Search NASA⌕ Search

SEARCH · Search NASA

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 595 records · Page 33

LAURA Users Manual: 5.6

This users manual provides in-depth information concerning installation and execution of Laura, version 5. Laura is a structured, multiblock, computational aerothermodynamic simulation code. Version 5 represents a major refactoring of the original Fortran 77 Laura code toward a modular structure afforded by Fortran 95. The refactoring improved usability and maintainability by eliminating the requirement for problem-dependent recompilations, providing more intuitive distribution of functionality, and simplifying inter- faces required for multi-physics coupling. As a result, Laura now shares gas-physics modules, MPI modules, and other low-level modules with the Fun3D unstructured-grid code. In addition to internal refactoring, several new features and capabilities have been added, e.g., a GNU-standard installation process, parallel load balancing, automatic trajectory point sequencing, free-energy minimization, and coupled ablation and flow field radiation.

Aerodynamics↗

DC 12m telescope. Preliminary calculations. Investigation of elevation axis position.

This paper examines some simple calculations of a 2D model of a telescope in order to understand how different design parameters affect the design. For the design of a telescope it is assumed that they need a design that minimizes deflections of the dish and also minimizes the size of the motors and torques needed to rotate in elevation. A common belief is that a lighter dish and minimum counterweight is desirable. However, these calculations show this is not necessarily true. The torque needed for rotation depends on the moment of inertia and if the telescope is balanced about the elevation axis. A light dish with no CW requires that the elevation axis be several meters in front of the dish (8-9m) in order to be balanced. This is not practical from a structural point of view. If the elevation axis is only 2m in front of the dish and there is no counterweight then the telescope will be unbalanced and the toruqes required will be very high - much higher than the torques needed only to overcome inertia. A heavy dish though can act as its own counterweight and the elevation axis only has to be 2-3m in front of the dish in order to achieve a balanced telescope. Also the struts that support the camera from the dish place a load on the dish which will put a bending moment on the dish. This bending moment will deform the dish and require it to be stiffer. A counterweight structure performs two functions. First, it allows the telescope to be balanced about the elevation axis. Second, it applies a force on the dish that opposes the forces from the camera struts, thereby reducing the bending moment and deformations of the dish.

47 OTHER INSTRUMENTATION↗

Analytical design and performance studies of the nuclear light bulb engine

Analytical studies were conducted to investigate in detail the heat balance characteristics of the nuclear light bulb engine. Distributions of energy deposition to all engine components from the fission process, conduction and convection, and thermal radiation were considered. Where uncertainties in basic data or heat transfer characteristics were encountered, ranges of heat loads were calculated and reference values were selected. The influence of these heat loads on engine performance, space radiator requirements, and cooling sequence and cooling circuit designs was determined. The analyses resulted in revisions to the previously reported reference engine characteristics, principally in the heat loads to some engine components and in the cooling sequence. These revisions were incorporated in the engine dynamics digital computer simulation program. No significant changes occurred in the dynamic response of the engine to perturbations in fuel injection rate, reactivity or exhaust nozzle area.

Rodgers, R. J.↗

LAURA Users Manual: 5.3-48528

This users manual provides in-depth information concerning installation and execution of LAURA, version 5. LAURA is a structured, multi-block, computational aerothermodynamic simulation code. Version 5 represents a major refactoring of the original Fortran 77 LAURA code toward a modular structure afforded by Fortran 95. The refactoring improved usability and maintainability by eliminating the requirement for problem-dependent re-compilations, providing more intuitive distribution of functionality, and simplifying interfaces required for multi-physics coupling. As a result, LAURA now shares gas-physics modules, MPI modules, and other low-level modules with the FUN3D unstructured-grid code. In addition to internal refactoring, several new features and capabilities have been added, e.g., a GNU-standard installation process, parallel load balancing, automatic trajectory point sequencing, free-energy minimization, and coupled ablation and flowfield radiation.

