Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “bottleneck analysis”

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 433 records · Page 24

Scalable and accurate multi-GPU-based image reconstruction of large-scale ptychography data

Abstract While the advances in synchrotron light sources, together with the development of focusing optics and detectors, allow nanoscale ptychographic imaging of materials and biological specimens, the corresponding experiments can yield terabyte-scale volumes of data that can impose a heavy burden on the computing platform. Although graphics processing units (GPUs) provide high performance for such large-scale ptychography datasets, a single GPU is typically insufficient for analysis and reconstruction. Several works have considered leveraging multiple GPUs to accelerate the ptychographic reconstruction. However, most of these works utilize only the Message Passing Interface to handle the communications between GPUs. This approach poses inefficiency for a hardware configuration that has multiple GPUs in a single node, especially while reconstructing a single large projection, since it provides no optimizations to handle the heterogeneous GPU interconnections containing both low-speed (e.g., PCIe) and high-speed links (e.g., NVLink). In this paper, we provide an optimized intranode multi-GPU implementation that can efficiently solve large-scale ptychographic reconstruction problems. We focus on the maximum likelihood reconstruction problem using a conjugate gradient (CG) method for the solution and propose a novel hybrid parallelization model to address the performance bottlenecks in the CG solver. Accordingly, we have developed a tool, called PtyGer ( Pty chographic G PU(multipl e )-based r econstruction), implementing our hybrid parallelization model design. A comprehensive evaluation verifies that PtyGer can fully preserve the original algorithm’s accuracy while achieving outstanding intranode GPU scalability.

97 MATHEMATICS AND COMPUTING↗

U.S. Department of Energy: National Virtual Biotechnology Laboratory (Technical Report)

With funding from the CARES Act, the U.S Department of Energy (DOE) established the National Virtual Biotechnology Laboratory (NVBL) in March 2020 to address key challenges associated with the COVID-19 crisis. The NVBL brought together the broad scientific and technical expertise and resources of DOE’s 17 national laboratories to help tackle medical supply shortages, discover potential drugs to fight the virus, develop and validate COVID-19 testing methods, model disease spread and impact across the nation, and understand virus transport in buildings and the environment. National laboratory resources leveraged for this effort include a suite of world-leading user facilities broadly available to the research community, such as light and neutron sources, nanoscale science research centers, sequencing and biocharacterization facilities, and high-performance computing facilities. As part of the NVBL framework, DOE rapidly assembled five project teams to (1) identify new targets for medical therapeutics; (2) develop innovations in testing capabilities; (3) provide epidemiological and logistical support; (4) understand viral fate and transport in the environment; and (5) address supply chain bottlenecks by harnessing extensive additive manufacturing capabilities. Each research team was charged with defining high-impact projects that could be completed in a 6-month sprint while coordinating their developments with academia, other government agencies, and the private sector. Within months, NVBL teams used DOE’s high-performance computers and light and neutron sources to identify promising candidates for antibodies and antivirals that universities and drug companies are now evaluating. NVBL researchers also developed new diagnostic targets and sample collection approaches, and supported efforts by the U.S. Food and Drug Administration, Centers for Disease Control and Prevention, and U.S. Department of Defense to establish national guidelines used in administering millions of tests. Researchers used artificial intelligence and high-performance computing to produce near-real-time data analysis to forecast disease transmission, stress on public health infrastructure, and economic impact, which supported decision-makers at the local, state, and national levels. To minimize virus uptake and protect human health, NVBL teams studied how to control indoor virus movement. Researchers also produced innovations in materials and advanced manufacturing that mitigated shortages in test kits and personal protective equipment, creating nearly 1,000 new jobs. Through its NVBL framework, DOE has contributed significantly to the nation’s COVID response, demonstrating in only a few months the critical impact of its national laboratories. NVBL’s accomplishments demonstrate not only the powerful resource represented by DOE’s national laboratories working together to meet national needs, but also the effectiveness of the integrated NVBL framework for rapidly responding to emergencies with research and development solutions. Going forward, the NVBL is poised to apply the unique capabilities and expertise of the national laboratory complex to future national and international emergencies, both natural and engineered. Through this framework, DOE will continue to be an integral component of agency-wide efforts to prepare for and respond to biorisks and other crises. This technical report describes the goals, progress, and results of NVBL’s five project teams—Molecular Design for COVID-19 Therapeutics, COVID-19 Testing, Epidemiological Modeling, Viral Fate and Transport, and Materials and Manufacturing of Critical Supplies—and lists each team’s publications and research output.

