Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “massive parallelization”

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 901 records · Page 50
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Efficient Treatment of Large Active Spaces through Multi-GPU Parallel Implementation of Direct Configuration Interaction

In this study, we have extended our graphical processing unit (GPU)-accelerated direct configuration interaction program to multiple devices, reducing iteration times for configuration spaces of 165 million determinants to only 3 s using NVIDIA P100 GPUs. Similar improvements in the one- and two-particle reduced density matrix formation allow for fast analytical energy gradients and electronic properties. Our parallel algorithm enables the calculation of arbitrarily large configuration spaces (limited only by available system memory), with iteration times of 13 min for an active space of 18 electrons in 18 orbitals (2.4 billion determinants) using six consumer grade NVIDIA 1080Ti GPUs. These advances enable routine molecular dynamics simulations, geometry optimizations, and absorption spectrum calculations for molecules with large configuration spaces, a task that has heretofore required massive computational effort. In this work, we demonstrate the utility of our program by generating the absorption spectrum for diphenyl acetylene at the floating occupation molecular orbital complete active space configuration interaction level of theory. Lastly, several active spaces were investigated to assess the dependence of spectral features on orbital space dimension.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRS1981

Concerning the Wolf-Rayet and other luminous early-type stars

Effective temperatures, radii, and luminosities were determined from S2/68, ANS, UBV, and uvby photometry for four B0/B1 supergiants, four O4 stars, and four WN7/WN8 stars as well as for four test stars having spectral types between B1.5 V and 09 V and five stars with known angular diameters and effective temperatures. The effective temperatures of B1 Ia+ stars are found to be near 17,000 K, those of O4 stars near 45,000, and those of WN7/WN8 stars near 26,000 K. The question of modeling the atmospheres of hot luminous stars is examined, and it is noted that the photosphere can be modeled adequately using a classical plane-parallel layer model atmosphere. In addition, it is found that the Wolf-Rayet stars of types WN7/WN8 fall in the H-R diagram near the B0 Ia stars, while the others fall near B0.5 III stars. The evolutionary relationship between the Wolf-Rayet and O stars is considered; it is suggested that a Wolf-Rayet spectrum is a short-lived phase in the life of a massive star.

Underhill, A. B.↗
NASA NTRS2012

Optical Signature Analysis of Tumbling Rocket Bodies via Laboratory Measurements

The NASA Orbital Debris Program Office has acquired telescopic lightcurve data on massive intact objects, specifically spent rocket bodies, in order to ascertain tumble rates in support of the Active Debris Removal (ADR) task to help remediate the LEO environment. Rotation rates are needed to plan and develop proximity operations for potential future ADR operations. To better characterize and model optical data acquired from ground-based telescopes, the Optical Measurements Center (OMC) at NASA/JSC emulates illumination conditions in space using equipment and techniques that parallel telescopic observations and source-target-sensor orientations. The OMC employs a 75-watt Xenon arc lamp as a solar simulator, an SBIG CCD camera with standard Johnson/Bessel filters, and a robotic arm to simulate an object's position and rotation. The light source is mounted on a rotary arm, allowing access any phase angle between 0 -- 360 degrees. The OMC does not attempt to replicate the rotation rates, but focuses on how an object is rotating as seen from multiple phase angles. The two targets studied are scaled (1:48), SL-8 Cosmos 3M second stages. The first target is painted in the standard government "gray" scheme and the second target is primary white, as used for commercial missions. This paper summarizes results of the two scaled rocket bodies, each rotated about two primary axes: (a) a spin-stabilized rotation and (b) an end-over-end rotation. The two rotation states are being investigated as a basis for possible spin states of rocket bodies, beginning with simple spin states about the two primary axes. The data will be used to create a database of potential spin states for future works to convolve with more complex spin states. The optical signatures will be presented for specific phase angles for each rocket body and shown in conjunction with acquired optical data from multiple telescope sources.