Mazaheri, Alireza↗

FPDeep: Scalable Acceleration of CNN Training on Deeply-Pipelined FPGA Clusters

In this paper, we propose a framework called FPDeep, which uses a hybrid of model and layer paral- lelism to configure distributed reconfigurable clusters to train DNNs. This approach has numerous benefits. First, the design does not suffer from batch size growth. Second, novel workload and weight partitioning leads to balanced loads of both among nodes. And third, the entire system is fine-grained pipeline. This leads to high parallelism and utilization and also minimizes the time features need to be cached while waiting for back-propagation.

Wang, Tianqi↗

Space Station flexible dynamics under plume impingement

Assembly of the Space Station requires numerous construction flights by the Space Shuttle. A particularly challenging problem is that of control of each intermediate station configuration when the shuttle orbiter is approaching it to deliver the next component. The necessary braking maneuvers cause orbiter thruster plumes to impinge on the station, especially its solar arrays. This in turn causes both overall attitude errors and excitation of flexible-body vibration modes. These plume loads are predicted to lead to CMG saturation during the approach of the orbiter to the SC-5 station configuration, necessitating the use of the station RCS jets for desaturation. They are also expected to lead to significant excitation of solar array vibrations. It is therefore of great practical importance to investigate the effects of plume loads on the flexible dynamics of station configuration SC-5 as accurately as possible. However, this system possesses a great many flexible modes (89 below 5 rad/s), making analysis time-consuming and complicated. Model reduction techniques can be used to overcome this problem, reducing the system model to one which retains only the significant dynamics, i.e. those which are strongly excited by the control inputs or plume disturbance forces and which strongly couple with the measured outputs. The particular technique to be used in this study is the subsystem balancing approach which was previously developed by the present investigator. This method is very efficient computationally. Furthermore, it gives accurate results even for the difficult case where the structure has many closed-spaced natural frequencies, when standard modal truncation can give misleading results. Station configuration SC-5 is a good example of such a structure.

Williams, Trevor↗

Adaptive Numerical Algorithms in Space Weather Modeling

Space weather describes the various processes in the Sun-Earth system that present danger to human health and technology. The goal of space weather forecasting is to provide an opportunity to mitigate these negative effects. Physics-based space weather modeling is characterized by disparate temporal and spatial scales as well as by different physics in different domains. A multi-physics system can be modeled by a software framework comprising of several components. Each component corresponds to a physics domain, and each component is represented by one or more numerical models. The publicly available Space Weather Modeling Framework (SWMF) can execute and couple together several components distributed over a parallel machine in a flexible and efficient manner. The framework also allows resolving disparate spatial and temporal scales with independent spatial and temporal discretizations in the various models. Several of the computationally most expensive domains of the framework are modeled by the Block-Adaptive Tree Solar wind Roe Upwind Scheme (BATS-R-US) code that can solve various forms of the magnetohydrodynamics (MHD) equations, including Hall, semi-relativistic, multi-species and multi-fluid MHD, anisotropic pressure, radiative transport and heat conduction. Modeling disparate scales within BATS-R-US is achieved by a block-adaptive mesh both in Cartesian and generalized coordinates. Most recently we have created a new core for BATS-R-US: the Block-Adaptive Tree Library (BATL) that provides a general toolkit for creating, load balancing and message passing in a 1, 2 or 3 dimensional block-adaptive grid. We describe the algorithms of BATL and demonstrate its efficiency and scaling properties for various problems. BATS-R-US uses several time-integration schemes to address multiple time-scales: explicit time stepping with fixed or local time steps, partially steady-state evolution, point-implicit, semi-implicit, explicit/implicit, and fully implicit numerical schemes. Depending on the application, we find that different time stepping methods are optimal. Several of the time integration schemes exploit the block-based granularity of the grid structure. The framework and the adaptive algorithms enable physics based space weather modeling and even forecasting.

Toth, Gabor↗

Platform-Independence and Scheduling In a Multi-Threaded Real-Time Simulation

Aviation research often relies on real-time, pilot-in-the-loop flight simulation as a means to develop new flight software, flight hardware, or pilot procedures. Often these simulations become so complex that a single processor is incapable of performing the necessary computations within a fixed time-step. Threads are an elegant means to distribute the computational work-load when running on a symmetric multi-processor machine. However, programming with threads often requires operating system specific calls that reduce code portability and maintainability. While a multi-threaded simulation allows a significant increase in the simulation complexity, it also increases the workload of a simulation operator by requiring that the operator determine which models run on which thread. To address these concerns an object-oriented design was implemented in the NASA Langley Standard Real-Time Simulation in C++ (LaSRS++) application framework. The design provides a portable and maintainable means to use threads and also provides a mechanism to automatically load balance the simulation models.

