Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Parallel Architectures”

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 523 records · Page 29
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Decentralized Schemes with Overlap for Solving Graph-Structured Optimization Problems

We present a new algorithmic paradigm for the decentralized solution of graph-structured optimization problems that arise in the estimation and control of network systems. A key and novel design concept of the proposed approach is that it uses overlapping subdomains to promote and accelerate convergence. We show that the algorithm converges if the size of the overlap is sufficiently large and that the convergence rate improves exponentially with the size of the overlap. The proposed approach provides a bridge between fully decentralized and centralized architectures and is flexible in that it enables the implementation of asynchronous schemes, handling of constraints, and balancing of computing, communication, and data privacy needs. The proposed scheme is tested in an estimation problem for a 9241-node power network and we show that it outperforms the alternating direction method of multipliers.

asynchronous↗
DOE OSTI2024

Evolution of the SLATE linear algebra library

SLATE (Software for Linear Algebra Targeting Exascale) is a distributed, dense linear algebra library targeting both CPU-only and GPU-accelerated systems, developed over the course of the Exascale Computing Project (ECP). While it began with several documents setting out its initial design, significant design changes occurred throughout its development. In some cases, these were anticipated: an early version used a simple consistency flag that was later replaced with a full-featured consistency protocol. In other cases, performance limitations and software and hardware changes prompted a redesign. Sequential communication tasks were parallelized; host-to-host MPI calls were replaced with GPU device-to-device MPI calls; more advanced algorithms such as Communication Avoiding LU and the Random Butterfly Transform (RBT) were introduced. Early choices that turned out to be cumbersome, error prone, or inflexible have been replaced with simpler, more intuitive, or more flexible designs. Applications have been a driving force, prompting a lighter weight queue class, nonuniform tile sizes, and more flexible MPI process grids. Of paramount importance has been building a portable library that works across several different GPU architectures – AMD, Intel, and NVIDIA – while keeping a clean and maintainable codebase. Here we explore the evolving design choices and their effects, both in terms of performance and software sustainability.

Gates, Mark↗
DOE OSTI2025

Optimizing transmit field inhomogeneity of parallel RF transmit design in 7T MRI using deep learning

Ultrahigh field (UHF) Magnetic Resonance Imaging (MRI) provides a higher signal-to-noise ratio and, thereby, higher spatial resolution. However, UHF MRI introduces challenges such as transmit radiofrequency (RF) field (B+1) inhomogeneities, leading to uneven flip angles and image intensity anomalies. These issues can significantly degrade imaging quality and its medical applications. This study addresses B+1 field homogeneity through a novel deep learning-based strategy. Traditional methods like Magnitude Least Squares (MLS) optimization have been effective but are time-consuming and dependent on the patient’s presence. Recent machine learning approaches, such as RF Shim Prediction by Iteratively Projected Ridge Regression and deep learning frameworks, have shown promise but face limitations like extensive training times and oversimplified architectures. We propose a two-step deep learning strategy. First, we obtain the desired reference RF shimming weights from multi-channel B+1 fields using random-initialized Adaptive Moment Estimation. Then, we employ Residual Networks (ResNets) to train a model that maps B+1 fields to target RF shimming outputs. Our approach does not rely on pre-calculated reference optimizations for the testing process and efficiently learns residual functions. Comparative studies with traditional MLS optimization demonstrate our method’s advantages in terms of speed and accuracy. The proposed strategy achieves a faster and more efficient RF shimming design, significantly improving imaging quality at UHF. This advancement holds potential for broader applications in medical imaging and diagnostics.

