performance on multiple benchmark datasets, and generalizes well to out-of-distribution datasets. Additionally, we utilize t-SNE visualization to explore correlations between molecular features and uncertainty, offering insight for dataset improvement. AutoGNNUQ has broad applicability in domains such as drug discovery and materials science, where accurate uncertainty quantification is crucial for decision-making.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2021

HiPACE++ v0.1

Hipace++ is the new implementation of the DESY-LBNL code Hipace (not distributed nor licensed) for plasma acceleration, relying on the quasi-static particle-in-cell (PIC) method, enabling fast and accurate simulations of plasma accelerators for problems satisfying the quasi-static approximation. Plasma acceleration is a priority in the ATAP division, and such problems represent a wide range of parameters. Hipace++ is an HPC (high-performance computing) code running on supercomputers and it provides the first efficient implementation of the quasi-static PIC method on Graphic Processing Units (GPU), showing 10x speedup and using much less resources than CPU implementations for production simulations. It currently runs on Nvidia GPUs and will ultimately be portable to more accelerated HPC platforms. The porting strategy is novel, and can be adapted to other flavors of quasi-static PIC. Besides, Hipace++ aims at using mesh refinement for higher efficiency, thus enabling simulations unfeasible today. For this purpose, Hipace++ depends on the open-source library AMReX (Adaptive Mesh Refinement at EXascale) developed at LBNL.

Vay, Jean-Luc↗

DOE OSTI2025

Size-Resolved Chemical Composition of Particles Collected Using STAC at the Ground Site During the SAIL Campaign in Gunnison, Colorado

Aerosol particles were collected using a four-stage Size and Time-resolved Aerosol Collector (STAC) during the SAIL field campaign. Each stage of STAC separates particles into distinct aerodynamic size fractions with 50% cut-off diameters: Stage A: 2.27 µm Stage B: 0.615 µm Stage C: 0.421 µm Stage D: 0.119 µm Each stage provides both size- and time-resolved sampling, enabling investigation of particle composition across different atmospheric regimes. Only a subset of samples was selected for analysis based on prevailing meteorological conditions (e.g., temperature, humidity, and air-mass influence) to capture representative aerosol types under distinct weather patterns. Collected substrates were first examined under Scanning Electron Microscopy (SEM) to evaluate particle loading, morphology, and spatial distribution. Subsequently, Computer-Controlled Scanning Electron Microscopy with Energy-Dispersive X-ray Spectroscopy (CCSEM/EDX) was performed to obtain size-resolved elemental composition of individual particles. A rule-based classification scheme was applied to categorize particles into major compositional groups (e.g., biological, carbonaceous, dust, sulfate, Na-rich, and mixed types). This dataset provides high-resolution morphological and chemical information on atmospheric particles collected during the SAIL campaign, offering insights into the influence of meteorology on aerosol composition and mixing state.

Size and Time-resolved Aerosol Collector↗

NASA NTRS1987

Transport of passive scalars in a turbulent channel flow

A direct numerical simulation of a turbulent channel flow with three passive scalars at different molecular Prandtl numbers is performed. Computed statistics including the turbulent Prandtl numbers are compared with existing experimental data. The computed fields are also examined to investigate the spatial structure of the scalar fields. The scalar fields are highly correlated with the streamwise velocity; the correlation coefficient between the temperature and the streamwise velocity is as high as 0.95 in the wall region. The joint probability distributions between the temperature and velocity fluctuations are also examined; they suggest that it might be possible to model the scalar fluxes in the wall region in a manner similar to the Reynolds stresses.

