Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Synchronization”

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 1,351 records · Page 75

Dynamic Time-Variant Connection Management for PGAS Models on InfiniBand

InfiniBand (IB) has established itself as a promising network infrastructure for high-end cluster computing systems as evidenced by its usage in the Top500 supercomputers today. While the IB standard describes multiple communication models (including reliable-connection (RC), and unreliable datagram (UD)), most of its promising features such as remote direct memory access (RDMA), hardware atomics and network fault tolerance are only available for the RC model which requires connections between communicating process pairs. In the past, several researchers have proposed on-demand connection management techniques that establish connections when there is a need to communicate, and not before. While such techniques work well for algorithms and applications that only communicate with a small set of processes in their life-time, there exists a broad set of applications that do not follow this trend. For example, applications that perform dynamic load balancing and adaptive work stealing have a small set of communicating neighbors at any given time, but over time the total number of neighbors can be very high; in some cases, equal to the entire system size. In this paper, we present a dynamic time-variant connection management approach that establishes connections on-demand like previous approaches, but further intelligently tears down some of the unused connections as well. While connection tear-down itself is relevant for any programming model, different models have different complexities. In this paper, we study the Global Arrays (GA) PGAS model for two reasons: (1) the simple one-sided communication primitives provided by GA and other PGAS models ensure that connection requests are always initiated by the origin process without explicit synchronization with the target process---this makes connection tear-down simpler to handle; and (2) GA supports several applications that demonstrate this behavior making it an obvious first target for the proposed enhancements. We evaluate our proposed approach using several micro-benchmarks as well as the NWChem computational chemistry application on more than 6000 processes, and show that our approach can significantly reduce the memory requirements of the communication library while maintaining its performance. To the best of our knowledge, this is the first design, implementation and evaluation of connection tear-down protocols over InfiniBand.

99 GENERAL AND MISCELLANEOUS↗

Fault-tolerance for exascale systems.

Periodic, coordinated, checkpointing to disk is the most prevalent fault tolerance method used in modern large-scale, capability class, high-performance computing (HPC) systems. Previous work has shown that as the system grows in size, the inherent synchronization of coordinated checkpoint/restart (CR) limits application scalability; at large node counts the application spends most of its time checkpointing instead of executing useful work. Furthermore, a single component failure forces an application restart from the last correct checkpoint. Suggested alternatives to coordinated CR include uncoordinated CR with message logging, redundant computation, and RAID-inspired, in-memory distributed checkpointing schemes. Each of these alternatives have differing overheads that are dependent on both the scale and communication characteristics of the application. In this work, using the Structural Simulation Toolkit (SST) simulator, we compare the performance characteristics of each of these resilience methods for a number of HPC application patterns on a number of proposed exascale machines. The result of this work provides valuable guidance on the most efficient resilience methods for exascale systems.

99 GENERAL AND MISCELLANEOUS↗

Singular value decomposition utilizing parallel algorithms on graphical processors

