Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “architecture optimization”

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 253 records · Page 14
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Achieving performance portability in Gaussian basis set density functional theory on accelerator based architectures in NWChemEx

The numerical integration of the exchange–correlation (XC) potential is one of the primary computational bottlenecks in Gaussian basis set Kohn–Sham density functional theory (KS-DFT). To achieve optimal performance and accuracy, care must be taken in this numerical integration to preserve local sparsity as to allow for near linear weak scaling with system size. This leads to an integration scheme with several performance critical kernels which must be hand optimized for each architecture of interest. As the set of available accelerator hardware goes more diverse, a key challenge for developers of KS-DFT software is to maintain performance portability across a wide range of computational architectures. In this article, we examine a modular software design pattern which decouples the implementation details of performance critical kernels from the expression of high-level algorithmic workflows in a device-agnostic language such as C++; thus allowing for developers to target existing and emerging accelerator hardware within a single code base. We consider the efficacy of such a design pattern in the numerical integration of the XC potential by demonstrating its ability to achieve performance portability across a set of accelerator architectures which are representative of those on current and future U.S. Department of Energy Leadership Computing Facilities.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Optimizing Performance of Combustion Chemistry Solvers on Intel's Many Integrated Core (MIC) Architectures

This work investigates novel algorithm designs and optimization techniques for restructuring chemistry integrators in zero and multidimensional combustion solvers, which can then be effectively used on the emerging generation of Intel's Many Integrated Core/Xeon Phi processors. These processors offer increased computing performance via large number of lightweight cores at relatively lower clock speeds compared to traditional processors (e.g. Intel Sandybridge/Ivybridge) used in current supercomputers. This style of processor can be productively used for chemistry integrators that form a costly part of computational combustion codes, in spite of their relatively lower clock speeds. Performance commensurate with traditional processors is achieved here through the combination of careful memory layout, exposing multiple levels of fine grain parallelism and through extensive use of vendor supported libraries (Cilk Plus and Math Kernel Libraries). Important optimization techniques for efficient memory usage and vectorization have been identified and quantified. These optimizations resulted in a factor of ~ 3 speed-up using Intel 2013 compiler and ~ 1.5 using Intel 2017 compiler for large chemical mechanisms compared to the unoptimized version on the Intel Xeon Phi. The strategies, especially with respect to memory usage and vectorization, should also be beneficial for general purpose computational fluid dynamics codes.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Enabling high areal capacity for Co-free high voltage spinel materials in next-generation Li-ion batteries

The rapidly growing technological demand for lithium-ion batteries has prompted the development of novel cathode materials with high energy density, low cost, and improved safety. High voltage spinel, LiNi 0.5 Mn 1.5 O 4 (LNMO), is one of the most promising candidates yet to be commercialized. The two primary obstacles for this material are the inferior electronic conductivity and fast capacity degradation in full cells due to the high operating voltage. By systematically addressing these limitations, we successfully develop a thick LNMO electrode with areal capacity loadings up to 3 mAh·cm –2 . The optimized thick electrode is paired with a commercial graphite anode at both the coin cell and pouch cell level, achieving a full cell capacity retention as high as 72% and 78%, respectively, after 300 cycles. We attribute this superior cycling stability to careful optimizations of cell components and testing conditions, with a specific focus improving electronic conductivity and high voltage compatibility. These results suggest precise control of materials quality, electrode architecture and electrolyte optimization can soon support the development of a cobalt-free battery system based on a thick LNMO cathode (>4 mAh·cm 2 ), which will eventually meet the needs of next-generation Li-ion batteries with reduced cost, improved safety, and assured sustainability.

25 ENERGY STORAGE↗
NASA NTRS1996

Design of Power System Architectures for Small Spacecraft Systems

The objective of this research is to perform a trade study on several candidate power system architectures for small spacecrafts to be used in NASA's new millennium program. Three initial candidate architectures have been proposed by NASA and two other candidate architectures have been proposed by Howard University. Howard University is currently conducting the necessary analysis, synthesis, and simulation needed to perform the trade studies and arrive at the optimal power system architecture. Statistical, sensitivity and tolerant studies has been performed on the systems. It is concluded from present studies that certain components such as the series regulators, buck-boost converters and power converters can be minimized while retaining the desired functionality of the overall architecture. This in conjunction with battery scalability studies and system efficiency studies have enabled us to develop more economic architectures. Future studies will include artificial neural networks and fuzzy logic to analyze the performance of the systems. Fault simulation studies and fault diagnosis studies using EMTP and artificial neural networks will also be conducted.