42 ENGINEERING↗

Performance Analysis of the NAS Y-MP Workload

This paper describes the performance characteristics of the computational workloads on the NAS Cray Y-MP machines, a Y-MP 832 and later a Y-MP 8128. Hardware measurements indicated that the Y-MP workload performance matured over time, ultimately sustaining an average throughput of 0.8 GFLOPS and a vector operation fraction of 87%. The measurements also revealed an operation rate exceeding 1 per clock period, a well-balanced architecture featuring a strong utilization of vector functional units, and an efficient memory organization. Introduction of the larger memory 8128 increased throughput by allowing a more efficient utilization of CPUs. Throughput also depended on the metering of the batch queues; low-idle Saturday workloads required a buffer of small jobs to prevent memory starvation of the CPU. UNICOS required about 7% of total CPU time to service the 832 workloads; this overhead decreased to 5% for the 8128 workloads. While most of the system time went to service I/O requests, efficient scheduling prevented excessive idle due to I/O wait. System measurements disclosed no obvious bottlenecks in the response of the machine and UNICOS to the workloads. In most cases, Cray-provided software tools were- quite sufficient for measuring the performance of both the machine and operating, system.

Bergeron, Robert J.↗

TU-EF-204-02: Hiigh Quality and Sub-MSv Cerebral CT Perfusion Imaging

Purpose: CT Perfusion (CTP) imaging is of great importance in acute ischemic stroke management due to its potential to detect hypoperfused yet salvageable tissue and distinguish it from definitely unsalvageable tissue. However, current CTP imaging suffers from poor image quality and high radiation dose (up to 5 mSv). The purpose of this work was to demonstrate that technical innovations such as Prior Image Constrained Compressed Sensing (PICCS) have the potential to address these challenges and achieve high quality and sub-mSv CTP imaging. Methods: (1) A spatial-temporal 4D cascaded system model was developed to indentify the bottlenecks in the current CTP technology; (2) A task-based framework was developed to optimize the CTP system parameters; (3) Guided by (1) and (2), PICCS was customized for the reconstruction of CTP source images. Digital anthropomorphic perfusion phantoms, animal studies, and preliminary human subject studies were used to validate and evaluate the potentials of using these innovations to advance the CTP technology. Results: The 4D cascaded model was validated in both phantom and canine stroke models. Based upon this cascaded model, it has been discovered that, as long as the spatial resolution and noise properties of the 4D source CT images are given, the 3D MTF and NPS of the final CTP maps can be analytically derived for a given set of processing methods and parameters. The cascaded model analysis also identified that the most critical technical factor in CTP is how to acquire and reconstruct high quality source images; it has very little to do with the denoising techniques often used after parametric perfusion calculations. This explained why PICCS resulted in a five-fold dose reduction or substantial improvement in image quality. Conclusion: Technical innovations generated promising results towards achieving high quality and sub-mSv CTP imaging for reliable and safe assessment of acute ischemic strokes. K. Li, K. Niu, Y. Wu: Nothing to disclose. G.-H. Chen: Research funded, GE Healthcare; Research funded, Siemens AX.

