Search NASA⌕ Search

SEARCH · Search NASA

Results for “strong scaling”

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 37 records · Page 2
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Disruption thermal load mitigation with shattered pellet injection on the Joint European Torus (JET)

Disruption mitigation remains a critical, unresolved challenge for ITER. To aid in addressing this challenge, a shattered pellet injection (SPI) system was installed on JET and experiments conducted at a range of thermal energy fractions and stored energies in excess of 7 MJ. The primary goals of these experiments were to investigate the efficacy of the SPI on JET and the ability of the plasma to assimilate multiple pellets. Single pellet injections produced a saturation in total radiated energy (Wrad) with increasing injected neon content, suggesting total radiation of stored thermal energy. Further increases in injected neon quantities resulted in reduced cooling times and current quench (CQ) durations, indicating higher impurity assimilation. No significant variation in CQ duration or W rad was observed when varying the deuterium content at fixed neon quantities. Additionally, higher assimilation, inferred by shorter CQ durations, was measured when a mechanical punch was used to launch the pellets and this was attributed to a lower pellet velocity leading to higher solid content in the pellet plume and larger fragments penetrating deeper into the plasma. Radiation asymmetries averaged over the cooling time were inferred from Emis3D and ranged from 1.6 to 1.9. Asymmetries averaged over the entire disruption sequence were found to increase at higher thermal energy fractions. The radiated energy fractions decreased with increasing thermal energy fractions but this trend was eliminated when toroidal asymmetries were accounted for with Emis3D. Pure deuterium pellets were able to produce cooling times of up to 75 ms with a gradual loss in thermal stored energy of up to 80%. Experiments with multiple pellet injection indicated W rad can be increased through pellet superposition and density can be increased with an additional D2 injection without a reduction in W rad . KPRAD modelling accurately reproduced the cooling times and the CQ duration at high thermal energies. Assimilation estimates from KPRAD indicated CQ rates scale strongly whilst W rad scales weakly and saturates with assimilated neon content. Comparable W rad can be achieved with lower assimilated neon quantities as longer cooling times are attained. Thus reduced neon content can be preferential in a thermal load mitigation scheme as it may reduce radiation asymmetries and prevent flash melting.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

Multi-GPU porting of a phase-change cascaded lattice Boltzmann method for three-dimensional pool boiling simulations

The Lattice Boltzmann method (LBM) has proven effective in simulating phase-change phenomena, such as melting, solidification, evaporation, and boiling. In this work, we develop a highly parallelized multi-GPU implementation of LBM for three-dimensional pool boiling simulations. The code is based on the OpenACC programming model, which enables the code to be deployed efficiently on multi-core CPUs, GPUs, and potentially other accelerators, without the need for architecture-specific rewrites. To support large-scale simulations, the domain is decomposed and distributed across multiple compute nodes using MPI. We demonstrate that the code exhibits excellent scaling properties, with ideal strong-scaling running with up to 256 GPUs on the MareNostrum5 cluster.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

PREEMPT: Scalable Epidemic Interventions Using Submodular Optimization on Multi-GPU Systems

Preventing and slowing the spread of epidemics is achieved through techniques such as vaccination and social distancing. Given practical limitations on the number of vaccines and cost of administration, optimization becomes a necessity. Previous approaches using mathematical programming methods have shown to be effective but are limited by computational costs. In this work, we make several contributions: First, we present a new approach for intervention via maximizing the influence of vaccinated nodes on the network. We call this method \preempt. Next, we prove submodular properties associated with the objective function of our method so that it aids in construction of an efficient greedy approximation strategy. Consequently, we present a new parallel algorithm based on greedy hill climbing for \preempt, and present an efficient parallel implementation for distributed CPU-GPU heterogeneous platforms. Our results demonstrate that \preempt{} is able to achieve a significant reduction (up to 6.75$\times$) in the percentage of people infected on a city-scale network. We also show strong scaling results of \preempt{} on 128 nodes of the Summit supercomputer. Our parallel implementation is able to significantly reduce time to solution, from hours to minutes on large networks. This work represents a first-of-its-kind effort in parallelizing greedy hill climbing and applying it toward devising effective interventions for epidemics.