Momoh, James A.↗
DOE OSTI2021

Review of Literature Related to Nuclear Data Mini-Apps

his document reviews nuclear data mini-apps prevalent in the published scientific literature. It is intended to identify algorithmic and computer architecture considerations of relevance to optimizing GIDI/MCGIDI for GPU-based architectures. While some of the works considered here focus on CPU-based architectures, the performance-limiting factors and mitigating strategies identified often extend to GPU-based architectures, particularly in the context of latency-bound performance. A brief overview of the cross section (XS) lookup process is given in Section 1 and a detailed analysis of nuclear data mini-apps follows in Section 2. Concluding remarks are offered in Section 3.

42 ENGINEERING↗
DOE OSTI2026

CRADA Number NFE-24-10110 with Qubit Engineering Inc. (CRADA Final Report)

Over the past year, the Qubit Engineering team has pushed the frontiers of power‑grid optimization, working in close collaboration with Oak Ridge National Laboratory (ORNL) and the Tennessee Valley Authority (TVA). Their progress is reflected in three newly submitted conference papers, “Unified Relational GNN Architecture for AC Optimal Power Flow Calculations in Electric Grids,” “Graph‑Based Attention Mechanisms for Solving the AC Optimal Power Flow Problem in Electrical‑Power Networks,” and “Enhanced Power‑Grid Maintenance Planning and Quantum‑Inspired Combinatorial Prospects.” These publications showcase state‑of‑the‑art graph‑neural‑network methods for AC‑OPF and novel quantum‑inspired heuristics for maintenance scheduling. Beyond the academic results, the Qubit team has converted the research into two production‑grade tools built on TVA data: Neuro‑Grid, an AI‑driven power‑flow simulator that provides instant, interactive full‑grid load‑flow visualizations, and Quanta‑Grid, a quantum‑inspired maintenance‑scheduling engine to support logistics optimization for power utilities. Together, these advances demonstrate how Qubit’s partnership with ORNL and TVA is delivering practical, physics‑grounded analytics for next‑generation grid management.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRS1992

Formalization and visualization of domain-specific software architectures

This paper describes a domain-specific software design system based on the concepts of software architectures engineering and domain-specific models and languages. In this system, software architectures are used as high level abstractions to formulate a domain-specific software design. The software architecture serves as a framework for composing architectural fragments (e.g., domain objects, system components, and hardware interfaces) that make up the knowledge (or model) base for solving a problem in a particular application area. A corresponding software design is generated by analyzing and describing a system in the context of the software architecture. While the software architecture serves as the framework for the design, this concept is insufficient by itself for supplying the additional details required for a specific design. Additional domain knowledge is still needed to instantiate components of the architecture and develop optimized algorithms for the problem domain. One possible way to obtain the additional details is through the use of domain-specific languages. Thus, the general concept of a software architecture and the specific design details provided by domain-specific languages are combined to create what can be termed a domain-specific software architecture (DSSA).

Bailor, Paul D.↗
DOE OSTI2017

Can We Practically Bring Physics-based Modeling Into Operational Analytics Tools?

We present that analytics software is increasingly used to improve and maintain operational efficiency in commercial buildings. Energy managers, owners, and operators are using a diversity of commercial offerings often referred to as Energy Information Systems, Fault Detection and Diagnostic (FDD) systems, or more broadly Energy Management and Information Systems, to cost-effectively enable savings on the order of ten to twenty percent. Most of these systems use data from meters and sensors, with rule-based and/or data-driven models to characterize system and building behavior. In contrast, physics-based modeling uses first-principles and engineering models (e.g., efficiency curves) to characterize system and building behavior. Historically, these physics-based approaches have been used in the design phase of the building life cycle or in retrofit analyses. Researchers have begun exploring the benefits of integrating physics-based models with operational data analytics tools, bridging the gap between design and operations. In this paper, we detail the development and operator use of a software tool that uses hybrid data-driven and physics-based approaches to cooling plant FDD and optimization. Specifically, we describe the system architecture, models, and FDD and optimization algorithms; advantages and disadvantages with respect to purely data-driven approaches; and practical implications for scaling and replicating these techniques. Finally, we conclude with an evaluation of the future potential for such tools and future research opportunities.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
NASA NTRSDate not supplied

