Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Computational Load Balancing”

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 541 records · Page 30

Mapping a battlefield simulation onto message-passing parallel architectures

Perhaps the most critical problem in distributed simulation is that of mapping: without an effective mapping of workload to processors the speedup potential of parallel processing cannot be realized. Mapping a simulation onto a message-passing architecture is especially difficult when the computational workload dynamically changes as a function of time and space; this is exactly the situation faced by battlefield simulations. This paper studies an approach where the simulated battlefield domain is first partitioned into many regions of equal size; typically there are more regions than processors. The regions are then assigned to processors; a processor is responsible for performing all simulation activity associated with the regions. The assignment algorithm is quite simple and attempts to balance load by exploiting locality of workload intensity. The performance of this technique is studied on a simple battlefield simulation implemented on the Flex/32 multiprocessor. Measurements show that the proposed method achieves reasonable processor efficiencies. Furthermore, the method shows promise for use in dynamic remapping of the simulation.

Nicol, David M.↗

A semi–automatic analytical methodology for characterizing the energy consumption of MRI systems using load duration curves

Background and purpose: Magnetic resonance imaging (MRI) scanners are a major contributor to greenhouse gas emissions from the healthcare sector, and efforts to improve energy efficiency and reduce energy consumption rely on quantification of the characteristics of energy consumption. The purpose of this work was to develop a semi-automatic analytical methodology for the characterization of the energy consumption of MRI systems using only the load duration curve (LDC). LDCs are a fundamental tool used across various fields to analyze and understand the behavior of loads over time. Methods: An electric current transformer sensor and data logger were installed on two 3T MRI scanners from two vendors, termed M1 (outpatient scanner) and M2 (inpatient/emergency scanner). Data was collected for 1 month (7/11/2023 to 8/11/2023). Active power was calculated, assuming a balanced three-phase system, using the average current measured across all three phases, a 480 V reference voltage for both machines, and vendor-provided power factors. An LDC was constructed for each system by sorting the active power values in descending order and computing the cumulative time (in units of percentage) for each data point. The first derivative of the LDC was then computed (LDC’), smoothed by convolution with a window function (sLDC’), and used to detect transitions between different system modes including (in descending power levels): scan, prepared-to-scan, idle, low-power, and off. The final, segmented LDC was used to measure time (% total time), total energy (kWh), and mean power (kW) for each system mode on both scanners. The method was validated by comparing mean power values, computed using the segmented 1-month LDC, for each nonproductive system mode (i.e., prepared-to-scan, idle, lower-power, and off) against power levels measured after a deliberate system shutdown was performed for each scanner (1 day worth of data). Results: The validation revealed differences in mean power values <1.4% for all nonproductive modes and both scanners. In the scan system mode, the mean power values ranged from 29.8 to 37.2 kW and the total energy consumed for 1 month ranged from 11 106 to 14 466 kWh depending on the scanner. Over the course of 1 month, the portion of time the scanners were in nonproductive modes ranged from 76% to 80% across scanners and the nonproductive energy consumption ranged from 8010 to 6722 kWh depending on the scanner. The M1 (outpatient) scanner consumed 99.9 and 183.9 kWh/day in idle mode for weekdays and weekends, respectively, because the scanner spent 23% more time proportionally in idle mode on the weekends. Conclusions: A semi-automatic method for quantifying energy consumption characteristics of MRI scanners was introduced and validated. This method is relatively simple to implement as it requires only power data from the scanners and avoids the technical challenges associated with extracting and processing scanner log files. Finally, the methodology enables quantitative evaluation of the power, time, and energy characteristics of MRI scanners in scan and nonproductive system modes, providing baseline data and the capability of identifying potential opportunities for enhancing the energy efficiency of MRI scanners.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