Cowardin, H.↗
NASA NTRS1995

Electromagnetic instabilities in solar wind interaction with dusty cometary plasmas

Dusty plasmas contain charged dust grains which are much more massive than protons, carry high negative charges due to preferential capture of electrons, and do not have a fixed charge. Fluctuations in the grain charges due to liberation or capture of additional electrons and protons translate as mass and momentum losses or gains for these species, which can render linear modes unstable. On the other hand, many authors have addressed the pickup of ions of cometary origin by the solar wind, which for the parallel part is due to relative streaming between cometary and solar wind ions which excites low-frequency electromagnetic turbulence. In the present work we look again at those instabilities by including effects due to the presence of charged dust in the cometary environments. We have investigated several frequency regimes: nonresonant below the cometary watergroup gyrofrequency, nonresonant below the cometary charged dust gyrofrequency (new and interesting but highly unlikely!) and resonant with the cometary watergroup ions. For most parameter ranges either the existing instabilities are enhanced, showing that the presence of charged dust facilitates the cometary ion pickup by the solar wind, or new instabilities have been shown to exist. Similar conclusions might be relevant for other kinds of astrophysical and heliospheric plasmas containing charged dust, as in planetary rings.

Verheest, Frank↗
DOE OSTI2024

Overview of ASDEX upgrade results in view of ITER and DEMO

Experiments on ASDEX Upgrade (AUG) in 2021 and 2022 have addressed a number of critical issues for ITER and EU DEMO. A major objective of the AUG programme is to shed light on the underlying physics of confinement, stability, and plasma exhaust in order to allow reliable extrapolation of results obtained on present day machines to these reactor-grade devices. Concerning pedestal physics, the mitigation of edge localised modes (ELMs) using resonant magnetic perturbations (RMPs) was found to be consistent with a reduction of the linear peeling-ballooning stability threshold due to the helical deformation of the plasma. Conversely, ELM suppression by RMPs is ascribed to an increased pedestal transport that keeps the plasma away from this boundary. Candidates for this increased transport are locally enhanced turbulence and a locked magnetic island in the pedestal. The enhanced D-alpha (EDA) and quasi-continuous exhaust (QCE) regimes have been established as promising ELM-free scenarios. Here, the pressure gradient at the foot of the H-mode pedestal is reduced by a quasi-coherent mode, consistent with violation of the high-n ballooning mode stability limit there. This is suggestive that the EDA and QCE regimes have a common underlying physics origin. In the area of transport physics, full radius models for both L- and H-modes have been developed. These models predict energy confinement in AUG better than the commonly used global scaling laws, representing a large step towards the goal of predictive capability. A new momentum transport analysis framework has been developed that provides access to the intrinsic torque in the plasma core. In the field of exhaust, the X-Point Radiator (XPR), a cold and dense plasma region on closed flux surfaces close to the X-point, was described by an analytical model that provides an understanding of its formation as well as its stability, i.e., the conditions under which it transitions into a deleterious MARFE with the potential to result in a disruptive termination. With the XPR close to the divertor target, a new detached divertor concept, the compact radiative divertor, was developed. Here, the exhaust power is radiated before reaching the target, allowing close proximity of the X-point to the target. No limitations by the shallow field line angle due to the large flux expansion were observed, and sufficient compression of neutral density was demonstrated. With respect to the pumping of non-recycling impurities, the divertor enrichment was found to mainly depend on the ionisation energy of the impurity under consideration. In the area of MHD physics, analysis of the hot plasma core motion in sawtooth crashes showed good agreement with nonlinear 2-fluid simulations. This indicates that the fast reconnection observed in these events is adequately described including the pressure gradient and the electron inertia in the parallel Ohm’s law. Concerning disruption physics, a shattered pellet injection system was installed in collaboration with the ITER International Organisation. Thanks to the ability to vary the shard size distribution independently of the injection velocity, as well as its impurity admixture, it was possible to tailor the current quench rate, which is an important requirement for future large devices such as ITER. Progress was also made modelling the force reduction of VDEs induced by massive gas injection on AUG. The H-mode density limit was characterised in terms of safe operational space with a newly developed active feedback control method that allowed the stability boundary to be probed several times within a single discharge without inducing a disruptive termination. Regarding integrated operation scenarios, the role of density peaking in the confinement of the ITER baseline scenario (high plasma current) was clarified. The usual energy confinement scaling ITER98(p,y) does not capture this effect, but the more recent H20 scaling does, highlighting again the importance of developing adequate physics based models. Advanced tokamak scenarios, aiming at large non-inductive current fraction due to non-standard profiles of the safety factor in combination with high normalised plasma pressure were studied with a focus on their access conditions. A method to guide the approach of the targeted safety factor profiles was developed, and the conditions for achieving good confinement were clarified. Based on this, two types of advanced scenarios (‘hybrid’ and ‘elevated’

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

Toward a persistent event-streaming system for high-performance computing applications

