Dissertations / Theses on the topic 'Dynamická SIMS'

Hmotnostná spektrometria sekundárnych iónov s analýzou doby letu (TOF-SIMS) patrí vďaka vysokej citlivosti na prvkové zloženie medzi významné metódy analýzy pevných povrchov. Táto práca demonštruje možnosti TOF-SIMS v troch odlišných oblastiach výskumu. Prvá časť práce sa zaoberá lokalizáciou defektov vysokonapäťových polovodičových súčiastok, ktorá je nevyhnutná k ich ďalšiemu skúmaniu metódou TOF-SIMS. Bola navrhnutá experimentálna zostava s riadiacim softvérom umožňujúca automatizované meranie záverného prúdu v rôznych miestach polovodičový súčiastok. Druhá časť práce sa zaoberá kvantifikáciou koncentrácie Mg dopantov v rôznych hĺbkach vzoriek AlGaN. Kvantifikácia je založená na metóde RSF a umožňuje charakterizáciu AlGaN heteroštruktúr určených na výrobu tranzistorov s vysokou elektrónovou mobilitou (HEMT) alebo na výrobu rôznych optoelektronických zariadení. Sada 12 AlGaN kalibračných vzoriek dopovaných Mg, určených na kvantifikáciu hĺbkových profilov, bola pripravená metódou iónovej implantácie. Posledná časť práce demonštruje možnosti metódy TOF-SIMS vo výskume heterogénnej katalýzy. Hlavným objektom nášho výskumu je dynamika oxidácie CO na oxid uhličitý na polykryštalickom povrchu platiny za tlakov vysokého vákua. V tejto práci prezentujem prvé TOF-SIMS pozorovanie časopriestorových vzorov v reálnom čase, ktoré vznikajú v dôsledku rôzneho pokrytia povrchu Pt reaktantmi. Výsledky TOF-SIMS experimentu boli porovnané s výsledkami podobného experiment v rastrovacom elektrónovom mikroskope (SEM).

The aim of this work is to connect dynamic identity management system developed under the project Sec6Net with a control of SDN network. The controller Pyretic is used for network control, which allows application development by using the match-action rules. Interface between the identity management system and controller Pyretic is designed and implemented in both systems. To prove the concept, selected use cases related to security, routing and accounting are created. The use cases are implemented as applications for Pyretic controller. All programs were tested in networking laboratory according to the possibilities. The main contribution of this work is to simplify and improve the management of computer networks while providing new capabilities to administrators of these networks and ultimately their users.

Au cours d’un traitement de surface de type dépôt assisté par plasma, les caractéristiques et propriétés de l’interface entre le dépôt et le substrat sont déterminées par la première couche atomique du dépôt, voire les premiers atomes qui commencent à recouvrir la surface du substrat. Aussi, la parfaite connaissance du comportement des particules incidentes et du réarrangement des atomes suite à l’impact d’une particule du plasma est-elle un élément essentiel à la description du comportement de la surface en cours de traitement et donc de ses propriétés ultérieures. Au cours de cette thèse, nous avons entrepris d’étudier, par une approche combinant expériences et simulation numérique par dynamique moléculaire, l’interaction d’espèces (C, Ti, W) avec une surface de silicium en fonction de paramètres tels que l’énergie, la fluence ou encore l’incidence des particules arrivant sur la surface. Une part importante de ce travail a consisté à adapter les codes de dynamique moléculaire (utilisation des champs de force réactifs) aux systèmes étudiés. La partie expérimentale a nécessité la mise en place de procédures spécifiques pour l’utilisation de l’équipement Storing Matter. Les résultats montrent que, quelles que soient l’espèce incidente, parmi celles étudiées, le coefficient de collage (SC) est dans la gamme [0.7 – 1] ; dans le cas de W, quasiment tous les atomes incidents restent sur la surface (SC~~1). Outre la détermination du coefficient de collage, pour différentes conditions initiales des espèces incidentes (énergie, incidence, fluence) les modifications apportées à la surface ont également été déterminées en termes d’implantation et de trajectoire dans le matériau des espèces incidentes, et de pulvérisation de la surface du substrat<br>During plasma assisted deposition, properties of the coating substrate interface depend on the first atomic layer of the deposit, or the atoms that first start to cover the surface. Therefore the good knowledge of the sticking coefficient and the reorganization of the surface following particle impact is an essential issue to achieve the description of the behavior of the processed surface and, therefore, its expected properties. Consequently, we investigated the interaction between incoming particles (C, Ti, W) and a silicon surface by using an approach combining molecular dynamic simulations and experiments. Various initial conditions were studied, energy, fluence and incidence angle of the incoming particles. An important part of this work has consisted in adapting the molecular dynamic codes (using reactive force fields) to the investigated systems. Meanwhile, experimental procedure specifically devoted to the use of the Storing Matter facility was also developed. Results show that the sticking coefficient (SC) value is in the range [0.7 – 1] irrespectively of the incoming species; in the case of W, almost all atoms stick on the surface (SC~~1). Besides the determination of sticking coefficient, the surface modification resulting from the particles impingement were determined for various initial conditions (energy, fluence, angle) in terms of implantation and displacement of the incoming species, and surface sputtering as well