60 APPLIED LIFE SCIENCES↗

Anodized Aluminum Oxide Membrane Ionic Memristors

Memory effect in ion transport (IT) at the solid–solution interface is uniquely attractive in that the conductance depends on or “memorizes” the previous states. Hysteretic and rectified transport properties offer exciting potential to developing advanced iontronics and neuromorphic functions, improving the efficiency of energy conversion and electrochemical processes, and overcoming the selectivity-throughput bottleneck in the enrichment of low abundant species for environment- and energy-friendly separations, among others. Herein, memory effects are discovered in the rectified electrokinetic IT through anodized aluminum oxide (AAO) membranes containing densely packed highly ordered nanochannels (10 10 per cm 2 ). Characteristic memristor responses of pinched current–potential loops are resolved in voltammetric experiments and successfully reproduced through finite element simulation. Excitatory and inhibitory conductance states are shown to arise from the enrichment and depletion of mobile charge carriers. Structurewise, the transport symmetry is broken by the barrier oxide layer (BOL) on the one end of the cylindrical nanochannels across the AAO membranes. Charge selectivity is attributed to the gradient(s) of the space charge density across the BOL characterized by depth profiling via X-ray photoelectron spectroscopy analysis. The space charge gradient(s) overcomes the fundamental limitation of widely exploited surface charge effects to enable intense rectification and hysteresis prevailing at very high ionic concentrations up to 1–2 M. A new strategy is developed for controlling the preferential IT direction and selectivity via counterion intercalation and extraction/exchange. Mechanistic understanding is further confirmed through parameter variations such as potential scan rate and ionic strength, which also demonstrates convenient controls of the related functions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Machine learning inversion of interatomic force constants from single-crystal inelastic neutron scattering

Atomic vibrations govern many macroscopic properties of materials, but experiments to comprehensively probe them remain challenging. Inelastic neutron scattering (INS) is a powerful technique to map phonon dispersions in crystals, especially when leveraging modern time-of-flight (ToF) spectrometers with large detectors. However, efficiently and robustly extracting interatomic force constants (FCs) parameterizing phonon dynamics from experimental spectra remains a bottleneck due to the complexity and high dimensionality of ToF INS datasets. Here, we present a machine learning approach for the direct inversion of FCs from single-crystal INS measurements. The framework leverages synthetic training data generated using universal machine-learned force fields and an efficient physics-based forward model. We benchmark two neural architectures–one emphasizing structured latent representation learning and the other direct, supervised spectral regression–across simulated datasets for two materials under idealized and noisy conditions. The latent-representation model is subsequently applied to experimental single-crystal INS data on germanium. The model is shown to reproduce FCs derived from both first-principles simulations and from iterative optimization, and furthermore achieves reliable inference even from sparse, single-orientation measurements representing short data acquisitions. Analysis of the learned latent space reveals semantically continuous and physically interpretable encodings that support strong cross-domain generalization. By bridging theoretical and experimental domains, we establish a path toward rapid inversion of experimental spectra and data-driven interpretation of temperature-dependent lattice dynamics.

42 ENGINEERING↗

Scenario Complexity for Unmanned Aircraft System Traffic

This work introduces an approach to estimate the complexity of a low-altitude air traffic scenario involving multiple UASs using mathematical programming. Given a set of multi-point UAS flight trajectories, vehicle dynamics, and a conflict resolution algorithm, an abstract model is developed such that it can be solved quickly using a mathematical programming optimization software without running high-fidelity simulations that can be computationally expensive and may not suit real-time apA quick and accurate assessment of complexity for a given traffic scenario can help plan and schedule flights to alleviate traffic bottleneck and mitigate operation risks, especially for unmanned aerial system traffic management where high traffic density or complexity is expected. This work introduces a traffic scenario complexity metric that was constructed based on the number of potential conflicts weighted by the conflict resolution cost associated. The cost associated with a conflict is calculated based on the corresponding conflict resolution maneuvers. To obtain the conflict resolution maneuvers, a MILP-based optimization was formulated with the vehicle model and conflict management parameters incorporated. To evaluate the complexity metrics, an approach of using measurements from high-fidelity simulations was proposed. The scenario complexity measurements for 920 random-generated scenarios were obtained through high-fidelity simulations and treated as the ground truth. Two statistics methods: Pearson and Alternative Conditional Expectations were applied for analysis. The results showed that the number of flights has low correlation with the scenario complexity according to the correlation coefficients calculated by both methods. The Alternative Conditional Expectations method shows that the proposed scenario complexity metric has better correlation with the ground truth than the number of potential conflicts.plications. In the abstract model, each vehicle is represented by a time-varied vector associated with position, speed, and heading information. The total extra distance that aircraft need to divert from their original routes to avoid collisions is computed and used to setup a quadratic programming formula. The metrics including the number of conflicts and extra distances travelled by all vehicles are then utilized to estimate the complexity of a given UAS flight scenario. Results and verification against high-fidelity simulations will be provided in the final draft.