One of the current challenges in underwater acoustic array signal processing is the detection of quiet targets in the presence of noise. In order to enable robust detection, one of the key processing steps requires data and replica whitening. This, in turn, involves the eigen-decomposition of the sample spectral matrix, Cx = 1/K xKX(k)XH(k) where X(k) denotes a single frequency snapshot with an element for each element of the array. By employing the singular value decomposition (SVD) method, the eigenvectors and eigenvalues can be determined directly from the data without computing the sample covariance matrix, reducing the computational requirements for a given level of accuracy (van Trees, Optimum Array Processing). (Recall that the SVD of a complex matrix A involves determining V, , and U such that A = U VH where U and V are orthonormal and is a positive, real, diagonal matrix containing the singular values of A. U and V are the eigenvectors of AAH and AHA, respectively, while the singular values are the square roots of the eigenvalues of AAH.) Because it is desirable to be able to compute these quantities in real time, an efficient technique for computing the SVD is vital. In addition, emerging multicore processors like graphical processing units (GPUs) are bringing parallel processing capabilities to an ever increasing number of users. Since the computational tasks involved in array signal processing are well suited for parallelization, it is expected that these computations will be implemented using GPUs as soon as users have the necessary computational tools available to them. Thus, it is important to have an SVD algorithm that is suitable for these processors. This work explores the effectiveness of two different parallel SVD implementations on an NVIDIA Tesla C2050 GPU (14 multiprocessors, 32 cores per multiprocessor, 1.15 GHz clock - peed). The first algorithm is based on a two-step algorithm which bidiagonalizes the matrix using Householder transformations, and then diagonalizes the intermediate bidiagonal matrix through implicit QR shifts. This is similar to that implemented for real matrices by Lahabar and Narayanan ("Singular Value Decomposition on GPU using CUDA", IEEE International Parallel Distributed Processing Symposium 2009). The implementation is done in a hybrid manner, with the bidiagonalization stage done using the GPU while the diagonalization stage is done using the CPU, with the GPU used to update the U and V matrices. The second algorithm is based on a one-sided Jacobi scheme utilizing a sequence of pair-wise column orthogonalizations such that A is replaced by AV until the resulting matrix is sufficiently orthogonal (that is, equal to U ). V is obtained from the sequence of orthogonalizations, while can be found from the square root of the diagonal elements of AH A and, once is known, U can be found from column scaling the resulting matrix. These implementations utilize CUDA Fortran and NVIDIA's CUB LAS library. The primary goal of this study is to quantify the comparative performance of these two techniques against themselves and other standard implementations (for example, MATLAB). Considering that there is significant overhead associated with transferring data to the GPU and with synchronization between the GPU and the host CPU, it is also important to understand when it is worthwhile to use the GPU in terms of the matrix size and number of concurrent SVDs to be calculated.

99 GENERAL AND MISCELLANEOUS↗

The On-Orbit Calibrations for the Fermi Large Area Telescope

The Large Area Telescope (LAT) on-board the Fermi Gamma-ray Space Telescope began its on-orbit operations on June 23, 2008. Calibrations, defined in a generic sense, correspond to synchronization of trigger signals, optimization of delays for latching data, determination of detector thresholds, gains and responses, evaluation of the perimeter of the South Atlantic Anomaly (SAA), measurements of live time, of absolute time, and internal and spacecraft boresight alignments. Here we describe on-orbit calibration results obtained using known astrophysical sources, galactic cosmic rays, and charge injection into the front-end electronics of each detector. Instrument response functions will be described in a separate publication. This paper demonstrates the stability of calibrations and describes minor changes observed since launch. These results have been used to calibrate the LAT datasets to be publicly released in August 2009.

79 ASTRONOMY AND ASTROPHYSICS↗

Towards Flexible Exascale Stream Processing System Simulation

Stream processing is an important emerging computational model for performing complex operations on and across multi-source, high-volume, unpredictable dataflows. We present Flow, a platform for parallel and distributed stream processing system simulation that provides a flexible modeling environment for analyzing stream processing applications. The Flow stream processing system simulator is a high-performance, scalable simulator that automatically parallelizes chunks of the model space and incurs near-zero synchronization overhead for acyclic stream application graphs. We show promising parallel and distributed event rates exceeding 149 million events per second on a cluster with 512 processor cores.

99 GENERAL AND MISCELLANEOUS↗

Status of the Short-Pulse X-ray Project at the Advanced Photon Source

The Advanced Photon Source Upgrade (APS-U) Project at Argonne will include generation of short-pulse x-rays based on Zholents deflecting cavity scheme. We have chosen superconducting (SC) cavities in order to have a continuous train of crabbed bunches and flexibility of operating modes. In collaboration with Jefferson Laboratory, we are prototyping and testing a number of single-cell deflecting cavities and associated auxiliary systems with promising initial results. In collaboration with Lawrence Berkeley National Laboratory, we are working to develop state-of-the-art timing, synchronization, and differential rf phase stability systems that are required for SPX. Collaboration with Advanced Computations Department at Stanford Linear Accelerator Center is looking into simulations of complex, multi-cavity geometries with lower- and higher-order modes waveguide dampers using ACE3P. This contribution provides the current R&D status of the SPX project.