De nombreux microprocesseurs modernes mettent en œuvre le multi-threading simultané (SMT) pour améliorer l'efficacité globale des processeurs superscalaires. SMT masque les opérations à longue latence en exécutant les instructions de plusieurs threads simultanément. Lorsque les threads exécutent le même programme (cas des applications SPMD), les mêmes instructions sont souvent exécutées avec des entrées différentes. Les architectures SMT traditionnelles exploitent le parallélisme entre threads, ainsi que du parallélisme de données explicite au travers d'unités d'exécution SIMD. L'exécution SIMD est efficace en énergie car le nombre total d'instructions nécessaire pour exécuter un programme est significativement réduit. Cette réduction du nombre d'instructions est fonction de la largeur des unités SIMD et de l'efficacité de la vectorisation. L'efficacité de la vectorisation est cependant souvent limitée en pratique. Dans cette thèse, nous proposons l'architecture de vectorisation dynamique inter-thread (DITVA) pour tirer parti du parallélisme de données implicite des applications SPMD en assemblant dynamiquement des instructions vectorielles à l'exécution. DITVA augmente un processeur à exécution dans l'ordre doté d'unités SIMD en lui ajoutant un mode d'exécution vectorisant entre threads. Lorsque les threads exécutent les mêmes instructions simultanément, DITVA vectorise dynamiquement ces instructions pour assembler des instructions SIMD entre threads. Les threads synchronisés sur le même chemin d'exécution partagent le même flot d'instructions. Pour conserver du parallélisme de threads, DITVA groupe de manière statique les threads en warps ordonnancés indépendamment. DITVA tire parti des unités SIMD existantes et maintient la compatibilité binaire avec les architectures CPU existantes<br>Many modern microprocessors implement Simultaneous Multi-Threading (SMT) to improve the overall efficiency of superscalar CPU. SMT hides long latency operations by executing instructions from multiple threads simultaneously. SMT may execute threads of different processes, threads of the same processes or any combination of them. When the threads are from the same process, they often execute the same instructions with different data most of the time, especially in the case of Single-Program Multiple Data (SPMD) applications.Traditional SMT architecture exploit thread-level parallelism and with the use of SIMD execution units, they also support explicit data-level parallelism. SIMD execution is power efficient as the total number of instructions required to execute a complete program is significantly reduced. This instruction reduction is a factor of the width of SIMD execution units and the vectorization efficiency. Static vectorization efficiency depends on the programmer skill and the compiler. Often, the programs are not optimized for vectorization and hence it results in inefficient static vectorization by the compiler.In this thesis, we propose the Dynamic Inter-Thread vectorization Architecture (DITVA) to leverage the implicit data-level parallelism in SPMD applications by assembling dynamic vector instructions at runtime. DITVA optimizes an SIMD-enabled in-order SMT processor with inter-thread vectorization execution mode. When the threads are running in lockstep, similar instructions across threads are dynamically vectorized to form a SIMD instruction. The threads in the convergent paths share an instruction stream. When all the threads are in the convergent path, there is only a single stream of instructions. To optimize the performance in such cases, DITVA statically groups threads into fixed-size independently scheduled warps. DITVA leverages existing SIMD units and maintains binary compatibility with existing CPU architectures