RingX: Scalable Parallel Attention for Long-Context Learning on HPC

The attention mechanism has become foundational for remarkable AI breakthroughs since the introduction of the Transformer, driving the demand for increasingly longer context to power frontier models such as large-scale reasoning language models and high-resolution image/video generators. However, its quadratic computational and memory complexities present substantial challenges. Current state-of-the-art parallel attention methods, such as ring attention, are widely adopted for long-context training but utilize a point-to-point communication strategy that fails to fully exploit the capabilities of modern HPC network architectures. In this work, we propose ringX, a scalable family of parallel attention methods optimized explicitly for HPC systems. By enhancing workload partitioning, refining communication patterns, and improving load balancing, ringX achieves up to 3.4 × speedup compared to conventional ring attention on the Frontier supercomputer. Optimized for both bi-directional and causal attention mechanisms, ringX demonstrates its effectiveness through training benchmarks of a Vision Transformer (ViT) on a climate dataset and a Generative Pre-Trained Transformer (GPT) model, Llama3 8B. Our method attains an end-to-end training speedup of approximately 1.5 × in both scenarios. To our knowledge, the achieved 38% model FLOPs utilization (MFU) for training Llama3 8B with a 1M-token sequence length on 4,096 GPUs represents one of the highest training efficiencies reported for long-context learning on HPC systems. Our code implementation is available at https://github.com/jqyin/ringX-attention.

Yin, Junqi [ORNL] (ORCID:0000000338435520)↗

Parallel Programming Strategies for Irregular Adaptive Applications

Achieving scalable performance for dynamic irregular applications is eminently challenging. Traditional message-passing approaches have been making steady progress towards this goal; however, they suffer from complex implementation requirements. The use of a global address space greatly simplifies the programming task, but can degrade the performance for such computations. In this work, we examine two typical irregular adaptive applications, Dynamic Remeshing and N-Body, under competing programming methodologies and across various parallel architectures. The Dynamic Remeshing application simulates flow over an airfoil, and refines localized regions of the underlying unstructured mesh. The N-Body experiment models two neighboring Plummer galaxies that are about to undergo a merger. Both problems demonstrate dramatic changes in processor workloads and interprocessor communication with time; thus, dynamic load balancing is a required component.

Biswas, Rupak↗

Measurement of Unsteady Pressure Data on a Large HSCT Semispan Wing and Comparison with Analysis

Experimental data from wind-tunnel tests of the Rigid Semispan Model (RSM) performed at NASA Langley's Transonic Dynamics Tunnel (TDT) are presented. The primary focus of the paper is on data obtained from testing of the RSM on the Oscillating Turntable (OTT). The OTT is capable of oscillating models in pitch at various amplitudes and frequencies about mean angles of attack. Steady and unsteady pressure data obtained during testing of the RSM on the OTT is presented and compared to data obtained from previous tests of the RSM on a load balance and on a Pitch and Plunge Apparatus (PAPA). Testing of the RSM on the PAPA resulted in utter boundaries that were strongly dependent on angle of attack across the Mach number range. Pressure data from all three tests indicates the existence of vortical flows at moderate angles of attack. The correlation between the vortical flows and the unusual utter boundaries from the RSM/PAPA test is discussed. Comparisons of experimental data with analyses using the CFL3Dv6 computational fluid dynamics code are presented.

Scott, Robert C.↗

Technical Challenges and Lessons from the Migration of the GLOBE Data and Information System to Utilize Cloud Computing Service

The Global Learning and Observation to Benefit the Environment (GLOBE) Data and Information System supports an international science and education program with capabilities to accept local environment observations, archive, display and visualize them along with global satellite observations. Since its inception twenty years ago, the Web and database system has been upgraded periodically to accommodate the changes in technology and the steady growth of GLOBEs education community and collection of observations. Recently, near the end-of-life of the system hardware, new commercial computer platform options were explored and a decision made to utilize Cloud services. Now the GLOBE DIS has been fully deployed and maintained using Amazon Cloud services for over two years now. This paper reviews the early risks, actual challenges, and some unexpected findings as a result of the GLOBE DIS migration. We describe the plans, cost drivers and estimates, highlight adjustments that were made and suggest improvements. We present the trade studies for provisioning, for load balancing, networks, processing, storage, as well as production, staging and backup systems. We outline the migration teams skills and required level of effort for transition, and resulting changes in the overall maintenance and operations activities. Examples include incremental adjustments to processing capacity and frequency of backups, and efforts previously expended on hardware maintenance that were refocused onto application-specific enhancements.