traffic complexity↗

SciDAC Visualization and Analytics Center for Enabling Technologies

This project focuses on leveraging scientific visualization and analytics software technology as an enabling technology for increasing scientific productivity and insight. Advances in computational technology have resulted in an "information big bang," which in turn has created a significant data understanding challenge. This challenge is widely acknowledged to be one of the primary bottlenecks in contemporary science. The vision for our Center is to respond directly to that challenge by adapting, extending, creating when necessary and deploying visualization and data understanding technologies for our science stakeholders. Using an organizational model as a Visualization and Analytics Center for Enabling Technologies (VACET), we are well positioned to be responsive to the needs of a diverse set of scientific stakeholders in a coordinated fashion using a range of visualization, mathematics, statistics, computer and computational science and data management technologies.

97 MATHEMATICS AND COMPUTING↗

High Fidelity Simulations of Large-Scale Wireless Networks (Part I)

Wireless systems and networks have experienced rapid growth over the last decade with the advent of smart devices for everyday use. These systems, which include smartphones, vehicular gadgets, and internet-of-things devices, are becoming ubiquitous and ever-more important. They pose interesting research challenges for design and analysis of new network protocols due to their large scale and complexity. In this work, we focus on the challenging aspect of simulating the inter-connectivity of many of these devices in wireless networks. The quantitative study of large scale wireless networks, with counts of wireless devices in the thousands, is a very difficult problem with no known acceptable solution. By necessity, simulations of this scale have to approximate reality, but the algorithms employed in most modern-day network simulators can be improved for wireless network simulations. In this report, we present advances that we have made and propositions for continuation of progress towards a framework for high fidelity simulations of wireless networks. This work is not complete in that a final simulation framework tool is yet to be produced. However, we highlight the major bottlenecks and address them individually with initial results showing enough promise.

42 ENGINEERING↗

Inflatable Antenna for CubeSat: A New Spherical Design for Increased X-Band Gain

Interplanetary CubeSats and small satellites have potential to provide means to explore space and to perform science in a more affordable way. As the goals for these spacecraft become more ambitious in space exploration, the communication systems currently implemented will need to be improved to support those missions. One of the bottlenecks is the antennas’ size, due to the close relation between antenna gain and dimensions. Hence, a possible solution is to develop inflatable antennas which can be packaged efficiently, occupying a small amount of space, and they can provide, once deployed, large dish dimension and correspondent gain. A prototype of a 1 m inflatable antenna for X-Band has been developed in a joint effort between JPL and ASU. After initial photogrammetry tests and radiation tests, it was discovered that the design was not able to meet the required gain. As a result, a new design, based on a spherical inflatable membrane, is proposed. This new design will allow reaching a more stable inflatable surface, hence improving the electromagnetic performance. This paper will detail the principle challenges in developing this new antenna focusing on: design, EM analysis, fabrication and tests.

Babuscia, Alessandra↗

Massive all-atom analysis of 2D materials with quantum properties (Final report)