Kim, John↗

NASA NTRS1987

Transport of passive scalars in a turbulent channel flow

Kim, John↗

DOE OSTI2015

Performance analysis of data intensive cloud systems based on data management and replication: a survey

As we delve deeper into the ‘Digital Age’, we witness an explosive growth in the volume, velocity, and variety of the data available on the Internet. For example, in 2012 about 2.5 quintillion bytes of data was created on a daily basis that originated from myriad of sources and applications including mobiledevices, sensors, individual archives, social networks, Internet of Things, enterprises, cameras, software logs, etc. Such ‘Data Explosions’ has led to one of the most challenging research issues of the current Information and Communication Technology era: how to optimally manage (e.g., store, replicated, filter, and the like) such large amount of data and identify new ways to analyze large amounts of data for unlocking information. It is clear that such large data streams cannot be managed by setting up on-premises enterprise database systems as it leads to a large up-front cost in buying and administering the hardware and software systems. Therefore, next generation data management systems must be deployed on cloud. The cloud computing paradigm provides scalable and elastic resources, such as data and services accessible over the Internet Every Cloud Service Provider must assure that data is efficiently processed and distributed in a way that does not compromise end-users’ Quality of Service (QoS) in terms of data availability, data search delay, data analysis delay, and the like. In the aforementioned perspective, data replication is used in the cloud for improving the performance (e.g., read and write delay) of applications that access data. Through replication a data intensive application or system can achieve high availability, better fault tolerance, and data recovery. In this paper, we survey data management and replication approaches (from 2007 to 2011) that are developed by both industrial and research communities. The focus of the survey is to discuss and characterize the existing approaches of data replication and management that tackle the resource usage and QoS provisioning with different levels of efficiencies. Moreover, the breakdown of both influential expressions (data replication and management) to provide different QoS attributes is deliberated. Furthermore, the performance advantages and disadvantages of data replication and management approaches in the cloud computing environments are analyzed. Open issues and future challenges related to data consistency, scalability, load balancing, processing and placement are also reported.

Malik, Saif Ur Rehman↗

DOE OSTI2023

Fourier-informed knot placement schemes for B-spline approximation

Fitting B-splines to scientific data is especially challenging when the given data contain noise, jumps, or corners. Here, we describe how periodic data sets with these features can be efficiently approximated with B-splines by analyzing the Fourier spectrum of the data. Our method uses a collection of spectral filters to compute high-order derivatives, smoothed versions of noisy data, and the locations of jump discontinuities. Further, these quantities are then combined to choose knots that capture the qualitative features of the data, leading to accurate B-spline approximations with few knots. The method we introduce is direct and does not require any intermediate B-spline fitting before choosing the final knot distribution. Aside from fast Fourier transforms to transfer to and from Fourier space, the method runs in linear time with very little communication. We assess performance on several test cases in one and two dimensions, including data sets with jump discontinuities and noise. These tests show the method fits discontinuous data without spurious oscillations and remains effective in the presence of noise.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Bayesian model updating with finite element vs surrogate models: Application to a miter gate structural system

Bayesian finite element (FE) model updating using direct model evaluations of large-scale high-fidelity FE models is extremely computationally expensive. Surrogate models can be used as fast emulators of FE models to accelerate the model calibration process. The physics/mechanics-based FE models are still the underpinning behind the surrogate models. Here, this paper evaluates the loss in accuracy and the gain in computational time while performing Bayesian model updating by using surrogate model evaluations compared to using direct FE model evaluations. This evaluation is crucial before entirely relying on surrogate models in model updating for structural health monitoring (SHM) and damage prognosis (DP) purposes. This paper also demonstrates Bayesian updating and surrogate model construction of large-scale high-fidelity FE models of infrastructure systems. In this regard, the miter gate structural system is considered as the testbed structure. Three predominant damage modes (loss of contact between gate and wall, loss of thickness due to corrosion, and loss of tension in the diagonal rods) are considered for model updating purposes. Bayesian model updating is performed using direct FE evaluations by leveraging parallel computing. Two types of surrogates, namely polynomial chaos expansion (PCE) and Gaussian process regression (GPR), are developed for the miter gate. Model updating is performed again using the trained surrogate models, and the updating results are compared with their counterparts obtained using the direct FE evaluation results. The posterior distribution of the FE model parameters obtained using the trained surrogates are sufficiently accurate with respect to the posterior obtained utilizing the direct FE evaluations. In addition, an approximate 4-fold decrease in the computational time was observed when using surrogate model evaluations instead of direct FE evaluations for model updating.