Lu, Zhengyi [Vanderbilt University]↗
DOE OSTI2025

Optimal Polynomial Smoothers and One‐Sided V‐Cycles for Poisson Problems

The solution to the Poisson equation arising from the spectral element discretization of the incompressible Navier‐Stokes equations needs robust preconditioning strategies. One such strategy is multigrid. To realize the potential of multigrid methods, effective smoothing strategies are needed. Chebyshev polynomial smoothers, in conjunction with pointwise Jacobi or additive Schwarz methods (ASMs), prove to be an effective smoother. Other polynomial smoothers, however, may provide superior convergence to the multigrid preconditioner. The authors compare the standard Chebyshev polynomial smoothers to both the novel fourth‐kind Chebyshev polynomial smoothers proposed by Lottes as well as smoothers based on the polynomial of best uniform approximation to as proposed by Kraus, Vassilevski, and Zikatanov. At the cost of symmetry, further improvements may be made. For example, a order polynomial smoother on both sides of the V‐cycle may be substituted with an order polynomial smoother on one side at no additional cost. The choice of omitting the postsmoother in favor of higher‐order polynomial presmoothing is advantageous in cases where the multigrid approximation property constant is large. The authors consider a 2D model problem based on finite differences to motivate the choice of polynomial smoother, order, and whether to apply postsmoothing for the target application of high‐order ‐geometric multigrid methods for GPU architectures. Results from both domains demonstrate the substantial improvement of these approaches over the standard Chebyshev polynomial smoother with a symmetric V‐cycle.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Performance Portability Evaluation of Fluid-Structure Interaction Simulations on Heterogeneous Platforms

The rapid proliferation of heterogeneous programming languages and multi-vendor hardware has underscored the critical need to evaluate the performance portability of scientific applications. In this work, we present the systematic porting and optimization of a massively parallel fluid-structure interaction code across multiple heterogeneous programming frameworks for deployment on leadership-class supercomputers from major vendors. Our analysis focuses on at-scale performance for simulations involving hundreds of millions of deformable cells, executed on a combination of CPUs and GPUs spanning thousands of nodes on exascale machines. We benchmark the performance of each implementation, highlighting the trade-offs inherent in adopting diverse programming models. Key insights regarding the portability of CUDA on multi-vendor platforms, the superior multi-core CPU performance from SYCL, and architectural considerations on performance optimization are distilled from our experience, offering guidance to other users of high performance computing based on our findings.

Martin, Aristotle [Duke University]↗
DOE OSTI2024

Dramatic changes in mitochondrial subcellular location and morphology accompany activation of the CO 2 concentrating mechanism

Dynamic changes in intracellular ultrastructure can be critical for the ability of organisms to acclimate to environmental conditions. Microalgae, which are responsible for ~50% of global photosynthesis, compartmentalize their Ribulose 1,5 Bisphosphate Carboxylase/Oxygenase (Rubisco) into a specialized structure known as the pyrenoid when the cells experience limiting CO 2 conditions; this compartmentalization is a component of the CO 2 Concentrating Mechanism (CCM), which facilitates photosynthetic CO 2 fixation as environmental levels of inorganic carbon (Ci) decline. Changes in the spatial distribution of mitochondria in green algae have also been observed under CO 2 limitation, although a role for this reorganization in CCM function remains unclear. We used the green microalga Chlamydomonas reinhardtii to monitor changes in mitochondrial position and ultrastructure as cells transition between high CO 2 and Low/Very Low CO 2 (LC/VLC). Upon transferring cells to VLC, the mitochondria move from a central to a peripheral cell location and orient in parallel tubular arrays that extend along the cell’s apico-basal axis. We show that these ultrastructural changes correlate with CCM induction and are regulated by the CCM master regulator CIA5. The apico-basal orientation of the mitochondrial membranes, but not the movement of the mitochondrion to the cell periphery, is dependent on microtubules and the MIRO1 protein, with the latter involved in membrane–microtubule interactions. Furthermore, blocking mitochondrial respiration in VLC-acclimated cells reduces the affinity of the cells for Ci. Overall, our results suggest that mitochondrial repositioning functions in integrating cellular architecture and energetics with CCM activities and invite further exploration of how intracellular architecture can impact fitness under dynamic environmental conditions.

CO2 concentrating mechanism↗
DOE OSTI2025

Sn-InAs Nanowire Shadow-Defined Josephson Junctions

Hybrid superconductor–semiconductor platforms are foundational to advancing quantum information technologies, motivating the integration of materials with clean interfaces, robust superconductivity, and scalable architectures. Here, in this work, we report the synthesis and analysis of inclined InAs nanowires, conformally coated with β-Sn shells. These nanowires extend in opposite in-plane directions, forming a self-aligned, criss-cross network. This enables the deterministic formation of nanowire-shadow Josephson junctions through angle-controlled, low-temperature Sn deposition. Structural characterization shows uniform polycrystalline β-Sn shells forming a sharp, diffusion-free interface with InAs. Low-temperature transport measurements reveal a hard induced superconducting gap ≈ 600 μeV, switching currents up to ≈ 500 nA, and parallel magnetic field resilience beyond 1T. These results establish β-Sn/InAs nanowire networks as a promising platform for superconducting qubits, low-noise microwave devices, and the exploration of exotic superconducting phases including triplet pairing and topological superconductivity.