Improvements in microscopy have enabled the acquisition of data at a scale that is difficult to process manually, making automated machine learning approaches to analyzing experimental images essential. In this project, we developed and applied machine learning (ML) workflows for atomic resolution scanning transmission electron microscopy (STEM) images. This development included improving both methodology as well as generating user-friendly codes. We developed machine learning architectures which, after training, automatically identify the location and types of defects throughout a material. We used these data to produce class-averaged images of 2D atomic coordinates with up to 0.3 pm precision, uncovering the structure and oscillations of long-range strain fields around point defects in WSe 2-2x Te 2x . We also resolved a long-standing problem in this field in the training of ML models, a lack of labeled experimental data, by developing a cycle-GAN that transformed simulated-generated labeled data into labeled data indistinguishable from experiment and therefore suitable for training. This removed the remaining parts of the ML data processing workflow where human intervention was still critical and therefore a bottleneck to working at scale. Codes have been developed and released for this full machine learning workflow. ML approaches to partially automate STEM acquisition were also developed. Finally we applied ML and other advanced data processing methods to several materials science problems in two-dimensional materials, including studying the evolution of hyperuniformity with defect concentration in WSe2, understanding phase transformations in transition metal dichalcogenides during in-situ heating in the STEM, and exploring how 2D interfaces transform from twisted into aligned structures.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Stress Intensity of Delamination in a Sintered-Silver Interconnection: Preprint

In automotive power electronics packages, conventional thermal interface materials such as greases, gels, and phase-change materials pose bottlenecks to heat removal and are also associated with reliability concerns. The industry trend is toward high thermal performance bonded interfaces for large-area attachments. However, because of coefficient of thermal expansion mismatches between materials/layers and resultant thermomechanical stresses, adhesive and cohesive fractures could occur, posing a reliability problem. These defects manifest themselves in increased thermal resistance. This research aims to investigate and improve the thermal performance and reliability of sintered-silver for power electronics packaging applications. This has been experimentally accomplished by the synthesis of large-area bonded interfaces between metalized substrates and copper base plates that have subsequently been subjected to thermal cycles. A finite element model of crack initiation and propagation in these bonded interfaces will allow for the interpretation of degradation rates by a crack-velocity (V)-stress intensity factor (K) analysis. A description of the experiment and the modeling approach are discussed.

30 DIRECT ENERGY CONVERSION↗

User's guide to the Reliability Estimation System Testbed (REST)

The Reliability Estimation System Testbed is an X-window based reliability modeling tool that was created to explore the use of the Reliability Modeling Language (RML). RML was defined to support several reliability analysis techniques including modularization, graphical representation, Failure Mode Effects Simulation (FMES), and parallel processing. These techniques are most useful in modeling large systems. Using modularization, an analyst can create reliability models for individual system components. The modules can be tested separately and then combined to compute the total system reliability. Because a one-to-one relationship can be established between system components and the reliability modules, a graphical user interface may be used to describe the system model. RML was designed to permit message passing between modules. This feature enables reliability modeling based on a run time simulation of the system wide effects of a component's failure modes. The use of failure modes effects simulation enhances the analyst's ability to correctly express system behavior when using the modularization approach to reliability modeling. To alleviate the computation bottleneck often found in large reliability models, REST was designed to take advantage of parallel processing on hypercube processors.

Nicol, David M.↗

MyTaxa: an advanced taxonomic classifier for genomic and metagenomic sequences

Determining the taxonomic affiliation of sequences assembled from metagenomes remains a major bottleneck that affects research across the fields of environmental, clinical and evolutionary microbiology. Here, we introduce MyTaxa, a homology-based bioinformatics framework to classify metagenomic and genomic sequences with unprecedented accuracy. The distinguishing aspect of MyTaxa is that it employs all genes present in an unknown sequence as classifiers, weighting each gene based on its (predetermined) classifying power at a given taxonomic level and frequency of horizontal gene transfer. MyTaxa also implements a novel classification scheme based on the genome-aggregate average amino acid identity concept to determine the degree of novelty of sequences representing uncharacterized taxa, i.e. whether they represent novel species, genera or phyla. Application of MyTaxa on in silico generated (mock) and real metagenomes of varied read length (100–2000 bp) revealed that it correctly classified at least 5% more sequences than any other tool. The analysis also showed that ~10% of the assembled sequences from human gut metagenomes represent novel species with no sequenced representatives, several of which were highly abundant in situ such as members of the Prevotella genus. Thus, MyTaxa can find several important applications in microbial identification and diversity studies.