Minutoli, Marco↗
DOE OSTI2023

A family of independent Variable Eddington Factor methods with efficient preconditioned iterative solvers

We present a family of discretizations for the Variable Eddington Factor (VEF) equations that have high-order accuracy on curved meshes and efficient preconditioned iterative solvers. The VEF discretizations are combined with the Discontinuous Galerkin transport discretization from to form effective high-order, linear transport methods. The VEF discretizations are derived by extending the unified analysis of Discontinuous Galerkin methods for elliptic problems presented by Arnold et al. to the VEF equations. This framework is used to define analogs of the interior penalty, second method of Bassi and Rebay, minimal dissipation local Discontinuous Galerkin, and continuous finite element methods. The analysis of subspace correction preconditioners, which use a continuous operator to iteratively precondition the discontinuous discretization, is extended to the case of the non-symmetric VEF system. Numerical results demonstrate that the VEF discretizations have arbitrary-order accuracy on curved meshes, preserve the thick diffusion limit, and are effective on a proxy problem from thermal radiative transfer in both outer transport iterations and inner preconditioned linear solver iterations. We demonstrate that the VEF solution converges to the S N transport solution as the mesh is refined on both problems with smooth and non-smooth behavior in angle. Parallel performance studies show that the interior penalty VEF discretization's linear solve weak scales out to 1024 processors and strong scales well on a single node. Particular attention is paid to the parallel performance of the VEF algorithm when used in combination with a parallel block Jacobi transport sweep.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Static Subspace Approximation for Random Phase Approximation Correlation Energies: Implementation and Performance

Developing theoretical understanding of complex reactions and processes at interfaces requires using methods that go beyond semilocal density functional theory to accurately describe the interactions between solvent, reactants and substrates. Methods based on many-body perturbation theory, such as the random phase approximation (RPA), have previously been limited due to their computational complexity. However, this is now a surmountable barrier due to the advances in computational power available, in particular through modern GPU-based supercomputers. In this work, we describe the implementation of RPA calculations within BerkeleyGW and show its favorable computational performance on large complex systems relevant for catalysis and electrochemistry applications. Our implementation builds off of the static subspace approximation which, by employing a compressed representation of the frequency dependent polarizability, enables the evaluation of the RPA correlation energy with significant acceleration and systematically controllable accuracy. We find that the computational cost of calculating the RPA correlation energy scales only linearly with system size for systems containing up to 50 thousand bands, and is expected to scale quadratically thereafter. We also show excellent strong scaling results across several supercomputers, demonstrating the performance and portability of this implementation.

algorithmic development↗
DOE OSTI2023

Domain Adaptation for Measurements of Strong Gravitational Lenses

Upcoming surveys are predicted to discover galaxy-scale strong lenses on the order of $10^5$, making deep learning methods necessary in lensing data analysis. Currently, there is insufficient real lensing data to train deep learning algorithms, but the alternative of training only on simulated data results in poor performance on real data. Domain Adaptation may be able to bridge the gap between simulated and real datasets. We utilize domain adaptation for the estimation of Einstein radius ($\Theta_E$) in simulated galaxy-scale gravitational lensing images with different levels of observational realism. We evaluate two domain adaptation techniques - Domain Adversarial Neural Networks (DANN) and Maximum Mean Discrepancy (MMD). We train on a source domain of simulated lenses and apply it to a target domain of lenses simulated to emulate noise conditions in the Dark Energy Survey (DES). We show that both domain adaptation techniques can significantly improve the model performance on the more complex target domain dataset. This work is the first application of domain adaptation for a regression task in strong lensing imaging analysis. Our results show the potential of using domain adaptation to perform analysis of future survey data with a deep neural network trained on simulated data.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2023

Domain Adaptation for Measurements of Strong Gravitational Lenses

Upcoming surveys are predicted to discover galaxy-scale strong lenses on the order of 10\textsuperscript{5}, making deep learning methods necessary in lensing data analysis. Currently, there is insufficient real lensing data to train deep learning algorithms, but the alternative of training only on simulated data results in poor performance on real data. Domain Adaptation may be able to bridge the gap between simulated and real datasets. We utilize domain adaptation for the estimation of Einstein radius ($\Theta_E$) in simulated galaxy-scale gravitational lensing images with different levels of observational realism. We evaluate two domain adaptation techniques - Domain Adversarial Neural Networks (DANN) and Maximum Mean Discrepancy (MMD). We train on a source domain of simulated lenses and apply it to a target domain of lenses simulated to emulate noise conditions in the Dark Energy Survey (DES). We show that both domain adaptation techniques can significantly improve the model performance on the more complex target domain dataset. This work is the first application of domain adaptation for a regression task in strong lensing imaging analysis. Our results show the potential of using domain adaptation to perform analysis of future survey data with a deep neural network trained on simulated data.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2023