B-Sn↗
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↗
DOE OSTI2024

ELECTRONIC STRUCTURE METHODS AND PROTOCOLS WITH APPLICATION TO DYNAMICS, KINETICS AND THERMOCHEMISTRY

Hydrocarbon combustion involves the reaction dynamics of a tremendous number of species beginning with many-component fuel mixtures and proceeding via a complex system of intermediates to form primary and secondary products. Combustion conditions corresponding to new advanced engines and/or alternative fuels rely increasingly on autoignition and low-temperature-combustion chemistry. In these regimes various transient radical species such as HO2, ROO·, ·QOOH, HCO, NO2, HOCO, and Criegee intermediates play important roles in determining the detailed as well as more general dynamics. A clear understanding and accurate representation of these processes is needed for effective modeling. Given the difficulties associated with making reliable experimental measurements of these systems, computation can play an important role in developing these energy technologies. Accurate calculations have their own challenges since even within the simplest dynamical approximations such as transition state theory, the rates depend exponentially on critical barrier heights and these may be sensitive to the level of quantum chemistry. Moreover, it is well-known that in many cases it is necessary to go beyond statistical theories and consider the dynamics. Quantum tunneling, resonances, radiative transitions, and non-adiabatic effects governed by spin-orbit or derivative coupling can be determining factors in those dynamics. Building upon progress made during a period of prior support through the DOE Early Career Program, this project combines developments in the areas of potential energy surface (PES) fitting and multistate multireference quantum chemistry to allow spectroscopically and dynamically/kinetically accurate investigations of key molecular systems (such as those mentioned above), many of which are radicals with strong multireference character and have the possibility of multiple electronic states contributing to the observed dynamics. An ongoing area of investigation is to develop general strategies for robustly convergent electronic structure theory for global multichannel reactive surfaces including diabatization of energy and other relevant surfaces such as dipole transition. Combining advances in ab initio methods with automated interpolative PES fitting allows the construction of high-quality PESs (incorporating thousands of high-level data) to be done rapidly through parallel processing on high-performance computing (HPC) clusters. In addition, new methods and approaches to electronic structure theory will be developed and tested through applications. This project will explore limitations in traditional multireference calculations (e.g., MRCI) such as those imposed by internal contraction, lack of high-order correlation treatment and poor scaling. Methods such as DMRG-based extended active-space CASSCF and various Quantum Monte Carlo (QMC) methods will be applied (including VMC/DMC and FCIQMC). Insight into the relative significance of different orbital spaces and the robustness of application of these approaches on leadership class computing architectures will be gained. Synergy with other components of this research program such as automated PES fitting and multireference quantum chemistry will be used to address challenges encountered by the standard approaches to computational thermochemistry (those being single-reference quantum chemistry and perturbative treatments of the anharmonic vibrational energy, which break down for some cases of electronic structure or floppy strongly coupled vibrational modes).