GLOBE↗

Technical Challenges and Lessons from the Migration of the GLOBE Data and Information System to Utilize Cloud Computing Service

The Global Learning and Observation to Benefit the Environment (GLOBE) Data and Information System supports an international science and education program with capabilities to accept local environment observations, archive, display and visualize them along with global satellite observations. Since its inception twenty years ago, the Web and database system has been upgraded periodically to accommodate the changes in technology and the steady growth of GLOBEs education community and collection of observations. Recently, near the end-of-life of the system hardware, new commercial computer platform options were explored and a decision made to utilize Cloud services. Now the GLOBE DIS has been fully deployed and maintained using Amazon Cloud services for over two years now. This paper reviews the early risks, actual challenges, and some unexpected findings as a result of the GLOBE DIS migration. We describe the plans, cost drivers and estimates, highlight adjustments that were made and suggest improvements. We present the trade studies for provisioning, for load balancing, networks, processing, storage, as well as production, staging and backup systems. We outline the migration teams skills and required level of effort for transition, and resulting changes in the overall maintenance and operations activities. Examples include incremental adjustments to processing capacity and frequency of backups, and efforts previously expended on hardware maintenance that were refocused onto application-specific enhancements.

GLOBE Data↗

Effect of Domain Size, Boundary, and Loading Conditions on Mechanical Properties of Amorphous Silica: A Reactive Molecular Dynamics Study

Mechanical properties are very important when choosing a material for a specific application. They help to determine the range of usefulness of a material, establish the service life, and classify and identify materials. The size effect on mechanical properties has been well established numerically and experimentally. However, the role of the size effect combined with boundary and loading conditions on mechanical properties remains unknown. In this paper, by using molecular dynamics (MD) simulations with the state-of-the-art ReaxFF force field, we study mechanical properties of amorphous silica (e.g., Young’s modulus, Poisson’s ratio) as a function of domain size, full-/semi-periodic boundary condition, and tensile/compressive loading. We found that the domain-size effect on Young’s modulus and Poisson’s ratio is much more significant in semi-periodic domains compared to full-periodic domains. The results, for the first time, revealed the bimodular and anisotropic nature of amorphous silica at the atomic level. We also defined a “safe zone” regarding the domain size, where the bulk properties of amorphous silica can be reproducible, while the computational cost and accuracy are in balance.

bimodular materials↗

Progressive Damage Response of 3D Woven Composites via the Multiscale Recursive Micromechanics Solution with Tailored Fidelity

Progressive failure simulations have been performed for orthogonal 3D woven composites consisting of RTM6 resin matrix and AS4 carbon fibers. The Multiscale Recursive Micromechanics approach has been used, which, while being computationally efficient, captures the primary effects of the microstructure at each considered length scale. This approach also enables use of any micromechanics theory at any length scale, and herein, the fidelity of the chosen theories across the scales has been tailored to strike a balance with computational efficiency. The Mori-Tanaka method is employed at the lowest length scale, the Generalized Method of Cells is used at intermediate scales, and the High-Fidelity Generalized Method of Cells is used at the highest woven composite repeating unit cell scale. Furthermore, two different damage models, also with different levels of fidelity and efficiency, have been used for the resin material at the lowest length scale. Results for the mechanical behavior in response to loading in various directions are compared for the two damage models and with available test data.