Domain Adaptation for Measurements of Strong Gravitational Lenses

Upcoming surveys are predicted to discover galaxy-scale strong lenses on the magnitude of 105, making deep learning methods necessary in lensing data analysis. Currently, there is insufficient real lensing data to train deep learning algorithms, but training only on simulated data results in poor performance on real data. Domain adaptation can bridge the gap between simulated and real datasets. We adopt domain adaptation on the estimation of Einstein radius in simulated galaxy-scale gravitational lensing images. We evaluate two domain adaptation techniques - domain adversarial neural networks (DANN) and maximum mean discrepancy (MMD). We train on a source domain of simulated lenses and apply it to a target domain with emulation of DES survey conditions. We show that both domain adaptation techniques can significantly improve the model performance on the more complex target domain datasets. Our results show the potential of using domain adaptation to perform analysis on future survey data with a deep neural network trained on simulated data.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Nek5000/RS performance on advanced GPU architectures

The authors explore performance scalability of the open-source thermal-fluids code, NekRS, on the U.S. Department of Energy's leadership computers, Crusher, Frontier, Summit, Perlmutter, and Polaris. Particular attention is given to analyzing performance and time-to-solution at the strong-scale limit for a target efficiency of 80%, which is typical for production runs on the DOE's high-performance computing systems. Several examples of anomalous behavior are also discussed and analyzed.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Image Gradient Decomposition for Parallel and Memory-Efficient Ptychographic Reconstruction

Ptychography is a popular microscopic imaging modality for many scientific discoveries and sets the record for highest image resolution. Unfortunately, the high image resolution for ptychographic reconstruction requires significant amount of memory and computations, forcing many applications to compromise their image resolution in exchange for a smaller memory footprint and a shorter reconstruction time. In this paper, we propose a novel image gradient decomposition method that significantly reduces the memory footprint for ptychographic reconstruction by tessellating image gradients and diffraction measurements into tiles. In addition, we propose a parallel image gradient decomposition method that enables asynchronous point-to-point communications and parallel pipelining with minimal overhead on a large number of GPUs. Our experiments on a Titanate material dataset (PbTiO3) with 16632 probe locations show that our Gradient Decomposition algorithm reduces memory footprint by 51 times. In addition, it achieves time-to-solution within 2.2 minutes by scaling to 4158 GPUs with a super-linear strong scaling efficiency at 364% compared to runtimes at 6 GPUs. This performance is 2.7 times more memory efficient, 9 times more scalable and 86 times faster than the state-of-the-art algorithm.

Wang, Xiao↗
DOE OSTI2025

TorchBraid: High-Performance Layer-Parallel Training of Deep Neural Networks with MPI and GPU Acceleration

TorchBraid is a high-performance implementation of layer-parallel training for deep neural networks (DNNs) supporting MPI-based parallelism and GPU acceleration. Layer-parallel training has been developed to overcome the serialization inherent in forward and backward propagation of DNNs that limits utilization of computational resources in the strong scaling limit. To achieve this, TorchBraid integrates the PyTorch neural network framework with the state-of-the-art XBraid time-parallel library. Furthermore, this article presents the use and performance of TorchBraid, in addition to solutions for overcoming the algorithmic challenges inherent in combining automatic differentiation with layer-parallel. Results are presented with and without GPU acceleration for the Tiny ImageNet and MNIST image classification data sets, as well as recurrent neural networks. Overall, TorchBraid enables fast training of DNNs, both in a strong and weak scaling context. In addition to the TorchBraid software, several new advances in applying layer-parallel algorithms are detailed. Integration of layer-parallel with data-parallel algorithms is presented for the first time, showing the computational advantages of the combination. Standard deep learning techniques, like batch-normalization, are developed for layer-parallel training. Finally, a new approach combining layer-parallel with spatial coarsening in order to accelerate training for 3D image classification shows roughly a 10× speedup over serial execution.