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2025

Knowledge Oriented Graph Unified Transformer (KOGUT) v0.1

KOGUT — Knowledge Oriented Graph Unified Transformer KOGUT implements the Relational Graph Transformer (RelGT) architecture for knowledge graph link prediction in biological domains, with a primary focus on microbial growth media prediction. While the original RelGT (arXiv:2505.10960) targets relational tables, time series, and multi-table databases, KOGUT adapts this architecture for heterogeneous biological knowledge graphs, providing first-in-class AI predictive models for microbial cultivation. Key Adaptations Beyond Original RelGT: - Knowledge Graph Focus: Applied to biological KGs with semantic node types (taxa, chemicals, media, phenotypes, environments) versus generic relational database tables, trained on the KG-Microbe knowledge graph (1.3M entities, 2.9M edges, 24 relation types). - Multimodal Node Encoding: Integrates node labels, categories, descriptions, and synonyms from KG metadata through learned embedding layers—adapting relational column features to graph node attributes with textual semantics. - Extended K-Hop Subgraph Strategy: Optimized neighborhood sampling (3-hop default, configurable up to 200 nodes) tuned for sparse biological networks, building on the original local-global attention framework with biological relation preservation. - Biolink Predicate Preservation: Type-specific transformations for 24 biological edge semantics (occurs_in, consumes, produces, has_phenotype, subclass_of) beyond standard relational foreign keys, enabling multi-relation link prediction. - Inductive Learning Support: Enables zero-shot predictions for novel taxa through feature-based embeddings (temperature, oxygen requirements, gram stain, cell shape), extending the original transductive relational benchmark scope to uncultured microorganisms. CheapSOTA Performance Optimizations (This Distribution): - VQ-EMA Centroid Attention: Vector quantization with exponential moving average for improved global context modeling (+5-10% MRR improvement). - HDF5 Precomputed Data Loading: One-time preprocessing of k-hop subgraphs to eliminate redundant graph traversals (2-5× training speedup). - Distributed Data Parallel Training: Multi-GPU support for scaling to larger knowledge graphs (tested on 4× NVIDIA A100 GPUs at NERSC Perlmutter). - Mixed Precision Training: Automatic mixed precision (AMP) for memory efficiency and faster training. Advantages Over Standard Knowledge Graph Embedding Models: Combines RelGT's proven multi-element tokenization (features, type, hop, structure) with graph-native biological representations, enabling interpretable link prediction across heterogeneous entities that standard embedding models (TransE, RotatE, ComplEx) and table-based transformers cannot directly model. Achieves near-perfect performance on microbial growth media prediction (MRR: 0.9966, Precision@1: 0.9932, Hit@10: 1.0000) while maintaining explainability through attention-based reasoning over biological pathways. Training Data: - KG-Microbe merged knowledge graph: 1,379,337 nodes, 2,960,472 edges - 24 biological relation types including taxonomic hierarchies, metabolic interactions, phenotype associations, and environmental relationships - Primary prediction task: Growth media suitability for microbial taxa (biolink:occurs_in, 50K edges) - Multi-relation capability: Predicts links for any of the 24 relation types, including chemical consumption/production, phenotype associations, and taxonomic classification Citation: Original RelGT Architecture: Dwivedi et al., "Relational Graph Transformer", arXiv:2505.10960, 2025 KOGUT Implementation: Knowledge Oriented Graph Unified Transformer for Microbial Growth Media Prediction Developed at Lawrence Berkeley National Laboratory (LBNL) Trained on NERSC Perlmutter supercomputer

Joachimiak, Marcin [Lawrence Berkeley National Lab↗
DOE OSTI2023

HIPLZ: Enabling performance portability for exascale systems

While heterogeneous computing has emerged as a dominant trend in current and future High-Performance Computing (HPC) systems, it is also widely recognized that this shift has led to increased software complexity due to a proliferation of programming systems for different heterogeneous processors. One such example is the Heterogeneous-Compute Interface for Portability from AMD (HIP ), which is composed of a C Runtime API and C++ Kernel Language. Many HPC applications will likely use HIP on future exascale systems (e.g., Frontier and El Capitan), but HIP currently only targets AMD and NVIDIA processors. This limitation creates challenges for users who would also like to run their applications on exascale systems based on other architectures (e.g., Aurora, which is based on Intel hardware) that are currently not targeted by HIP . In this paper, we introduce the design and implementation of HIPLZ , a compiler and runtime system that uses the Intel Level Zero API to support HIP on Intel GPU architectures. We discuss the design of HIPLZ , derived from HIPCL (an implementation of HIP on top of OpenCL ), and portability issues that occur from using the Level Zero runtime as a backend. We evaluate our implementation by running several performance benchmarks and mini-apps written in HIP on Intel architectures using HIPLZ . Our results show that this approach provides competitive performance relative to Intel's OpenCL implementations on Intel Gen9 and UHD Graphics 770 GPUs, while providing good coverage of features needed by HPC applications. Overall, this approach is a promising demonstration of enabling performance portability for exascale systems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Lab-Scale Cable-Driven Parallel Robot Prototype for Automated Prefabricated Component Manipulation

This paper presents the design and evaluation of a lab-scale cable-driven parallel robot (CDPR) developed as a flexible platform for automated installation of prefabricated components onto exterior building envelopes. Traditional manual installation methods for prefabricated components, which depend on scaffolding, cranes, cherry pickers, and verbal coordination, are not only labor-intensive and error-prone but also face significant limitations in dense urban environments due to site access constraints. To address these challenges, we developed a lab-scale CDPR platform capable of autonomously transporting building envelope components from a designated pickup zone to their target installation location, minimizing the need for human intervention. This study describes the system’s mechanical design, actuation architecture, real-time feedback system, and control strategy of the CDPR, and evaluates its performance in a laboratory environment. The robot’s actuation system uses torque control for end-effector manipulation. The robot’s real-time pose feedback comes from a construction-grade total station and a wireless inertial measurement unit (IMU), which together support precise end-effector control. Experimental results demonstrate the successful integration of the hardware, sensing, state estimation, and control subsystems. Preliminary tests showed that our lab-scale prototype can position the end effector with an error of less than 3 mm, which is a level of precision not previously achieved by existing CDPRs in construction applications. The key findings are twofold: (1) torque-only control is necessary but not sufficient for minimizing final pose error, and (2) incorporating real-time pose feedback can achieve the desired placement accuracy.