3D woven↗

Model-Free Building Temperature Control and Power Allocation Under Measurement Time Delays

Taking a step towards a greener planet has created an increased need for a higher integration of renewable energy resources into the electric grid. Nonetheless, the intermittency and uncertainty associated with renewable generation have slowed down this integration. Demand response (DR) has been recently adopted to address this challenge by utilizing demand side flexibility and enabling the participation of many grid-interactive efficient buildings (GEBs). However, existing DR methods require significant modeling and/or training efforts and are computationally expensive. To address the aforementioned issues, we propose a model-free control (MFC)-based strategy that is robust to the time delays in the temperature measurements of the thermostatically controlled loads (TCLs). It assigns to each GEB a local controller to maintain the TCLs’ temperatures within desired comfort levels, while the load aggregator (LA) allocates the assigned reference power provided by the distribution system operator (DSO) to support a specific grid service, such as demand peak reduction, load shifting, balancing supply and demand, and consuming the solar photovoltaic power locally. We investigate the effects of such loss of information on the local control action as well as on meeting the power allocation constraint. We conclude that, for an appropriate choice of design parameters, the proposed MFC controller is satisfactorily robust to measurement time delays.

Telsang, Bhagyashri↗

DARMA-Kokkos Data and Execution Interoperability WBS 2.3.1.04 Milestone 16 (ECP Milestone Report)

DARMA (Distributed Asynchronous Resilient Models for Applications) is a runtime library supporting the Sandia ATDM (Advanced Technology Development and Mitigation) program. The main application drivers fall within the ECP milestone 2.2.5.03 ADNN03-ASC ATDM SNL Application, which includes applications that require load balancing and asynchronous communication for high performance. The DARMA runtime infrastructure has been modified to be compatible with Kokkos/OpenMP parallelization within tasks, which is a critical requirement for high performance for the Sandia ATDM apps. DARMA development has occurred in parallel with a verification milestone for ATDM in FY18. For FY19, DARMA should impact ATDM by enabling dynamic load balancing and communication through only incremental changes to the existing verified MPI codes. DARMA can now support the intra-kernel thread parallelization in the parent MPI apps, allowing DARMA to be easily added without rewriting individual math kernels. The results presented here demonstrate the DARMA results for an MPI mini-app.

97 MATHEMATICS AND COMPUTING↗

Key Residential Building Equipment Technologies for Control and Grid Support PART I (Residential)

Electrical energy consumption of the residential sector is a crucial area of research that has in the past primarily focused on increasing the efficiency of household devices such as water heaters, dishwashers, air conditioners, and clothes washer and dryer units. However, the focus of this research is shifting as objectives such as developing the smart grid and ensuring that the power system remains reliable come to the fore, along with the increasing need to reduce energy use and costs. Load research has started to focus on mechanisms to support the power system through demand reduction and/or reliability services. The power system relies on matching generation and load, and day-ahead and real-time energy markets capture most of this need. However, a separate set of grid services exist to address the discrepancies in load and generation arising from contingencies and operational mismatches, and to ensure that the transmission system is available for delivery of power from generation to load. Currently, these grid services are mostly provided by generation resources. The addition of renewable resources with their inherent variability can complicate the issue of power system reliability and lead to the increased need for grid services. Using load as a resource, through demand response programs, can fill the additional need for flexible resources and even reduce costly energy peaks. Loads have been shown to have response that is equal to or better than generation in some cases. Furthermore, price-incentivized demand response programs have been shown to reduce the peak energy requirements, thereby affecting the wholesale market efficiency and overall energy prices. The residential sector is not only the largest consumer of electrical energy in the United States, but also has the highest potential to provide demand reduction and power system support, as technological advancements in load control, sensor technologies, and communication are made. The prevailing loads based on the largest electrical energy consumers in the residential sector are space heating and cooling, washer and dryer, water heating, lighting, computers and electronics, dishwasher and range, and refrigeration. As the largest loads, these loads provide the highest potential for delivering demand response and reliability services. Many residential loads have inherent flexibility that is related to the purpose of the load. Depending on the load type, electric power consumption levels can either be ramped, changed in a step-change fashion, or completely removed. Loads with only on-off capability (such as clothes washers and dryers) provide less flexibility than resources that can be ramped or step-changed. Add-on devices may be able to provide extra demand response capabilities. Still, operating residential loads effectively requires awareness of the delicate balance of occupants health and comfort and electrical energy consumption. This report is Phase I of a series of reports aimed at identifying gaps in automated home energy management systems for incorporation of building appliances, vehicles, and renewable adoption into a smart grid, specifically with the intent of examining demand response and load factor control for power system support. The objective is to capture existing gaps in load control, energy management systems, and sensor technology with consideration of PHEV and renewable technologies to establish areas of research for the Department of Energy. In this report, (1) data is collected and examined from state of the art homes to characterize the primary residential loads as well as PHEVs and photovoltaic for potential adoption into energy management control strategies; and (2) demand response rules and requirements across the various demand response programs are examined for potential participation of residential loads. This report will be followed by a Phase II report aimed at identifying the current state of technology of energy management systems, sensors, and communication technologies for demand response and load factor control applications for the residential sector. The purpose is to cover the gaps that exist in the information captured by the sensors for energy management system to be able to provide demand response and load factor control. The vision is the development of an energy management system or other controlling enterprise hardware and software that is not only able to control loads, PHEVs, and renewable generation for demand response and load factor control, but also to do so with consumer comforts in mind and in an optimal fashion.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Nonlinear coupled rotor-fuselage helicopter vibration studies with higher harmonic control