Sugden, Paul P.↗

Modeling Smart Microgrids for the Developing World with Probabilistic Supply and Demand Inputs

Developing world microgrids often balance insufficient supply with growing, unpredictable demand. Deterministic and probabilistic simulators exist to model these microgrids, and each focuses on different technical aspects. With the addition of smart meters into microgrids, monitoring and control is now available at high granularity, which enriches microgrid planning and operation. This research is designing a new simulator to model smart microgrids with discrete probability distributions as supply and demand inputs. In our model, smart meters allow real-time power clipping for demand side management, effectively smoothing the system load curve as needed. To compare clipping schemes for grid operation and generation mixes for planning, we aggregate inputs by convolution then compute expected energy sold and probability of avoiding power cuts. The simulator plots these values for different combinations of power clipping threshold and number of customers clipped.

13 HYDRO ENERGY↗

ECP Milestone Deliverable Memo 2.3.1.04.15

The DARMA many-task framework provides asynchronous communication and load balancing functionality. This functionality is embedded in standard , modern C++ through the use of the template wrapper classes similar to futures. DARMA codes previously could not interoperate with MPI. A new C++ interface and extended semantics now allow quiescence of DARMA kernels and transfer of data ownership back into MPI, allowing 1) isolated DARMA kernels to be inserted in a larger MPI code or 2) reuse of MPI libraries like solvers within a DARMA application.

97 MATHEMATICS AND COMPUTING↗

TBAA20: Task-Based Algorithms and Applications

The new challenges posed by Exascale system architectures have resulted in difficulty achieving a desired scalability using traditional distributed ­memory runtimes. Task­-based programming models show promise in addressing these challenges, providing application developers with a productive and performant approach to programming on next generation systems. Empirical studies show that task-based models can overcome load ­balancing issues that are inherent to traditional distributed ­memory runtimes, and that task-­based runtimes perform comparably to those systems when balanced. This panel is designed to explore the advantages of task-­based programming models on modern and future HPC systems from an industry, university, and national lab perspective. It aims at gathering application experts and proponents of these models to present concrete and practical examples of using task­-based runtimes to overcome the challenges posed by Exascale system architectures. This report describes the objectives, activities, and outcomes of the panel TBAA: Task­-Based Algorithms and Applications which was held at the International Conference for High Performance Computing, Networking, Storage, and Analysis (SC 20) on November 18, 2020.

97 MATHEMATICS AND COMPUTING↗

Creep testing of foil-gage metals at elevated temperature using an automated data acquisition system

A method is being developed to obtain creep data on foil gage metals at elevated temperatures using an automated data acquisition system in conjunction with a mechanically counter balanced extensometer. The automated system components include the Hewlett-Packard (HP) 9845A desktop computer, the HP 3455A digital voltmeter and the HP 3495A scanner. Software for test monitoring and data collection was developed; data manipulation, including curve plotting was done with a HP regression analysis software package. Initial creep tests were conducted on .003 in. thick foil specimens of Ti-6A1-4V at temperatures of 800 F and 1000 F and at stress levels of 25 ksi and 45 ksi. For comparison, duplicate tests were run on .049 in. thick specimens sheet of the same alloy. During testing, the furnace and specimen temperature, bridge voltage, strain and load output were automatically monitored and recorded at predetermined intervals. Using the HP regression analysis program, recorded strain output was plotted as a function of time. These resultant creep curves indicate that, under similar conditions of temperature and stress, foil gage specimens exhibit a higher creep rate than sheet specimens.