Layer-parallel↗
DOE OSTI2025

Sensitivity of tropical orographic precipitation to wind speed with implications for future projections

Abstract. Some of the rainiest regions on Earth lie upstream of tropical mountains, where the interaction of prevailing winds with orography produces frequent precipitating convection. Yet the response of tropical orographic precipitation to the large-scale wind and temperature variations induced by anthropogenic climate change remains largely unconstrained. Here, we quantify the sensitivity of tropical orographic precipitation to background cross-slope wind using theory, idealized simulations, and observations. We build on a recently developed theoretical framework that characterizes the orographic enhancement of seasonal mean precipitation, relative to upstream regions, as a response of convection to cooling and moistening of the lower free troposphere by stationary orographic gravity waves. Using this framework and convection-permitting simulations, we show that higher cross-slope wind speeds deepen the penetration of the cool and moist gravity wave perturbation upstream of orography, resulting in a mean rainfall increase of 20 % (m s−1)−1 to 30 % (m s−1)−1 increase in cross-slope wind speed. Additionally, we show that orographic precipitation in five tropical regions exhibits a similar dependence on changes in cross-slope wind at both seasonal and daily timescales. Given next-century changes in large-scale winds around tropical orography projected by global climate models, this strong scaling rate implies wind-induced changes in some of Earth's rainiest regions that are comparable with any produced directly by increases in global mean temperature and humidity.

Nicolas, Quentin (ORCID:000000024116973X)↗
DOE OSTI2024

Optimizing Distributed Training on Frontier for Large Language Models

Large language models (LLMs) have demonstrated remarkable success as foundational models, benefiting various downstream applications through fine-tuning. Loss scaling studies have demonstrated the superior performance of larger LLMs compared to their smaller counterparts. Nevertheless, training LLMs with billions of parameters poses significant challenges and requires considerable computational resources. For example, training a one trillion parameter GPT-style model on 20 trillion tokens requires a staggering 120 million exaflops. This research explores efficient distributed training strategies to extract this computation from Frontier, the world's first exascale supercomputer. We enable and investigate various model and data parallel training techniques, such as tensor parallelism, pipeline parallelism, and sharded data parallelism, to facilitate training a trillion-parameter model on Frontier. We empirically assess these techniques and their associated parameters to determine their impact on memory footprint, communication latency, and GPU's computational efficiency. We analyze the complex interplay among these techniques and find a strategy to combine them to achieve high throughput through hyperparameter tuning. We have identified efficient strategies for training large LLMs of varying sizes through empirical analysis and hyperparameter tuning. For 22 Billion, 175 Billion, and 1 Trillion parameters, we achieved GPU throughputs of 38.38%, 36.14%, and 31.96%, respectively. For the training of the 175 Billion parameter model and the 1 Trillion parameter model, we achieved 100% weak scaling efficiency on 1024 and 3072 Mi250X GPUs, respectively. We also achieved strong scaling efficiencies of 89% and 87% for these two models. We trained these models only tens of iterations instead of training till completion.

Yin, Junqi↗
DOE OSTI2025

Breaking the mold: Overcoming the time constraints of molecular dynamics on general-purpose hardware

The evolution of molecular dynamics (MD) simulations has been intimately linked to that of computing hardware. For decades following the creation of MD, simulations have improved with computing power along the three principal dimensions of accuracy, atom count (spatial scale), and duration (temporal scale). Since the mid-2000s, computer platforms have, however, failed to provide strong scaling for MD, as scale-out central processing unit (CPU) and graphics processing unit (GPU) platforms that provide substantial increases to spatial scale do not lead to proportional increases in temporal scale. Important scientific problems therefore remained inaccessible to direct simulation, prompting the development of increasingly sophisticated algorithms that present significant complexity, accuracy, and efficiency challenges. While bespoke MD-only hardware solutions have provided a path to longer timescales for specific physical systems, their impact on the broader community has been mitigated by their limited adaptability to new methods and potentials. In this work, we show that a novel computing architecture, the Cerebras wafer scale engine, completely alters the scaling path by delivering unprecedentedly high simulation rates up to 1.144 M steps/s for 200 000 atoms whose interactions are described by an embedded atom method potential. This enables direct simulations of the evolution of materials using general-purpose programmable hardware over millisecond timescales, dramatically increasing the space of direct MD simulations that can be carried out. In this paper, we provide an overview of advances in MD over the last 60 years and present our recent result in the context of historical MD performance trends.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

GR-Athena++: General-relativistic Magnetohydrodynamics Simulations of Neutron Star Spacetimes