Performance and Portability of a Linear Solver Across Emerging Architectures

A linear solver algorithm used by a large-scale unstructured-grid computational fluid dynamics application is examined for a broad range of familiar and emerging architectures. Efficient implementation of a linear solver is challenging on recent CPUs offering vector architectures. Vector loads and stores are essential to effectively utilize available memory bandwidth on CPUs, and maintaining performance across different CPUs can be difficult in the face of varying vector lengths offered by each. A similar challenge occurs on GPU architectures, where it is essential to have coalesced memory accesses to utilize memory bandwidth effectively. In this work, we demonstrate that restructuring a computation, and possibly data layout, with regard to architecture is essential to achieve optimal performance by establishing a performance benchmark for each target architecture in a low level language such as vector intrinsics or CUDA. In doing so, we demonstrate how a linear solver kernel can be mapped to Intel® Xeon™ and Xeon Phi™, Marvell® ThunderX2®, NEC® SX-Aurora™ TSUBASA Vector Engine, and NVIDIA® and AMD® GPUs. We further demonstrate that the required code restructuring can be achieved in higher level programming environments such as OpenACC, OCCA, and Intel® OneAPI™/SYCL, and that each generally results in optimal performance on the target architecture. Relative performance metrics for all implementations are shown, and subjective ratings for ease of implementation and optimization are suggested.

Programming models↗
DOE OSTI2023

Ready for the Frontier: Preparing Applications for the World’s First Exascale System

Frontier, a supercomputer at the Oak Ridge Leadership Computing Facility (OLCF), debuted atop the Top500 list of the world’s most powerful supercomputers in June 2022 as the very first computer to produce exascale performance. Making sure scientific applications are optimized on this architecture is the critical link necessary to translate the newly available computational power into scientific insight and solutions. To that goal, the OLCF developed the Center for Accelerated Application Readiness (CAAR) program to ensure that a suite of highly optimized applications is ready for scientific runs at the onset of production operations for Frontier. This paper describes our experience in porting and optimizing such suite of applications in the OLCF’s CAAR program.

Budiardja, Reuben↗
DOE OSTI2019

Fine-tuning the photosynthetic light harvesting apparatus for improved photosynthetic efficiency and biomass yield

Photosynthetic electron transport rates in higher plants and green algae are light-saturated at approximately one quarter of full sunlight intensity. This is due to the large optical cross section of plant light harvesting antenna complexes which capture photons at a rate nearly 10-fold faster than the rate-limiting step in electron transport. As a result, 75% of the light captured at full sunlight intensities is reradiated as heat or fluorescence. Previously, it has been demonstrated that reductions in the optical cross-section of the light-harvesting antenna can lead to substantial improvements in algal photosynthetic rates and biomass yield. By surveying a range of light harvesting antenna sizes achieved by reduction in chlorophyll b levels, we have determined that there is an optimal light-harvesting antenna size that results in the greatest whole plant photosynthetic performance. We also uncover a sharp transition point where further reductions or increases in antenna size reduce photosynthetic efficiency, tolerance to light stress, and impact thylakoid membrane architecture. Plants with optimized antenna sizes are shown to perform well not only in controlled greenhouse conditions, but also in the field achieving a 40% increase in biomass yield.

09 BIOMASS FUELS↗
NASA NTRSDate not supplied

A New Architecture for Parallelization of Complex Spacecraft Trajectory Optimization Scans

This paper describes CopScanner, a new component of the Copernicus ecosystem for spacecraft trajectory design and optimization. CopScanner is a Python library being developed at the NASA JSC which enables easy parallelization of Copernicus scans. CopScanner is currently being developed and implemented for production of Copernicus trajectory scans for upcoming Artemis Missions (Artemis II and beyond). On the backend, CopScanner utilizes Dask, an open-source Python library for parallel computing which enables parallelization over both multi-core local machines and large-scale distributed computing clusters. CopScanner abstracts the trajectory scanning process into a DAG which is constructed using a chain of individual subscans. Each node in the DAG executes a python module, called the callable, for which there are built-in defaults, or users may specify their own. Support for custom callables makes CopScanner a versatile trajectory optimization software. All output files and associated metadata from a CopScanner scan are compressed and stored in a two-file output, collectively called the FileStore, consisting of a SQLite database and a compressed JSON MessagePack file, for which CopScanner provides a Python class for interaction.