Blackburn, L. B.↗

PC-RPL: Joint Control of Routing Topology and Transmission Power in Real Low-Power and Lossy Networks

We present PC-RPL , a transmission power-controlled IPv6 routing protocol for low-power and lossy wireless networks that significantly improves the end-to-end packet delivery performance under heavy traffic compared to the standard RPL. We show through actual design, implementation, and experiments that a multihop wireless network can achieve better throughput and routing stability when transmission power and routing topology are “jointly and adaptively” controlled. Our experiments show that the predominant “fixed and uniform” transmission power strategy with “link quality and hop distance”–based routing topology construction (i.e., RPL) loses significant bandwidth due to hidden terminal and load imbalance problems. We design an adaptive and distributed control mechanism for transmission power and routing topology, named PC-RPL , on top of the standard RPL routing protocol for hidden terminal mitigation and load balancing. We implement PC-RPL on real embedded devices and evaluate its performance on a 49-node multihop testbed. PC-RPL reduces total end-to-end packet losses by approximately sevenfold without increasing hop distance compared to RPL with the highest transmission power, resulting in 17% improvement in aggregate bandwidth and 64% improvement for the worst-case node by successfully alleviating both hidden terminal and load imbalance problems.

Computer Science↗

Simulation capability for dynamics of two-body flexible satellites

An analysis and computer program were prepared to realistically simulate the dynamic behavior of a class of satellites consisting of two end bodies separated by a connecting structure. The shape and mass distribution of the flexible end bodies are arbitrary; the connecting structure is flexible but massless and is capable of deployment and retraction. Fluid flowing in a piping system and rigid moving masses, representing a cargo elevator or crew members, have been modeled. Connecting structure characteristics, control systems, and externally applied loads are modeled in easily replaced subroutines. Subroutines currently available include a telescopic beam-type connecting structure as well as attitude, deployment, spin and wobble control. In addition, a unique mass balance control system was developed to sense and balance mass shifts due to the motion of a cargo elevator. The mass of the cargo may vary through a large range. Numerical results are discussed for various types of runs.

Austin, F.↗

Passive Rocket Diffuser Theory: A Re-Examination of Minimum Second Throat Size

Second-throat diffusers serve to isolate rocket engines from the effects of ambient back pressure during testing without using active control systems. Among the most critical design parameters is the relative area of the diffuser throat to that of the nozzle throat. A smaller second throat is generally desirable because it decreases the stagnation-to-ambient pressure ratio the diffuser requires for nominal operation. There is a limit, however. Below a certain size, the second throat can cause pressure buildup within the diffuser and prevent it from reaching the start condition that protects the nozzle from side-load damage. This paper presents a method for improved estimation of the minimum second throat area which enables diffuser start. The new 3-zone model uses traditional quasi-one-dimensional compressible flow theory to approximate the structure of two distinct diffuser flow fields observed in Computational Fluid Dynamics (CFD) simulations and combines them to provide a less-conservative estimate of the second throat size limit. It is unique among second throat sizing methods in that it accounts for all major conical nozzle and second throat diffuser design parameters within its limits of application. The performance of the 3-zone method is compared to the historical normal shock and force balance methods, and verified against a large number of CFD simulations at specific heat ratios of 1.4 and 1.25. Validation is left as future work, and the model is currently intended to function only as a first-order design tool.

Jones, Daniel R.↗

Simulation capability for dynamics of rotating counterweight space stations.

An analysis and computer program were prepared to realistically simulate the dynamic behavior of a class of space stations consisting of two end bodies separated by a connecting structure. The shape and mass distribution of the flexible end bodies are arbitrary; the connecting structure is flexible but massless and is capable of deployment and retraction. Fluid flowing in a piping system and rigid moving masses, representing a cargo elevator or crew members, have been modeled. Connecting structure characteristics, control systems, and externally applied loads are modeled in easily replaced subroutines. Subroutines currently available include a telescopic beam-type connecting structure as well as attitude, deployment, spin, and wobble control. In addition, a unique mass balance control system was developed to sense and balance mass shifts due to the motion of a cargo elevator. The mass of the cargo may vary through a large range. Numerical results are discussed for various types of runs.