59 BASIC BIOLOGICAL SCIENCES↗

Exponential acceleration of macroscopic quantum tunneling in a Floquet Ising model

The exponential suppression of macroscopic quantum tunneling (MQT) in the number of elements to be reconfigured is an essential element of broken symmetry phases. This suppression is also a core bottleneck in quantum algorithms, such as traversing an energy landscape in optimization, and adiabatic state preparation more generally. In this work, we demonstrate exponential acceleration of MQT through Floquet engineering with the application of a uniform, high frequency transverse drive field. Using the ferromagnetic phase of the transverse field Ising model in one and two dimensions as a prototypical example, we identify three phenomenological regimes as a function of drive strength. For weak drives, the system exhibits exponentially decaying tunneling rates but robust magnetic order; in the crossover regime at intermediate drive strength, we find polynomial decay of tunnelling alongside vanishing magnetic order; and at very strong drive strengths both the Rabi frequency and time-averaged magnetic order are approximately constant with increasing system size. We support these claims with extensive full wavefunction and tensor network numerical simulations, and theoretical analysis. An experimental test of these results presents a technologically important and novel scientific question accessible on NISQ-era quantum computers.

Grattan, George↗

XPRESS: eXascale PRogramming Environment and System Software, Final Report

The University of Oregon (UO) was a partner in the XPRESS project together with Sandia National Laboratories (SNL), Indiana University (IU), Lawrence Berkeley National Laboratory (LBL), Louisiana State University (LSU), Oak Ridge National Laboratory (ORNL), University of Houston (UH), and University of North Carolina (RENCI). UO's objective in the project was to develop "APEX (Asynchronous Performance Environment for eXascale)": performance measurement, analysis, feedback and control infrastructure to asynchronous tasking runtimes, specifically HPX (LSU) and HPX-5 (IU). During the project period, our efforts focused on the need to measure complex tasking models where the traditional instrumentation methods would not work due to an absent/meaningless call-path context and runtime control of task pre-emption and resumption, usually on a different operating system thread. In addition to the challenges of performance measurement, we developed a framework to provide information from all levels of the stack (hardware, operating system, runtime, application) to a feedback & control mechanism for controlling concurrency, task granularity, networking parameters, and other aspects of execution in order to improve execution performance, throughput and/or efficiency. UO demonstrated these developments on leading edge HPC machines at large scale, including a 2017 Gordon Bell submission in which APEX was used to optimize key bottlenecks in an astrophysics application. The following final report summarizes our accomplishments during the XPRESS project.

99 GENERAL AND MISCELLANEOUS↗

Sustainable production of ammonia fertilizers from biomass

Nitrogen is the most limiting plant nutrient. Inexpensive natural gas has substantially reduced costs of ammonia fertilizer for intensive agriculture in the developed world, but its excessive use negatively impacts downstream ecosystems. By contrast, the availability of ammonia fertilizer is a major economic bottleneck for agriculture in developing countries. A dedicated lignocellulosic biomass crop could supply sufficient substrate to generate optimal nitrogen fertilizer on less than 5% of a grower’s food crop acreage. However, reimagining ammonia generation using biomass could significantly enhance sustainable agricultural productivity in developing countries when combined with emerging catalytic technologies.

09 BIOMASS FUELS↗

Wave propagation in spatially modulated tubes

We investigate wave propagation in rotationally symmetric tubes with a periodic spatial modulation of cross section. Using an asymptotic perturbation analysis, the governing quasi-two-dimensional reaction-diffusion equation can be reduced into a one-dimensional reaction-diffusion-advection equation. Assuming a weak perturbation by the advection term and using projection method, in a second step, an equation of motion for traveling waves within such tubes can be derived. Both methods predict properly the nonlinear dependence of the propagation velocity on the ratio of the modulation period of the geometry to the intrinsic width of the front, or pulse. As a main feature, we observe finite intervals of propagation failure of waves induced by the tube’s modulation and derive an analytically tractable condition for their occurrence. For the highly diffusive limit, using the Fick-Jacobs approach, we show that wave velocities within modulated tubes are governed by an effective diffusion coefficient. Furthermore, we discuss the effects of a single bottleneck on the period of pulse trains. We observe period changes by integer fractions dependent on the bottleneck width and the period of the entering pulse train.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