43 PARTICLE ACCELERATORS↗

Terahertz radiation as a bunch diagnostic for laser-wakefield-accelerated electron bunches

Experimental results are reported from two measurement techniques (semiconductor switching and electro-optic sampling) that allow temporal characterization of electron bunches produced by a laser-driven plasma-based accelerator. As femtosecond electron bunches exit the plasma-vacuum interface, coherent transition radiation (at THz frequencies) is emitted. Measuring the properties of this radiation allows characterization of the electron bunches. Theoretical work on the emission mechanism is presented, including a model that calculates the THz wave form from a given bunch profile. It is found that the spectrum of the THz pulse is coherent up to the 200 {micro}m thick crystal (ZnTe) detection limit of 4 THz, which corresponds to the production of sub-50 fs (rms) electron bunch structure. The measurements demonstrate both the shot-to-shot stability of bunch parameters that are critical to THz emission (such as total charge and bunch length), as well as femtosecond synchronization among bunch, THz pulse, and laser beam.

43 PARTICLE ACCELERATORS↗

Empirical Evaluation of GPS Clock Accuracy for Isochronous Droop-Based Inverters: Preprint

Grid-forming inverters rely upon an internal reference to regulate the grid’s voltage and frequency. Without a reliable reference, issues such as loss of synchronism, frequency instability, and unequal power sharing can occur. One proposed internal phase reference for grid-forming inverters is generated by the controller’s clock, but this is affected by nonidealities in the crystal oscillator leading to time drift in the phase reference. To improve the controller’s clock accuracy, the Global Positioning System (GPS) can be used to calibrate the internal clock. Previously, GPS-based clock calibration was explored theoretically and various methods were developed for mitigating frequency/phase drift at the power system level. This paper experimentally validates the direct performance of GPS time-calibration at the controller level and justifies its use in maintaining an accurate phase reference for droop-based grid-forming inverters.

clock drift↗

On the Low Risk of Subsynchronous Resonance (SSR) in Type III Wind Turbines Operating in Grid-Forming Control

We have shown in a previous work that the risk of sub-synchronous resonance (SSR) between a wind power plant with Type III wind turbines and series-compensated transmission lines is low when the wind turbines in the plant are operated in grid-forming mode, instead of the standard grid-following mode. The fundamental mechanism behind the improved damping characteristics is explained in this paper by modeling the positive and negative sequence impedances of Type III wind turbines for GFM operation mode. It is discovered that the GFM control naturally acts against the negative resistance of Type III wind turbines that results from an interaction between the proportional gain of the rotor-side converter current controller and negative slip. The developed sequence impedance models and improved damping behavior are verified using PSCAD simulations of a 2.5-MW Type III GFM wind turbine. The modeling predictions are also supported by experimentally measuring the sequence impedance response of a 2.5-MW Type III wind turbine during operation in GFL and GFM modes.

grid-following turbine↗

Study of Seamless Microgrid Transition Operation Using Grid-Forming Inverters

This paper investigates operational techniques to achieve seamless (smooth) microgrid (MG) transitions by dispatching a grid-forming (GFM) inverter. In traditional approaches, the GFM inverter must switch between grid-following (GFL) and GFM control modes during MG transition operation. Today's inverter technology allows GFM inverters to always operate in GFM control mode, so it is worth exploring how to use them to achieve smooth MG transition operation. This paper proposes three operational techniques: a traditional scheme of switching between GFL and GFM control; a new scheme of consistent GFM control and shifting the droop intercept up before islanding operation; and a new scheme of consistent GFM control and shifting the droop intercept up before synchronization operation. A full hardware setup is established to compare the three techniques and showcase their implementations in real-world applications. The results show that the third technique outperforms the others and exhibits the best transition performance because the GFM inverter maintains the same operating points during the transition operation. Therefore, we conclude that ensuring smooth MG transition operation requires that the GFM inverter(s) maintain the same operating points (v, f, P, Q, and phase angle) during the transition operation in addition to minimizes the point of common coupling power flow.