This paper addresses the problem of vibration prediction and vibration reduction in helicopters by means of active control methodologies. The nonlinear equations of a coupled rotor/flexible-fuselage system have been derived using computer algebra, thus relegating this tedious task to the computer. In the solution procedure the trim state and vibratory response of the helicopter are obtained in a single pass by using a harmonic balance technique and simultaneously satisfying the trim and the vibratory response of the helicopter in all the rotor and fuselage degrees of freedom. Using this solution procedure, the influence of the fuselage flexibility on the vibratory response is studied. In addition, it is shown that the conventional single frequency HHC is capable of reducing either the hub loads or only the fuselage vibrations but not both simultaneously. A new scheme called MHHC, having multiple higher harmonic pitch inputs, was used to accomplish this task of simultaneously reducing both the vibratory hub loads and fuselage vibratory response. In addition, the uniqueness of this MHHC scheme is explained in detail.

Friedmann, P. P.↗

Rocket ascent G-limited moment-balanced optimization program (RAGMOP)

This document describes the RAGMOP (Rocket Ascent G-limited Momentbalanced Optimization Program) computer program for parametric ascent trajectory optimization. RAGMOP computes optimum polynomial-form attitude control histories, launch azimuth, engine burn-time, and gross liftoff weight for space shuttle type vehicles using a search-accelerated, gradient projection parameter optimization technique. The trajectory model available in RAGMOP includes a rotating oblate earth model, the option of input wind tables, discrete and/or continuous throttling for the purposes of limiting the thrust acceleration and/or the maximum dynamic pressure, limitation of the structural load indicators (the product of dynamic pressure with angle-of-attack and sideslip angle), and a wide selection of intermediate and terminal equality constraints.

Lyons, J. T.↗

A simplified CFD model for spectral radiative heat transfer in high-pressure hydrocarbon–air combustion systems

Detailed radiation modeling in piston engines has received relatively little attention to date. Recently, it is being revisited in light of current trends towards higher operating pressures and higher levels of exhaust-gas recirculation (EGR), both of which enhance molecular gas radiation. Advanced high-efficiency engines also are expected to function closer to the limits of stable operation, where even small perturbations to the energy balance can have a large influence on system behavior. Detailed radiation modeling using sophisticated tools like photon Monte Carlo/line-by-line (PMC/LBL) is computationally expensive. Here, guided by results from PMC/LBL, a simplified stepwise-gray spectral model in combination with a first-order spherical harmonics (P1 method) radiative transfer equation (RTE) solver is proposed and tested for engine-relevant conditions. Radiative emission, reabsorption and radiation reaching the walls are computed for a heavy-duty compression-ignition engine at part-load and full-load operating conditions with different levels of EGR and soot. The results are compared with those from PMC/LBL, P1/FSK (P1 with a full-spectrum k-distribution spectral model) and P1/Gray radiation models to assess the proposed model’s accuracy and computational cost. The results show that the proposed P1/StepwiseGray model can calculate reabsorption locally and globally with less than 10% error (with respect to PMC/LBL) at a small fraction of the computational cost of PMC/LBL (a factor of 30) and P1/FSK (a factor of 15). In contrast, error in computed reabsorption by the P1/Gray model is as high as 60%. Finally, it is expected that the simplified model should be broadly applicable to high-pressure hydrocarbon–air combustion systems, with or without soot.