Dans le combustible nucléaire UO2, utilisé dans les réacteurs à eau pressurisée (REP), le Cs, élément volatil compte parmi les produits de fission (PF) les plus abondamment produits. De plus, l’isotope 137Cs est connu pour être particulièrement radiotoxique. En cas d’accident, le relâchement de cet isotope est donc problématique et son étude est cruciale pour la sûreté nucléaire. En France, l’IRSN (Institut de Radioprotection et de sureté nucléaire) développe des codes de prédictions du relâchement des PF depuis le combustible, tels que MFPR (Module for Fission Product Release). Ces codes nécessitent d’être alimentés par des données fondamentales sur le comportement des PF. Ainsi, la connaissance des coefficients de diffusion de ces éléments dans la matrice combustible en fonction de la température et de l’atmosphère (pouvant oxyder le combustible en UO2+x) est primordiale. Dans ce contexte, l’objectif de cette thèse, menée en collaboration avec l’IRSN, est d’étudier la migration du Cs dans le dioxyde d’uranium stœchiométrique et sur-stœchiométrique, en conditions représentatives d’un fonctionnement normal et accidentel d’un REP, avec et sans la présence de Mo. Ce dernier est un PF abondamment produit qui agit comme tampon d’oxydation du combustible et est capable d’avoir des interactions chimiques avec le césium. De telles interactions pourraient affecter le comportement du Cs, et donc son relâchement depuis le combustible. Il a donc été nécessaire d’envisager les éventuelles interactions entre le Cs et le Mo dans le cadre de notre étude. La démarche expérimentale a consisté à simuler la présence de Cs et/ou Mo dans des pastilles d’UO2 ou d’UO2+x. par implantations ioniques des isotopes stables 133Cs et/ou 95Mo. Des recuits à haute température (950-1600°C) sous atmosphère contrôlée ou des irradiations en régime électronique couplées en température ont ensuite été réalisés, permettant d’induire la migration du Cs et du Mo. La spectrométrie de masse à ionisation secondaire (SIMS) a été utilisée pour suivre l’évolution des profils de concentration des éléments implantés, permettant d’extraire les coefficients de diffusion apparents du Cs dans UO2 et UO2+x en fonction des différents traitements. Une étude complémentaire de la microstructure a été réalisée par spectroscopie Raman et microscopie électronique en transmission (MET). Le Cs est très mobile dans UO2 sous atmosphère réductrice même si une partie et piégée sous forme de bulles à faible profondeur. Nous avons mis en évidence que la présence de Mo diminuait fortement cette mobilité. La même tendance est observée dans UO2+x sous atmosphère oxydante. Néanmoins les mécanismes d’immobilisation du Cs par le Mo diffèrent selon les conditions redox de recuit. En atmosphère réductrice, les expériences MET ont montré la formation de paires bulles de Cs-précipités métalliques de Mo dans les échantillons co-implantés. En atmosphère oxydante, l’absence de mobilité du Cs pourrait être liée à l’oxydation du Mo rendant possible des interactions chimiques Cs-Mo. Pour la première fois, des potentiels semi-empiriques ont été utilisés pour réaliser des calculs de dynamique moléculaire sur la diffusion du Cs et du Mo dans UO2 et UO2+x. Ces calculs nous ont aussi permis de caractériser les mécanismes de diffusion de l’oxygène dans ces matériaux en présence de ces deux PF<br>In the nuclear fuel UO2, which is widely used in Pressurized Water Reactor (PWR), Cs is a volatile element and is one of the most abundant fission product (FP). Furthermore, 137Cs is known to be highly radiotoxic. During a hypothetical accident, release of Cs would be particularly problematic for the environment. Hence, study of this element is of major concern for nuclear safety. To assess this issue, the French nuclear safety institute (IRSN) develops codes to predict FP release from nuclear fuel in normal and accidental conditions. This code requires fundamental data on FP behavior such as diffusion coefficient of these elements in UO2 as a function of temperature and atmosphere conditions (leading to UO2+x formation in oxidative conditions). The aim of this PhD, supported by the IRSN, is to study Cs migration in stoichiometric and hyper-stoichiometric uranium dioxide with and without the presence of Mo, in normal and accidental conditions of a PWR. This latter element is also an abundant FP, which is important to consider because it acts as an oxygen buffer in the fuel and may interact chemically with Cs. Such interactions may affect Cs behavior, hence its release from the fuel. Therefore, Cs-Mo interactions are considered in our study. The experimental procedure consists in simulating the Cs and/or Mo presence in UO2 and UO2+x pellets by ion implantation of stable isotopes 133Cs and/or 95Mo. Then, high temperature annealing (950 °C - 1600 °C) under controlled atmosphere or electronic excitations induced by irradiation coupled with temperature are performed to induce Cs and Mo migration. Secondary Ion Mass Spectrometry (SIMS) is used to follow the concentration profile evolution of these elements, allowing extracting effective diffusion coefficients of Cs in UO2 and UO2+x as a function of irradiation or thermal treatment. Microstructure characterizations were made by Raman spectroscopy and transmission electron microscopy (TEM). We show that Cs is mobile in UO2 under reducing atmosphere, even though some of the Cs is trapped in Cs-bubbles located near the surface. We evidence that Mo presence prevents Cs to be mobile. The same tendency is observed in UO2+x under oxidizing atmosphere. Nevertheless, Cs immobilization mechanisms in presence of Mo vary upon redox conditions used during annealing. In reducing conditions, TEM experiments showed formation of Cs bubbles associated with Mo metallic precipitates in co-implanted samples. In oxidative conditions, absence of Cs mobility could be explained by Mo oxidation leading to possible Cs-Mo chemical interactions. For the first time, semi-empirical potentials were used to perform molecular dynamic (MD) calculations on Cs and Mo diffusion in UO2 and UO2+x. These simulations also allowed characterizing oxygen diffusion mechanisms in these matrixes in presence of Cs and Mo