grid-forming↗

Study of Seamless Microgrid Transition Operation Using Grid-Forming Inverters: Preprint

This paper investigates operational techniques to achieve smooth microgrid (MG) transitions by dispatching the grid-forming (GFM) inverter. In traditional approaches, the GFM inverter must switch between grid-following (GFL) and GFM control mode during MG transition operation. Today's inverter technology allows GFM inverters to always operate in GFM control mode, so it is worthy to explore how to use them to achieve smooth MG transition operation. This paper proposes three operational techniques: a traditional scheme of switching between GFL and GFM control; a new scheme of consistent GFM control and shifting the droop intercept up before islanding operation; and a new scheme of consistent GFM control and shifting the droop intercept up before synchronization operation. A full hardware setup is established to compare the three techniques and showcase their implementations in real-world applications. As expected, the third technique outperforms the others and exhibits the best transition performance because the GFM inverter maintains the same operating points during the transition operation. Therefore, we conclude that ensuring smooth MG transition operation requires the GFM inverter(s) to maintain the same operating points (v, f, P, Q, and phase angle) during the transition operation in addition to minimizes the point of common coupling power flow.

grid-following control↗

Upgrade of High-Energy X-Ray real-time radioscopy for KROTOS experiment

As part of its R and D programs on severe accidents, in particular on understanding of corium-water interaction, CEA is commissioning an update of the KROTOS experiment at Cadarache. The Xray imaging setup (high energy real-time radioscopy) is upgraded to provide the best performances for the new experimental program. In order to fit the performance needs (faster acquisition, smaller detection limit and higher field of view), two radioscopy setups, with two linear accelerators (linacs 9 MV and 6 MV), are placed in the irradiation cell. Having discussed the expected performances in terms of frequency, detection limit and field-of-view, this article details each stage of both radioscopy chains: principles and technical characteristics. Then, linacs and cameras synchronization (at few hundred Hertz), data flows and storage setups are detailed. Finally, experimental characterizations and performance validations on phantom are presented. (authors)

43 PARTICLE ACCELERATORS↗

Infrastructure-Based Cooperative Perception at a Traffic Intersection: Overview and Challenges: Preprint

Recent advancement in autonomous driving vehicles and V2X communication has attracted increasing attention towards Intelligent Transportation Systems to build a safe and reliable traffic intersection. However, most of the systems are still at the initial stages and require significant progress to become a reality. This paper presents an overview of NREL Infrastructure Perception and Control (IPC) framework which is an open-source track-data fusion engine which takes input from infrastructure-based perception sensors and cooperatively shared messages from Connected Autonomous Vehicles (CAVs) and Connected Vehicle (CVs) and the challenges associated with deploying such cooperative perception framework at a four-way traffic intersection in the city of Colorado Springs, CO, USA. The sensor data is collected by deploying two radars and two LiDAR sensors on the IPC mobile lab and two radars on diagonally opposite traffic poles at the proposed intersection. The sensor output results imply the need for rapid sensor calibration to bring the collective perception to a common coordinate frame, the importance of time synchronization between the sensors in order to capture accurate spatial and temporal alignment of the objects, and the need for a health monitoring system with fail safe closed-loop detection model for real-time deployment.

camera↗

Demonstration and Optimization of a Drive Laser for an X-Band Photoinjector

Recently, a drive laser for an S-band (2.86 GHz) rf photoinjector, designed to provide a pulse that has a flat temporal and spatial profile, has been built, commissioned, and put into service as part of the LLNL Compton-scattering monoenergetic {gamma}-ray source program. This laser is based on an all-fiber oscillator and front-end amplification system, and provides both the laser light to generate the electrons as well as the rf signal that is amplified to accelerate them. Now, a new 11.424 GHz photoinjector is being developed, which has required a revised design for the laser system. The higher frequency has placed more stringent requirements on the synchronization stability, delivered pulse length, and pulse rise times to maintain the desired emittance. Presented here are the overall design and measured performance of the current system and a discussion of what changes are being made to address observed shortcomings and more demanding requirements to make the system ready for the next-generation Compton-scattering monoenergetic {gamma}-ray source.