High-performance computing (HPC) applications have traditionally relied on parallel file systems and file transfer services to manage data movement and storage. Alternative approaches have been proposed that use direct communications between application components, trading persistence and fault tolerance for speed. Event-driven architectures, as popularized in enterprise contexts, present a compelling middle ground, avoiding the performance cost and API constraints of parallel file systems while retaining persistence and offering impedance matching between application components. However, adapting streaming frameworks to HPC workloads requires addressing challenges unique to HPC systems. This paper investigates the potential for a streaming framework designed for HPC infrastructures and use cases. We introduce Mofka, a persistent event-streaming framework designed specifically for HPC environments. Mofka combines the capabilities of a traditional streaming service with optimizations tailored to the HPC context, such as support for massively multicore nodes, efficient scaling for large producer-consumer workflows, RDMA-enabled high-performance network communications, specialized network fabrics with multiple links per node, and efficient handling of large scientific data payloads. Built using the Mochi suite of HPC data service components, Mofka provides a lightweight, modular, and high-performance solution for persistent streaming in HPC systems. We present the architecture of Mofka and evaluate its performance against Kafka and Redpanda using benchmarks on diverse platforms, including Argonne's Polaris and Oak Ridge's Frontier supercomputers, showing up to 8× improvement in throughput in some scenarios. We then demonstrate its utility in several real-world applications: a tomographic reconstruction pipeline, a workflow for the discovery of metal-organic frameworks for carbon capture, and the instrumentation of Dask workflows for provenance tracking and performance analysis.

HPC↗
NASA NTRS2001

Hyperspectral Image Classification using a Self-Organizing Map

The use of hyperspectral data to determine the abundance of constituents in a certain portion of the Earth's surface relies on the capability of imaging spectrometers to provide a large amount of information at each pixel of a certain scene. Today, hyperspectral imaging sensors are capable of generating unprecedented volumes of radiometric data. The Airborne Visible/Infrared Imaging Spectrometer (AVIRIS), for example, routinely produces image cubes with 224 spectral bands. This undoubtedly opens a wide range of new possibilities, but the analysis of such a massive amount of information is not an easy task. In fact, most of the existing algorithms devoted to analyzing multispectral images are not applicable in the hyperspectral domain, because of the size and high dimensionality of the images. The application of neural networks to perform unsupervised classification of hyperspectral data has been tested by several authors and also by us in some previous work. We have also focused on analyzing the intrinsic capability of neural networks to parallelize the whole hyperspectral unmixing process. The results shown in this work indicate that neural network models are able to find clusters of closely related hyperspectral signatures, and thus can be used as a powerful tool to achieve the desired classification. The present work discusses the possibility of using a Self Organizing neural network to perform unsupervised classification of hyperspectral images. In sections 3 and 4, the topology of the proposed neural network and the training algorithm are respectively described. Section 5 provides the results we have obtained after applying the proposed methodology to real hyperspectral data, described in section 2. Different parameters in the learning stage have been modified in order to obtain a detailed description of their influence on the final results. Finally, in section 6 we provide the conclusions at which we have arrived.

Martinez, P.↗
NASA NTRS2019

An Integrated Data Analytics Platform

An Integrated Science Data Analytics Platform is an environment that enables the confluence of resources for scientific investigation. It harmonizes data, tools and computational resources which subsequently enable the research community to focus on the investigation rather than spending time on security, data preparation, management, etc. OceanWorks is a NASA technology integration project to establish a cloud-based Integrated Ocean Science Data Analytics Platform at NASA’s Physical Oceanography Distributed Active Archive Center (PO.DAAC) for big ocean science. It focuses on advancement and maturity by bringing together several NASA open-source, big data projects for parallel analytics, anomaly detection, in-situ to satellite data matchup, quality-screened data subsetting, search relevancy, and data discovery. Our communities are relying on data distributed through data centers such as the PO.DAAC, COAPS, NCAR, and many others to conduct their research. In typical investigations, scientists would engage in: search for data, evaluate the relevance of that data, download it, and then apply algorithms to identify trends. Such workflow cannot scale if the research involves a massive amount of data or multi-variate measurements. NASA’s Surface Water and Ocean Topography (SWOT) mission is expected to produce massive amount of observational data during its 3-year nominal mission. Collections like SWOT challenges all existing Earth Science data archival, distribution and analysis paradigms. In this paper, we will discuss how OceanWorks enhances the analysis of physical ocean data where the computation is done on an elastic cloud platform next to the archive to deliver fast, web-accessible services for working with oceanographic measurements.