Austin, F.↗

Design of a Low Speed Fan Stage for Noise Suppression

This report describes the design of a low tip speed, moderate pressure rise fan stage for demonstration of noise reduction concepts. The fan rotor is a fixed-pitch configuration delivering a design pressure ratio of 1.378 at a specific flow of 43.1 lbm/sec/sq ft. Four exit stator configurations were provided to demonstrate the effectiveness of circumferential and axial sweep in reducing rotor-stator interaction tone noise. The fan stage design was combined with an axisymmetric inlet, conical convergent nozzle, and nacelle to form a powered fan-nacelle subscale model. This model has a 22-inch cylindrical flow path and employs a rotor with a 0.30 hub-to-tip radius ratio. The design is fully compatible with an existing NASA force balance and rig drive system. The stage aerodynamic and structural design is described in detail. Three-dimensional (3-D) computational fluid dynamics (CFD) tools were used to define optimum airfoil sections for both the rotor and stators. A fan noise predictive system developed by Pratt & Whitney under contract to NASA was used to determine the acoustic characteristics of the various stator configurations. Parameters varied included rotor-to-stator spacing and vane leading edge sweep. The structural analysis of the rotor and stator are described herein. An integral blade and disk configuration was selected for the rotor. Analysis confirmed adequate low cycle fatigue life, vibratory endurance strength, and aeroelastic suitability. A unique load carrying stator arrangement was selected to minimize generation of tonal noise due to sources other than rotor-stator interaction. Analysis of all static structural components demonstrated adequate strength, fatigue life, and vibratory characteristics.

Dalton, W. N.↗

Data Locality Enhancement of Dynamic Simulations for Exascale Computing (Final Report)