42 ENGINEERING↗

A 10-Hz Terawatt Class Ti:Sapphire Laser System: Development and Applications

We developed a two stage Ti:Sapphire laser system to generate 16 mJ/80fs laser pulses at the pulse repetition rate of 10 Hz. The key deriver for the present design is implementing a highly efficient symmetric confocal pre-amplifier and employing a simple, inexpensive synchronization scheme relying only on a commercial digital delay-generator. We characterized the amplified pulses in spatial-, spectral-, and temporal-domains. The laser system was used to investigate various nonlinear optical processes, and to modify the optical properties of metal- and semiconductor-surfaces. We are currently building a third amplifier to boost the laser power to the multi-terawatt range.

43 PARTICLE ACCELERATORS↗

Experiments with Memory-to-Memory Coupling for End-to-End fusion Simulation Workflows

Scientific applications are striving to accurately simulate multiple interacting physical processes that comprise complex phenomena being modeled. Efficient and scalable parallel implementations of these coupled simulations present challenging interaction and coordination requirements, especially when the coupled physical processes are computationally heterogeneous and progress at different speeds. In this paper, we present the design, implementation and evaluation of a memory-to-memory coupling framework for coupled scientific simulations on high-performance parallel computing platforms. The framework is driven by the coupling requirements of the Center for Plasma Edge Simulation, and it provides simple coupling abstractions as well as efficient asynchronous (RDMA-based) memory-to-memory data transport mechanisms that complement existing parallel programming systems and data sharing frameworks. The framework enables flexible coupling behaviors that are asynchronous in time and space, and it supports dynamic coupling between heterogeneous simulation processes without enforcing any synchronization constraints. We evaluate the performance and scalability of the coupling framework using a specific coupling scenario, on the Jaguar Cray XT5 system at Oak Ridge National Laboratory.

99 GENERAL AND MISCELLANEOUS↗

Maestro: Data Orchestation and Tuning for OpenCL Devices

Abstract. As heterogeneous computing platforms become more prevalent, the programmer must account for complex memory hierarchies in addition to the diculties of parallel programming. OpenCL is an open standard for parallel computing that helps alleviate this diculty by providing a portable set of abstractions for device memory hierarchies. However, OpenCL requires that the programmer explicitly controls data transfer and device synchronization, two tedious and error-prone tasks. This paper introduces Maestro, an open source library for data orchestration on OpenCL devices. Maestro provides automatic data transfer, task decomposition across multiple devices, and autotuning of dynamic execution parameters for some types of problems.

99 GENERAL AND MISCELLANEOUS↗

Stereo photogrammetry reveals substantial drag on cloud thermals: STEREO PHOTOGRAMMETRY OF CLOUD THERMALS

Stereo photogrammetry, which uses two synchronized cameras to measure three-dimensional positions, is applied here to ascertain whether drag plays a role in the ascent of cloud thermals. In particular, stereo cameras are used to measure the sizes and speeds of cloud thermals in Florida. Using the vertical momentum equation, it is found that a substantial amount of drag (a drag coefficient on the order of 1) is needed to match both the stereo-photogrammetric data and the known buoyancy of clouds from previous in situ measurements and large-eddy simulations. Empirical data on form drag and theoretical calculations of wave drag reveal that, for the observed Froude numbers of cloud thermals, a drag coefficient of about one is to be expected.

54 ENVIRONMENTAL SCIENCES↗