Yang, Chaowei↗
DOE OSTI2023

Enhanced negative energy with a massless Dirac field

Motivated by traversable wormhole constructions that require large amounts of negative energy, we explore constraints on the amount of negative energy that can be carried by a free Dirac field in a slab-shaped region between two parallel spatial planes. Specifically, we ask what is the minimum possible uniform energy density that can exist at some time, considering all possible states and all possibilities for the physics outside the slab. The vacuum state where we identify the two sides of the slab with antiperiodic boundary conditions gives one possible state with uniform negative energy, but we argue that states with more negative energy exist above 1+1 dimensions. Technically, we reduce the problem to studying a massive Dirac field on an interval in 1+1 dimensions and numerically search for states with uniform energy density in a lattice regulated model. We succeed in finding states with enhanced negative energy (relative to the antiperiodic vacuum) which also appear to have a sensible continuum limit. Our results for the mass-dependence of the minimum uniform energy density in 1+1 dimensions suggest that for a 3+1 dimensional massless Dirac fermion, it is possible to have states with arbitrarily large uniform negative energy density in an arbitrarily wide slab.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

Towards performance portability in the Spark astrophysical magnetohydrodynamics solver in the Flash-X simulation framework

Simulations of core-collapse supernovae, and other astrophysical phenomena, are quintessential extreme-scale computing challenges. For core-collapse supernova simulations to be carried out by the ExaStar project under the Exascale Computing Project umbrella, a robust, efficient, and state-of-the-art magnetohydrodynamics solver is a critical requirement. In Flash-X, the primary software instrument for ExaStar, a new magnetohydrodynamics solver has been designed and implemented from the ground up to achieve accuracy and efficiency for simulations of complex astrophysical flows. This new solver, dubbed Spark, uses high-order spatial reconstruction, Runge-Kutta time integration, and an efficient cell-centered approach to satisfying the divergence-free condition for the magnetic fields. Spark was written to be optimized for data locality in cache hierarchy of CPUs. Since data locality optimizations for cache hierarchy are not directly compatible with those of accelerators, we have taken the approach of using program synthesis to avoid massive amounts of code replication that would be necessary if we were to maintain two different versions of the solver. Our program synthesis relies on a simple key-dictionary approach, implemented in python, that enables us to assemble the version of the solver suitable for the target hardware from code fragments identified by specific keys. In this work, we describe the data locality optimizations of the solver for CPUs and accelerators and the program synthesis tools that enable this portability. We also detail the parallel performance of Spark for both CPUs and accelerators.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2019

Cosmic Pathways for Compact Groups in the Milli-Millennium Simulation

We detected 10 compact galaxy groups (CGs) at z=0 in the semianalytic galaxy catalog of Guo et al. for themilli-Millennium Cosmological Simulation (sCGs in mGuo2010a). We aimed to identify potential canonicalpathways for compact group evolution and thus illuminate the history of observed nearby CGs. By constructingmerger trees for z=0 sCG galaxies, we studied the cosmological evolution of key properties and compared themwith z=0 Hickson CGs (HCGs). We found that, once sCG galaxies come within 1 (0.5) Mpc of their mostmassive galaxy, they remain within that distance until z=0, suggesting sCG "birth redshifts." At z=0 stellarmasses of sCG most massive galaxies are within 1010M*/Me1011. In several cases, especially in the twofour- and five-member systems, the amount of cold gas mass anticorrelates with stellar mass, which in turncorrelates with hot gas mass. We define the angular difference between group members' 3D velocity vectors,vel, and note that many of the groups are long-lived because their small values of vel indicate a significantparallel component. For triplets in particular, vel values range between 20° and 40° so that galaxies are comingtogether along roughly parallel paths, and pairwise separations do not show large pronounced changes after closeencounters. The best agreement between sCG and HCG physical properties is for M* galaxy values, but HCGvalues are higher overall, including for star formation rates (SFRs). Unlike HCGs, due to a tail at low SFR and M*and a lack of M*1011Me galaxies, only a few sCG galaxies are on the star-forming main sequence

Tzanavaris, Panayiotis↗
DOE OSTI2020

Enabling Highly Efficient Capsule Networks Processing Through A PIM-Based Architecture Design