We present the extension of GR-Athena++ to general-relativistic magnetohydrodynamics (GRMHD) for applications to neutron star spacetimes. The new solver couples the constrained transport implementation of Athena++ to the Z4c formulation of the Einstein equations to simulate dynamical spacetimes with GRMHD using oct-tree adaptive mesh refinement. We consider benchmark problems for isolated and binary neutron star spacetimes demonstrating stable and convergent results at relatively low resolutions and without grid symmetries imposed. The code correctly captures magnetic field instabilities in nonrotating stars with total relative violation of the divergence-free constraint of 10 –16 . It handles evolutions with a microphysical equation of state and black hole formation in the gravitational collapse of a rapidly rotating star. For binaries, we demonstrate correctness of the evolution under the gravitational radiation reaction and show convergence of gravitational waveforms. We showcase the use of adaptive mesh refinement to resolve the Kelvin–Helmholtz instability at the collisional interface in a merger of magnetised binary neutron stars. GR-Athena++ shows strong scaling efficiencies above 80% in excess of 10 5 CPU cores and excellent weak scaling is shown up to ~5 × 10 5 CPU cores in a realistic production setup. GR-Athena++ allows for the robust simulation of GRMHD flows in strong and dynamical gravity with exa-scale computers.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Simulating many-engine spacecraft: Exceeding 1 quadrillion degrees of freedom via information geometric regularization

We present an optimized implementation of the recently proposed information geometric regularization (IGR) for unprecedented scale simulation of compressible fluid flows applied to multi-engine spacecraft boosters. We improve upon state-of-the-art computational fluid dynamics (CFD) techniques in terms of computational cost, memory footprint, and energy-to-solution metrics. Unified memory on coupled CPU–GPU or APU platforms increases problem size with negligible overhead. Mixed half/single-precision storage and computation are used on well-conditioned numerics. We simulate flow at 200 trillion grid points and 1 quadrillion degrees of freedom, exceeding the current record by a factor of 20. A factor of 4 wall-time speedup is achieved over optimized baselines. Ideal weak scaling is observed on OLCF Frontier, LLNL El Capitan, and CSCS Alps using the full systems. Strong scaling is near ideal at extreme conditions, including 80% efficiency on CSCS Alps with an 8 node baseline and stretching to the full system.

Wilfong, Benjamin [Georgia Institute of Technology↗
DOE OSTI2022

Early-Stage Evaluation of Catalyst Manufacturing Cost and Environmental Impact Using CatCost

The costs and environmental impacts of catalyst manufacture are often neglected during early-stage research because of a lack of accessible, standardized tools to assess them. Here we report the key features of CatCost, a free and public estimation tool for the evaluation of catalyst cost. We demonstrate its functionality with a case study of diverse catalysts (ZSM-5, Pt/TiO2 and Mo2C) for the catalytic fast pyrolysis of biomass. We quantified the economic and environmental improvements made by replacing circulating-bed ZSM-5 with more stable, fixed-bed Pt/TiO2 and Mo2C catalysts, while revealing the effects of synthesis methods and production scale on catalyst costs. The manufacture of ZSM-5 had a large processing cost contribution that was strongly scale dependent, whereas the costs of the other catalysts were dominated by raw materials at all scales. Furthermore, while ZSM-5 costs the least per kilogram, the more stable catalysts cost half as much per gallon of fuel.

BIOMASS FUELS,ENERGY PLANNING, POLICY, AND ECONOMY↗
DOE OSTI2026

Open quantum system approach to inclusive jet production in heavy-ion collisions

We derive a factorization formula for inclusive jet production in heavy-ion collisions using the tools of Effective Field Theory (EFT). We show how physics at widely separated scales in this process can be systematically separated by matching to EFTs at successively lower virtualities. Owing to a strong scale separation, we recover a vacuum-like DGLAP evolution above the jet scale, while the additional low-energy scales induced by the medium effectively probe the internal structure of the jet. As a result, the cross section can be written as a series with an increasing number of subjets characterized by perturbative matching coefficients each of which is convolved with a distinct function. These functions encode broadening, medium-induced radiations as well as quantum interference such as the Landau-Pomeranchuk-Migdal effect and color coherence dynamics to all orders in perturbation theory. As a first application of this EFT framework, we investigate the case of an unresolved jet and show how the cross section can be factorized and fully separate the jet dynamics from the universal physics of the medium. To compare to the existing literature, we explicitly compute the medium jet function at next-to-leading order in the coupling and leading order in medium opacity.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