Liu, Yifang [Oak Ridge National Laboratory (ORNL),↗
DOE OSTI2021

True Load Balancing for Matricized Tensor Times Khatri-Rao Product

MTTKRP is the bottleneck operation in algorithms used to compute the CP tensor decomposition. For sparse tensors, utilizing the compressed sparse fibers (CSF) storage format and the CSF-oriented MTTKRP algorithms is important for both memory and computational efficiency on distributed-memory architectures. Existing intelligent tensor partitioning models assume the computational cost of MTTKRP to be proportional to the total number of nonzeros in the tensor. However, this is not the case for the CSF-oriented MTTKRP on distributed-memory architectures. We outline two deficiencies of nonzero-based intelligent partitioning models when CSF-oriented MTTKRP operations are performed locally: failure to encode processors' computational loads and increase in total computation due to fiber fragmentation. We focus on existing fine-grain hypergraph model and propose a novel vertex weighting scheme that enables this model encode correct computational loads of processors. We also propose to augment the fine-grain model by fiber nets for reducing the increase in total computational load via minimizing fiber fragmentation. In this way, the proposed model encodes minimizing the load of the bottleneck processor. In conclusion, parallel experiments with real-world sparse tensors on up to 1024 processors prove the validity of the outlined deficiencies and demonstrate the merit of our proposed improvements in terms of parallel runtimes.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Expanded Capabilities and Opportunities for Virtual Beamline Code

In the 25 years since LLNL broke ground on the National Ignition Facility (NIF) - a cornerstone of the National Nuclear Security Administration’s Stockpile Stewardship Program - the NIF and Photon Science (NIF&PS) Directorate has steadily pushed the boundaries of laser physics, nonlinear optics, and photonics in service of inertial confinement fusion experiments, and advanced photon sources development. Beyond the 192 beamlines of the primary laser facility with its world-record energy and fusion yields, NIF&PS capabilities have expanded to include chirped-pulse amplification, kilojoule petawatt-class short-pulse systems to generate hard x-ray radiographic probes, high-average power lasers as secondary sources for generation or wake field particle acceleration along directed energy applications, and other scientific pursuits. The Computing Directorate partners with NIF&PS on numerous projects and technologies from diagnostic measurements and high performance control systems to data analysis, information technology infrastructure, and scientific simulation code development. “This strong collaboration enables cutting-edge science, particularly with regard to physics-based modeling and simulation,” says physicist Jean-Michel Di Nicola. One crucial capability for NIF&PS is the Virtual Beamline (VBL) laser simulation code. It can model all the major laser physics and technology involved in the design optimization, commissioning, and operations of advanced laser architectures from tabletop to NIF-scale. VBL provides researchers with high-fidelity models and high-resolution calculations of laser performance predictions—including for the entire NIF laser system, the Advanced Radiographic Capability, parts of the High Repetition-Rate Advanced Petawatt Laser System laser delivered to the Czech Republic, and the Optical Science Laser. After more than two decades of experimentally verified physics and computing enhancements, this unique workhorse code is wrapping up another major milestone: migration from Java to C++ with a flurry of user interface and optimization features, as well as laser physics and high-resolution enhancements thanks to parallel execution on Livermore Computing platforms.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2025

Antielectrophoretic Response-Driven Bending–Tilting Deformation of Cationic Polyelectrolyte Brushes Drives Nonlinear Electroosmotic Transport in Brush-Grafted Nanochannels