In recent years, the CNNs have achieved great successes in the image processing tasks, e.g., image recognition and object detection. Unfortunately, traditional CNN's classication is found to be easily misled by increasingly complex image features due to the usage of pooling operations, hence unable to preserve accurate position and pose information of the objects. To address this challenge, a novel neural network structure called Capsule Network has been proposed, which introduces equivariance through capsules to signicantly enhance the learning ability for image segmentation and object detection. Due to its requirement of performing a high volume of matrix operations, CapsNets have been generally accelerated on modern GPU platforms that provide highly optimized software library for common deep learning tasks. However, based on our performance characterization on modern GPUs, CapsNets exhibit low effciency due to the special program and execution features of their routing procedure, including massive unshareable intermediate variables and intensive syn- chronizations, which are very dicult to optimize at software level. To address these challenges, we propose a hybrid computing architecture design named PIM-CapsNet. It preserves GPU's on-chip computing capability for accelerating CNN types of layers in CapsNet, while pipelining with an off-chip in-memory acceleration solution that effectively tackles routing procedure's ineffciency by leveraging the processing-in-memory capability of today's 3D stacked memory. Using routing procedure's inherent parallellization feature, our design enables hierarchical improvements on CapsNet inference effciency through minimizing data movement and maximizing parallel processing in memory. Evaluation results demonstrate that our proposed design can achieve substantial improvement on both performance and energy savings for CapsNet inference, with almost zero accuracy loss. The results also suggest good performance scalability in optimizing the routing procedure with increasing network size.

Zhang, Xingyao↗
NASA NTRS1990

Cometary compact H II regions are stellar-wind bow shocks

Comet-shaped H II regions, like G34.3 + 0.2, are easily explained as bow shocks created by wind-blowing massive stars moving supersonically through molecular clouds. The required velocities of the stars through dense clumps are less than about 10 km/s, comparable to the velocity dispersion of stars in OB associations. An analytic model of bow shocks matches the gross characteristics seen in the radio continuum and the velocity structure inferred from hydrogen recombination and molecular line observations. The champagne flow model cannot account for these structures. VLBI observations of masers associated with the shells of cometary compact H II regions should reveal tailward proper motions predominantly parallel to the shell, rather than perpendicular. It is predicted that over a decade baseline, high signal-to-noise VLA observations of this class of objects will show headward pattern motion in the direction of the symmetry axis, but not expansion. Finally, shock-generated and coronal infrared lines are also predicted.

Van Buren, Dave↗
DOE OSTI2022

Management and Storage of Scientific Data

Scientific discoveries rely heavily on efficient access, search, and management of massive data sets. Data management technologies have, for decades, provided foundational capabilities for scientific computing. Just as storage, input/output (I/O), and data management have been fundamental to simulation-based science for many years, so too are capable data-management technologies key to the success of today’s scientific workflows utilizing data intensive and machine learning (ML) techniques. The Department of Energy, Office of Science, Advanced Scientific Computing Research (ASCR) program has invested broadly in data-management research focused on high-performance computing (HPC) systems, from parallel file systems that store data to application software that makes these systems more productive. Still, advances in technology combined with growing diversity of supported science strongly motivate continued investment in this area. In January 2022, ASCR convened a workshop to identify priority research directions in the area of data management for high-performance and scientific computing. Attendees were challenged to identify promising approaches that would support the breadth of the DOE mission, including the explosion of artificial intelligence (AI) uses and the growing needs of experimental and observational science. Technological and science drivers were identified and considered as they relate to key aspects of data management such as interfaces, architectural design, and FAIR principles (Findable, Accessible, Interoperable, and Reusable). The thoughts of the workshop participants were distilled into a set of four priority research directions with the potential for high impact on DOE science. These research directions are summarized in the following pages.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗
DOE OSTI2024