42 ENGINEERING↗

NASA NTRS1991

Rotating reactor studies

Undesired gravitational effects such as convection or sedimentation in a fluid can sometimes be avoided or decreased by the use of a closed chamber uniformly rotated about a horizontal axis. In a previous study, the spiral orbits of a heavy or buoyant particle in a uniformly rotating fluid were determined. The particles move in circles, and spiral in or out under the combined effects of the centrifugal force and centrifugal buoyancy. A optimization problem for the rotation rate of a cylindrical reactor rotated about its axis and containing distributed particles was formulated and solved. Related studies in several areas are addressed. A computer program based on the analysis was upgraded by correcting some minor errors, adding a sophisticated screen-and-printer graphics capability and other output options, and by improving the automation. The design, performance, and analysis of a series of experiments with monodisperse polystyrene latex microspheres in water were supported to test the theory and its limitations. The theory was amply confirmed at high rotation rates. However, at low rotation rates (1 rpm or less) the assumption of uniform solid-body rotation of the fluid became invalid, and there were increasingly strong secondary motions driven by variations in the mean fluid density due to variations in the particle concentration. In these tests the increase in the mean fluid density due to the particles was of order 0.015 percent. To a first approximation, these flows are driven by the buoyancy in a thin crescent-shaped depleted layer on the descending side of the rotating reactor. This buoyancy distribution is balanced by viscosity near the walls, and by the Coriolis force in the interior. A full analysis is beyond the scope of this study. Secondary flows are likely to be stronger for buoyant particles, which spiral in towards the neutral point near the rotation axis under the influence of their centrifugal buoyancy. This is because the depleted layer is thicker and extends all the way around the reactor.

Roberts, Glyn O.↗

DOE OSTI2016

Thermodynamics, Kinetics and Structural Evolution of ε-LiVOPO 4 over Multiple Lithium Intercalation

ε -Li x VOPO 4 is a promising multi-electron cathode for rechargeable lithium-ion bat- teries that has an extremely high theoretical capacity of 318 mAh/g. In this w ork, we demonstrate the stable cycling of more than one Li in solid-state-syn thesized ε - LiVOPO 4 over more than 20 cycles for the first time. Using a combination of densit y functional theory (DFT) calculations, X-ray pair distribution funct ion (PDF) analy- sis and X-ray Absorption Near Edge Structure (XANES) measurements, we pre sent a comprehensive analysis of the thermodynamics, kinetics and stru ctural evolution of ε -Li x VOPO 4 over the entire lithiation range. We identify two intermediate ph ases at x = 1 . 5 and 1.75 in the low-voltage regime using DFT calculations, and the comput ed and electrochemically measured voltage profiles are in excellent agree ment. Operando PDF techniques show a reversible hysteretic change in the short ( < 2° A) V-O bond lengths coupled with an irreversible extension of the long V-O bond ( > 2.4 °A) dur- ing low-voltage cycling. These observations are confirmed with EXAFS spe ctra. We identify hydrogen intercalation from the electrolyte decomposition as a likely explana- tion for the ~ 2 . 4°A V-O bond and its irreversible extension. In terms of electronic conductivity, we find Li x VOPO 4 to be a large band gap insulator across the entire lithiation range, and calculated small polaron migration barriers are similar to those of the olivine LiMPO 4 cathodes. Finally, we demonstrate that ε -LiVOPO 4 is likely to be a one-dimensional diffuser using climbing-image nudged elastic ban d calculations. These results highlight the importance of nano-sizing and carbon coating in achieving good electrochemical performance in this material.

Lin, Yuh-Chieh↗