In this paper, we use all-atom molecular dynamics (MD) simulations to describe a non-linearly enhanced electroosmotic (EOS) flow, where, in a nanochannel grafted with cationic PMETAC ([Poly(2-(Methacryloyloxy)Ethyl) Trimethylammonium Chloride]) brushes, a two-fold increase in the electric field strength leads to a several-fold (more than two-fold) increase in the EOS flow strength and volume flow rate. The electric field enforces the PMETAC brushes to undergo a bending-tilting driven deformation with a significant portion of the brush layer becoming parallel to the grafting surface. In response, a substantial fraction of the counterions leave the brush layer (hence become more mobile), but instead of going into the bulk, accumulate at the brush-bulk interface, i.e., stay in proximity of the brush segments aligned parallel to the grafting surface. This creates an interesting situation, where the counterions are not completely within the brush layer, yet they fully screen the brush charges. Such “freer” conditions enable the counterions to achieve very high velocity, thereby ensuring that the water solvating the counterions themselves move very fast triggering the significantly augmented EOS transport. Probing deeper we can identify that the bending-tilting driven brush deformation, enforcing the brushes to align parallel to the substrate, results from the anti-electrophoretic behavior of the brushes, where despite being positively charged, the brushes move against the electric field direction. Such an anti-electrophoretic behavior of the PE brushes, which has not been reported before, can be associated with the very fast velocities of the negatively charged counterions and the electrostatic and hydrodynamic coupling of the counterions with the positive functional groups of the brushes. Here, we anticipate that the findings of this paper will shed light on strategies for nanochannel flow, the anti-electrophoretic response of charged polymer chains, and the significance of capturing the detailed chemical architecture of polyelectrolytes in nanoscale science and engineering.

36 MATERIALS SCIENCE↗
DOE OSTI2023

DGaaS: GPU as a Service on Distributed Computing System

In the rapidly evolving landscape of scientific computing, Graphics Processing Units (GPUs) have become indispensable for their unparalleled ability to handle parallel tasks in complex calculations, simulations, and data analysis. Their utility is further magnified in machine learning and AI applications, where they significantly accelerate model training and predictive analytics. Within this context, the Triton Inference Server emerges as a pivotal open-source tool, specializing in AI inferencing and optimizing GPU utilization across various platforms and frameworks. This paper presents an in-depth study on distributed High Throughput Computing (HTC), specifically focusing on the HTCondor framework and its resource provisioning tools, GlideinWMS and HEPCloud. These systems enable large-scale scientific experiments like CMS and DUNE to efficiently access and utilize vast computational resources. The paper explores the core architectural components of GlideinWMS, including jobs, user pools, and worker nodes, and discusses their integration with GPUs and the Triton server. The primary aim of this research is to develop a solution that optimizes GPU utilization by leveraging Glideins and containers. This approach allows computational jobs, particularly those involving AI models, to use GPUs only when essential, thereby facilitating efficient sharing of limited GPU resources. To validate this architecture, the study conducted three key tests involving custom scripts, container-based servers, and Triton server deployments. However, the study faces challenges, notably in locating the Triton server and ensuring secure remote access. To address these issues, future work will focus on developing a proxy mechanism and enhancing security protocols. In conclusion, this study offers a comprehensive roadmap for effective and efficient GPU utilization in distributed High Throughput Computing. It aims to contribute significantly to the scientific community by solving pressing problems and implementing robust solutions in collaboration with the GlideinWMS and HEPCloud teams. The research sets the stage for a more efficient, scalable, and cost-effective paradigm in scientific computing.

97 MATHEMATICS AND COMPUTING↗
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 OSTI2020

Long-Range Lamellar Alignment in Diblock Bottlebrush Copolymers via Controlled Oscillatory Shear

A simple strategy is presented to achieve well-ordered nanostructures in microphase-separated diblock bottlebrush copolymers (dbBB) for potential opportunities in nanotechnology that require large grain sizes and directed orientation. The unique architecture and relaxation processes of dbBBs offer a pathway to enhanced dynamic ordering and reorientation over macroscopic scales. Here, controlled shear is used to align lamellar domains. The high molecular mobility of dbBBs in the melt state, as shown by linear and nonlinear viscoelastic characterization, supports this alignment process. When subjected to large-amplitude oscillatory shear at controlled frequency (ω), strain amplitude (γ), and temperature (T shear ), microphase-separated lamellae transition into highly aligned states as determined by small-angle X-ray scattering (SAXS) and transmission electron microscopy (TEM). High throughput synchrotron SAXS is used to rapidly characterize microphase separation and lamellar alignment across the entire bulk sample. Lamellae align parallel to shear planes with an orientation order parameter (S) approaching S = 0.8 over a large cubic volume (V ~ mm 3 ).

36 MATERIALS SCIENCE↗