Quentin Moore↗
DOE OSTI2019

Beyond 30% Conversion Efficiency in Silicon Solar Cells: A Numerical Demonstration

We demonstrate through precise numerical simulations the possibility of flexible, thin-film solar cells, consisting of crystalline silicon, to achieve power conversion efficiency of 31%. Our optimized photonic crystal architecture consists of a 15 μm thick cell patterned with inverted micro-pyramids with lattice spacing comparable to the wavelength of near-infrared light, enabling strong wave-interference based light trapping and absorption. Unlike previous photonic crystal designs, photogenerated charge carrier flow is guided to a grid of interdigitated back contacts with optimized geometry to minimize Auger recombination losses due to lateral current flow. Front and back surface fields provided by optimized Gaussian doping profiles are shown to play a vital role in enhancing surface passivation. We carefully delineate the drop in power conversion efficiency when surface recombination velocities exceed 100 cm/s and the doping profiles deviate from prescribed values. These results are obtained by exact numerical simulation of Maxwell’s wave equations for light propagation throughout the cell architecture and a state-of-the-art model for charge carrier transport and Auger recombination.

14 SOLAR ENERGY↗
DOE OSTI2018

Attacking the Opioid Epidemic: Determining the Epistatic and Pleiotropic Genetic Architectures for Chronic Pain and Opioid Addiction

We describe the CoMet application for large-scale epistatic Genome-Wide Association Studies (eGWAS) and pleiotropy studies. High performance is attained by transforming the underlying vector comparison methods into highly performant generalized distributed dense linear algebra operations. The 2-way and 3-way Proportional Similarity metric and Custom Correlation Coefficient are implemented using native or adapted GEMM kernels optimized for GPU architectures. By aggressive overlapping of communications, transfers and computations, high efficiency with respect to single GPU kernel performance is maintained up to the full Titan and Summit systems. Nearly 300 quadrillion element comparisons per second and over 2.3 mixed precision ExaOps are reached on Summit by use of Tensor Core hardware on the Nvidia Volta GPUs. Performance is four to five orders of magnitude beyond comparable state of the art. CoMet is currently being used in projects ranging from bioenergy to clinical genomics, including for the genetics of chronic pain and opioid addiction.

Joubert, Wayne↗
DOE OSTI2026

Random 3D Interpenetrating Electrode Design for Energy Storage Applications

We introduce the concept of random interpenetrating electrode structures, inspired by spinodal decomposition, as a transformative approach for energy storage applications. While ordered architectures, such as double gyroids and Kelvin cell structures, have long been favored for their structural uniformity and surface area utilization, we demonstrate that random structures offer distinct advantages including shorter diffusion pathways, higher packing density, and enhanced reaction kinetics, particularly at low temperatures. Using a combination of computational and electrochemical analysis, we show that random architecture outperforms their ordered counterparts in scenarios where diffusion dominates. This work represents the first systematic exploration of random structures in energy storage, highlighting their potential to overcome limitations of traditional designs. However, challenges such as structural reproducibility and scalability remain, necessitating further investigation into fabrication techniques and material selection. By establishing a framework to optimize random electrode architectures, we provide critical insights into the interplay between geometry, diffusion, and reaction kinetics. This study not only introduces a new design principle for 3D electrodes but also opens pathways for next-generation energy storage systems that demand both high performance and adaptability to a range of operating conditions.

Materials science↗
DOE OSTI2018

Scaling Deep Learning workloads: NVIDIA DGX-1/Pascal and Intel Knights Landing

Deep Learning (DL) algorithms have become ubiquitous in data analytics. As a result, major computing vendors—including NVIDIA, Intel, AMD, and IBM—have architectural road maps influenced by DL workloads. Furthermore, several vendors have recently advertised new computing products as accelerating large DL workloads. Unfortunately, it is difficult for data scientists to quantify the potential of these different products. Here, this article provides a performance and power analysis of important DL workloads on two major parallel architectures: NVIDIA DGX-1 (eight Pascal P100 GPUs interconnected with NVLink) and Intel Knights Landing (KNL) CPUs interconnected with Intel Omni-Path or Cray Aries. Our evaluation consists of a cross section of convolutional neural net workloads: CifarNet, AlexNet, GoogLeNet, and ResNet50 topologies using the Cifar10 and ImageNet datasets. The workloads are vendor-optimized for each architecture. We use sequentially equivalent implementations to maintain iso-accuracy between parallel and sequential DL models. Our analysis indicates that although GPUs provide the highest overall performance, the gap can close for some convolutional networks; and the KNL can be competitive in performance/watt. We find that NVLink facilitates scaling efficiency on GPUs. However, its importance is heavily dependent on neural network architecture. Furthermore, for weak-scaling—sometimes encouraged by restricted GPU memory—NVLink is less important.

97 MATHEMATICS AND COMPUTING↗