6957792 research outputs found
Sort by
One Year Term Review as a Participating Guest in the Detonator and Detonation Physics Group
No full textExploiting the capabilities of modern GPUs for dense matrix computations Exploiting the capabilities of modern GPUs for dense matrix computations Utilización de GPUs deúltima generación para cálculo matricial denso Exploiting the capabilities of modern GP
No full textAbstract: We present several algorithms to compute the solution of a linear system of equations on a GPU, as well as general techniques to improve their performance, such as padding and hybrid GPU-CPU computation. We compare single and double precision performance of a modern GPU with unified architecture, and show how iterative refinement with mixed precision can be used to regain full accuracy in the solution of linear systems, exploiting the potential of the processor for single precision arithmetic. Experimental results on a GTX280 using CUBLAS 2.0, the implementation of BLAS for NVIDIA R GPUs with unified architecture, illustrate the performance of the different algorithms and techniques proposed. Keywords: Graphics processors (GPUs), general purpose computing on GPU, linear algebra, BLAS, high performance. Resumen: El presente informe describe diferentes algoritmos para calcular la solución de un sistema lineal sobre una GPU, así como técnicas generales para mejorar su renidimento, como padding y técnicas híbridas CPU-GPU. Además, se hace uso de técnicas de refinamiento iterativo con precisión mixta, para conseguir mayor precisión en la solución obtenida. Se incluyen resultados experimentales sobre el procesador GTX280 para simple y doble precisión, haciendo uso de CUBLAS 2.0, la implementación de BLAS desarrollada por NVIDIA R para GPUs con arquitectura unificada. Palabras clave: Procesadores gráficos (GPUs), procesamiento de carácter general sobre GPUs,álgebra lineal, BLAS, altas prestaciones. Abstract We present several algorithms to compute the solution of a linear system of equations on a GPU, as well as general techniques to improve their performance, such as padding and hybrid GPU-CPU computation. We compare single and double precision performance of a modern GPU with unified architecture, and show how iterative refinement with mixed precision can be used to regain full accuracy in the solution of linear systems, exploiting the potential of the processor for single precision arithmetic. Experimental results on a GTX280 using CUBLAS 2.0, the implementation of BLAS for NVIDIA R GPUs with unified architecture, illustrate the performance of the different algorithms and techniques proposed
Application of Optimal Control Theory to a Batch Crystallizer using Orbital Flatness
No full textAbstract In this contribution we suggest an efficient application of the Pontryagin Minimum Principle (PMP) to the solution of an optimal control problem for a standard moment model of a batch crystallizer. The application is based on a time scaling that turns the moment model differentially flat, and on a simplifying assumption. We present efficient solutions and also consider uniqueness of the solutions. We demonstrate ways to assess and partially correct the error due to simplification. Finally, we use a case study to demonstrate the optimal trajectories, numerical effort and the magnitude of the error
Content Management Implemented as Shared Service: A Public Sector Case Study
No full textAbstract. Sharing services has gained the interest of governments to reduce costs. The basic idea is that services provided by one department can be provided to others with relatively few efforts. A new emerging trend is the implementation of content management (CM) shared services. As a new phenomenon, there is little understanding of this concept. This paper addresses this knowledge gap by investigating a public sector case study and analyzing the decision process concerning the introduction of a CM for Shared Service Centers (SSCs). The case is analyzed using a decision framework based on a multi-theory approach found in outsourcing literature. The differences with other types of SSCs are highlighted. The complexity of this arrangement originates from the need to manage content in the many parts of the organization and the involvement of many different roles. A CM SSC requires a holistic decision-making approach by balancing the management, technology and content dimensions carefully, as these dimensions influence the resulting arrangement and potential benefits
Measuring Distance with Mobile Phones Using Single-Camera Stereo Vision
No full textAbstract-Computer stereo vision is an important technique for robotic navigation and other mobile scenarios where depth perception is needed, but it usually requires two cameras with a known horizontal displacement. In this paper, we present a solution for mobile devices with just one camera, which is a first step towards making computer stereo vision available to a wide range of devices that are not equipped with stereo cameras. We have built a prototype using a state-of-the-art mobile phone, which has to be manually displaced in order to record images from different lines of sight. Since the displacement between the two images is not known in advance, it is measured using the phone's inertial sensors. We evaluated the accuracy of our single-camera approach by performing distance calculations to everyday objects in different indoor and outdoor scenarios, and compared the results with that of a stereo camera phone. As a main advantage of a single moving camera is the possibility to vary its relative position between taking the two pictures, we investigated the effect of different camera displacements on the accuracy of distance measurements
Migrating from Closed Source Software to OSS -Security Issues and Challenges
No full textAbstract Open source software (OSS
An Experimental Investigation of Effects of Specimen Size Parameters on Compressive and Tensile Properties in a Closed Cell Al Foam * 1
No full textEffects of specimen size parameters, i.e. the specimen volume and the ratio of cross sectional dimension to cell size, on compressive and tensile properties were investigated in a closed cell Al foam. For compressive tests, the stress in a plateau region decreased with decreasing specimen volume. This is likely because constraint of cell walls at the free surface is reduced with decreasing specimen volume. The scatter of the stress was large for the small volume specimens. The critical ratio of cross sectional dimension to cell size was 4 for negligible scatter of the stress. For tensile tests, the ultimate tensile strength decreased, the elongation to failure increased and the work hardening coefficient decreased with decreasing specimen volume. It is suggested that the reduced constraint of cell walls at the free surface by decreasing specimen volume affects the tensile strength as well as the compressive strength
Enhanced Power Saving Through Increasing Unavailability Interval in the IEEE 802.16e Systems
Get PDFAbstract-The power saving mechanisms of the current IEEE 802.16e system are designed to take no consideration of the harmonization between power saving classes (PSCs) I and II. Using the IEEE 802.16e standard as a basis, we propose a dynamic power saving mechanism that increases unavailability interval when a mobile station (MS) uses PSCs. The proposed mechanism adjusts the timing of the sleep window of PSC I to maximize the unavailability interval of the MSs. As a result, the proposed scheme achieves power saving of the MSs. Through numerical analysis and simulations, we show that the proposed mechanism can reduce the power consumption of the MSs considerably compared with conventional mechanisms. Index Terms-IEEE 802.16e, wireless MAN, MAC, power saving, sleep mode
Measuring Temporal Lags in Delay-Tolerant Networks
Get PDFAbstract-Delay-tolerant networks (DTNs) are characterized by a possible absence of end-to-end communication routes at any instant. In most cases, however, a form of connectivity can be established over time and space. This particularity leads to consider the relevance of a given route not only in terms of hops (topological length), but also in terms of time (temporal length). The problem of measuring temporal distances between individuals in a social network was recently addressed, based on a posteriori analysis of interaction traces. This paper focuses on the distributed version of this problem, asking whether every node in a network can know precisely and in real time how out-ofdate it is with respect to every other. Answering affirmatively is simple when contacts between the nodes are punctual, using the temporal adaptation of vector clocks provided i