42 ENGINEERING↗

An Analysis of Performance Enhancement Techniques for Overset Grid Applications

The overset grid methodology has significantly reduced time-to-solution of high-fidelity computational fluid dynamics (CFD) simulations about complex aerospace configurations. The solution process resolves the geometrical complexity of the problem domain by using separately generated but overlapping structured discretization grids that periodically exchange information through interpolation. However, high performance computations of such large-scale realistic applications must be handled efficiently on state-of-the-art parallel supercomputers. This paper analyzes the effects of various performance enhancement techniques on the parallel efficiency of an overset grid Navier-Stokes CFD application running on an SGI Origin2000 machine. Specifically, the role of asynchronous communication, grid splitting, and grid grouping strategies are presented and discussed. Results indicate that performance depends critically on the level of latency hiding and the quality of load balancing across the processors.

Djomehri, J. J.↗

Fault-Tolerant and Elastic Streaming MapReduce with Decentralized Coordination

The MapReduce programming model, due to its simplicity and scalability, has become an essential tool for processing large data volumes in distributed environments. Recent Stream Processing Systems (SPS) extend this model to provide low-latency analysis of high-velocity continuous data streams. However, integrating MapReduce with streaming poses challenges: first, the runtime variations in data characteristics such as data-rates and key-distribution cause resource overload, that inturn leads to fluctuations in the Quality of the Service (QoS); and second, the stateful reducers, whose state depends on the complete tuple history, necessitates efficient fault-recovery mechanisms to maintain the desired QoS in the presence of resource failures. We propose an integrated streaming MapReduce architecture leveraging the concept of consistent hashing to support runtime elasticity along with locality-aware data and state replication to provide efficient load-balancing with low-overhead fault-tolerance and parallel fault-recovery from multiple simultaneous failures. Our evaluation on a private cloud shows up to 2:8 improvement in peak throughput compared to Apache Storm SPS, and a low recovery latency of 700 -1500 ms from multiple failures.

97 MATHEMATICS AND COMPUTING↗

A Comparison Study of Offshore Wind Support Structures with Monopiles and Jackets for U.S. Waters: Preprint

U.S. experience in offshore wind is limited, and high costs are expected unless innovations are introduced in one or multiple aspects of the project, from the installed technology to the balance of system (BOS). The substructure is the main single component responsible for the BOS capital expenditure (CapEx) and thus one that, if improved, could yield significant levelized cost of energy (LCOE) savings. For projects in U.S. waters, multimember, lattice structures (also known as jackets) can render required stiffness for transitional water depths at potentially lower costs than monopiles (MPs). In this study, we used a systems engineering approach to evaluate the LCOE of prototypical wind power plants at six locations along the eastern seaboard and the Gulf of Mexico for both types of support structures. Using a reference wind turbine and actual metocean conditions for the selected sites, we calculated loads for a parked and an operational situation, and we optimized the MP- and jacket-based support structures to minimize their overall mass. Using a suite of cost models, we then computed their associated LCOE. For all water depths, the MP-based configurations were heavier than their jacket counterparts, but the overall costs for the MPs were less than they were for jackets up to depths of slightly less than 30 m. When the associated manufacturing and installation costs were included, jackets resulted in lower LCOE for depths greater than 40 m. These results can be used by U.S. stakeholders to understand the potential for different technologies at different sites, but the methodology illustrated in this study can be further employed to analyze the effects of innovations and design choices throughout wind power plant systems.

17 WIND ENERGY↗