The development of modern processors exhibits two trends that complicate the optimizations of modern software. The first is the increasing sensitivity of processors' throughput to irregularities in computation. With more processors produced through a massive integration of simple cores, future systems will increasingly favor regular data-level parallel computations, but deviate from the needs of applications with complex patterns. Some evidences are already shown on Graphic Processing Units (GPU): Irregular data accesses (e.g., indirect references A[D[i]]) and conditional branches are limiting many GPU applications' performance at a level an order of magnitude lower than the peak of GPU. The second hardware trend is the growing gap between memory bandwidth and the aggregate speed—that is, the sum of all cores' computing power—of a Chip Multiprocessor (CMP). Despite the capped growth of the peak CPU speed, the aggregate speed of a CMP keeps increasing as more cores get into a single chip. It is expected that by 2018, node concurrency in an exascale system will increase by hundreds of times, whereas, memory bandwidth will expand by only 10 to 20 times. Consequently, data movement and storage is expected to consume more than 70% of the total system power. Bridging this gap is difficult; the complexities of modern CMP memory hierarchy make it even harder: Data cache becomes shared among computing units, and the sharing is often non-uniform—whether two computing units share a cache depends on their proximity and the level of the cache. On the recent IBM Power7 architecture, for instance, four hardware contexts (or SMT threads) in a core share the entire memory hierarchy, all cores in one chip share an on-chip L3 cache, and cores across chips share L3 and main memory through off-chip connections. These two trends complicate the translation of computing power into performance, especially for a program with either intensive data accesses or complex patterns in data accesses or control flow paths. Unfortunately, both attributes present and will persist in a class of important applications. For instance, many scientific simulations deal with a large volume of data. And meanwhile, as most real-world processes are non-uniform and evolving (e.g., the evolution of a galaxy or the process of a drug injection), both the computations and data accesses of these programs tend to be irregular and dynamically changing. Currently, the lack of support to these applications on modern CMP severely limits their performance. On GPU, as our recent study shows and other studies echo, performance enhancement of a factor of integers is possible when memory accesses or control flows are streamlined for a set of GPU applications. On multicore CPU, our studies show that traditional locality enhancement, for being oblivious to the new features of multicore memory hierarchy, may even cause large slowdown to data-intensive dynamic applications. The severity of the issues is expected to worsen as the two hardware trends continue. Some recent studies try to match software with the trends, but in a limited scope or manner. For irregularities on GPU, most studies focus on irregularities analyzable through static analysis (e.g., data accesses in regular loops). Dynamic irregularities are harder to address because the needed analysis and transformations typically have to happen at run time. Some other research resorts to hardware extensions, an actual adoption of which is unclear for the entailed space cost and complexity. For data locality, recent years have seen some exploitations of the new memory hierarchy on multicore for performance, but most of them are on process or thread scheduling, rather than program transformations. Our recent study reveals that program-level transformations may magnify the scheduling benefits by a factor of seven, concluding that program-level transformation should play a central role for data locality enhancement on modern CMP. But research in this direction has been sparse, and most have focused on data layout or cache performance modeling, rather than program transformations to match with the new memory hierarchy features. Overall, it is still an open question how to bridge the gap between dynamic computations and the two prominent properties of modern processors. The goal of this project is to develop a set of techniques and software tools to enhance the matching between memory accesses in dynamic simulations and the prominent features of modern and future CMP, alleviating the memory performance issues for petascale and exascale computing. This report summarizes the discoveries and products produced throughout this project. It includes free launch, a new software approach to overcoming the shortcomings of both methods; coherence-free multiview, an approach that allows multiple views of a single data object to co-exist on GPU memory during a GPU kernel execution; algorithmic optimizations to data analytics problems, especially those that involve lots of distance calculations. Supporting dynamic parallelism is important for GPU to benefit a broad range of applications. There are currently two fundamental ways for programs to exploit dynamic parallelism on GPU: a software-based approach with software-managed worklists, and a hardware-based approach through dynamic subkernel launches. Neither is satisfactory. The former is complicated to program and is often subject to some load imbalance; the latter suffers large runtime overhead. In this work, we propose free launch, a new software approach to overcoming the shortcomings of both methods. It allows programmers to use subkernel launches to express dynamic parallelism. It employs a novel compiler-based code transformation named subkernel launch removal to replace the subkernel launches with the reuse of parent threads. Coupled with an adaptive task assignment mechanism, the transformation reassigns the tasks in the subkernels to the parent threads with a good load balance. The technique requires no hardware extensions, immediately deployable on existing GPUs. It keeps the programming convenience of the subkernel launch-based approach while avoiding its large runtime overhead. Meanwhile, its superior load balancing makes it outperform manual worklist-based techniques by 3X on average. The work was published at the 48th Annual IEEE/ACM International Symposium on Microarchitecture (Micro'2015). A Graphic Processing Unit (GPU) system is typically equipped with many types of memory (e.g., global, constant, texture, shared, cache). Data placement determines what data are placed on which type of memory, essential for GPU memory performance. Prior optimizations of data placement always require a single view of a data object on memory, which limits the optimization effectiveness. In this work, we propose coherence-free multiview, an approach that allows multiple views of a single data object to co-exist on GPU memory during a GPU kernel execution. We demonstrate that under certain conditions, the multiple views can remain incoherent while facilitating enhanced data placement. We present a theorem and some compiler support to ensure the soundness of the usage of coherence-free multiview. We further develop reference-discerning data placement, a new way to enhance data placements on GPU. It enables more flexible data placements by using coherence-free multiview to leverage the slack in coherence requirement of some GPU programs. Experiments on three types of GPU systems show that, with less than 200KB space cost, the new data placement technique can provide a 1.6X average (up to 4.27X) speedup. In addition, we have examined algorithmic optimizations to data analytics problems, especially those that involve lots of distance calculations. Computing distances among data points is an essential part of many important algorithms in data analytics, graph analysis, and other domains. In each of these domains, developers have spent significant manual effort optimizing algorithms, often through novel applications of the triangle equality, in order to minimize the number of distance computations in the algorithms. In this work, we observe that many algorithms across these domains can be generalized as an instance of a generic distance-related abstraction. Based on this abstraction, we derive seven principles for correctly applying the triangular inequality to optimize distance-related algorithms. Guided by the findings, we develop Triangular OPtimizer (TOP), the first software framework that is able to automatically produce optimized algorithms that either matches or outperforms manually designed algorithms for solving distance-related problems. TOP achieves up to 237x speedups and 2.5X on average. The work has been published at the 32nd International Conference on Machine Learning (ICML'15) and the 41st International Conference on Very Large Data Bases (VLDB'15).

42 ENGINEERING↗