Towards reverse mode automatic differentiation of Kokkos-based codes

Derivative computation is a key component of optimization, sensitivity analysis, uncertainty quantification, and the solving of nonlinear problems. Automatic differentiation (AD) is a powerful technique for evaluating such derivatives, and in recent years, has been integrated into programming environments such as Jax, PyTorch, and TensorFlow to support derivative computations needed for training of machine learning models, facilitating wide-spread use of these technologies. The C++ language has become the de facto standard for scientific computing due to numerous factors, yet language complexity has made the wide-spread adoption of AD technologies for C++ difficult, hampering the incorporation of powerful differentiable programming approaches into C++ scientific simulations. This is exacerbated by the increasing emergence of architectures, such as GPUs, with limited memory capabilities and requiring massive thread-level concurrency. C++ AD tools must effectively use these environments to bring novel scientific simulations to next-generation DOE experimental and observational facilities. In this project, we investigated source transformation-based automatic differentiation using LLVM compiler infrastructure to automatically generate portable and efficient gradient computations of Kokkos-based code. We have demonstrated that our proposed strategy is feasible by investigating the usage of a prototype LLVM-based source transformation tool to generate gradients of simple functions made of sequences of simple Kokkos parallel regions. Speedups of up to 500x compared to Sacado were observed on NVIDIA V100 GPU.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1991

Hubble Space Telescope imaging of Eta Carinae

New high spatial resolution observations of the material around Eta Carinae, obtained with the Hubble Space Telescope Wide Field/Planetary Camera, are presented. The star Eta Carinae is one of the most massive and luminous stars in the Galaxy, and has been episodically expelling significant quantities of gas over the last few centuries. The morphology of the bright central nebulosity (the homunculus) indicates that it is a thin shell with very well defined edges, and is clumpy on 0.2 arcsec (about 10 to the 16th cm) scales. An extension to the northeast of the star (NN/NS using Walborn's 1976 nomenclature) appears to be a stellar jet and its associated bow shock. The bow shock is notable for an intriguing series of parallel linear features across its face. The S ridge and the W arc appear to be part of a 'cap' of emission located to the SW and behind the star. Together, the NE jet and the SW cap suggest that the symmetry axis for the system runs NE-SW rather than SE-NW, as previously supposed. Overall, the data indicate that the material around the star may represent an oblate shell with polar blowouts, rather than a bipolar flow.

Hester, J. J.↗
NASA NTRS2016

Evolution of Flexible Multibody Dynamics for Simulation Applications Supporting Human Spaceflight

During the course of transition from the Space Shuttle and International Space Station programs to the Orion and Journey to Mars exploration programs, a generic flexible multibody dynamics formulation and associated software implementation has evolved to meet an ever changing set of requirements at the NASA Johnson Space Center (JSC). Challenging problems related to large transitional topologies and robotic free-flyer vehicle capture/ release, contact dynamics, and exploration missions concept evaluation through simulation (e.g., asteroid surface operations) have driven this continued development. Coupled with this need is the requirement to oftentimes support human spaceflight operations in real-time. Moreover, it has been desirable to allow even more rapid prototyping of on-orbit manipulator and spacecraft systems, to support less complex infrastructure software for massively integrated simulations, to yield further computational efficiencies, and to take advantage of recent advances and availability of multi-core computing platforms. Since engineering analysis, procedures development, and crew familiarity/training for human spaceflight is fundamental to JSC's charter, there is also a strong desire to share and reuse models in both the non-realtime and real-time domains, with the goal of retaining as much multibody dynamics fidelity as possible. Three specific enhancements are reviewed here: (1) linked list organization to address large transitional topologies, (2) body level model order reduction, and (3) parallel formulation/implementation. This paper provides a detailed overview of these primary updates to JSC's flexible multibody dynamics algorithms as well as a comparison of numerical results to previous formulations and associated software.

Multibody dynamics↗
DOE OSTI2025

MoE-Inference-Bench: Performance Evaluation of Mixture of Expert Large Language and Vision Models

Mixture of Experts (MoE) models have enabled the scaling of Large Language Models (LLMs) and Vision Language Models (VLMs) by achieving massive parameter counts while maintaining computational efficiency. However, MoEs introduce several inference-time challenges, including load imbalance across experts and the additional routing computational overhead. To address these challenges and fully harness the benefits of MoE, a systematic evaluation of hardware acceleration techniques is essential. We present MoE-Inference-Bench, a comprehensive study to evaluate MoE performance across diverse scenarios. We analyze the impact of batch size, sequence length, and critical MoE hyperparameters such as FFN dimensions and number of experts on throughput. We evaluate several optimization techniques on Nvidia H100 GPUs, including pruning, Fused MoE operations, speculative decoding, quantization, and various parallelization strategies. Our evaluation includes MoEs from the Mixtral, DeepSeek, OLMoE and Qwen families. The results reveal performance differences across configurations and provide insights for the efficient deployment of MoEs.

Chitty-Venkata, Krishna Teja↗