Recent advances in graphics processing units (GPUs) have resulted in massively parallel hardware that is easily programmable and widely available in commodity desktop computer systems. GPUs typically use single-instruction, multiple-data (SIMD) pipelines to achieve high performance with minimal overhead for control hardware. Scalar threads running the same computing kernel are grouped together into SIMD batches, sometimes referred to as warps. While SIMD is ideally suited for simple programs, recent GPUs include control flow instructions in the GPU instruction set architecture and programs using these instructions may experience reduced performance due to the way branch execution is supported by hardware. One solution is to add a stack to allow different SIMD processing elements to execute distinct program paths after a branch instruction. The occurrence of diverging branch outcomes for different processing elements significantly degrades performance using this approach. In this thesis, we propose dynamic warp formation and scheduling, a mechanism for more efficient SIMD branch execution on GPUs. It dynamically regroups threads into new warps on the fly following the occurrence of diverging branch outcomes. We show that a realistic hardware implementation of this mechanism improves performance by an average of 47% for an estimated area increase of 8%.

The waves traveling on the surface of the world’s oceans carry atremendous amount of energy. The ability to convert this energy forhuman use has the potential to help solves the worlds energy problem. Adirect-drive linear generator point absorber is a wave energy converter(WEC) that aims to reduce the complexity of the overall system andshelter the most vulnerable parts of the system by placing them on theseabed. This concept builds around the buoy moving up and down indifferent sea-states which leads to a correlating vertical movement of thestroke in the generator resulting in the conversion of mechanical energyto power. This report aims to explore the possibility to use the open codeWEC-Sim to model the Uppsala University direct-drive linear generatorWEC in extreme sea states and to identify the resulting extreme loads. Theconstructed WEC-Sim model constrained the buoys motion in heave andsurge and limited its range of motion by modeling the generators upperend-stop spring. Simulations were run for different sea-states and theresulting forces on the system were analyzed. The peak line force for thedifferent sea states was calculated and compared to previous studies. Theresults validated the model as they showed a good correlation for mostsea-states. It was only for larger significant wave heights that there was adivergence compared to the results in previous studies.

Trends in computer engineering place renewed emphasis on increasing parallelism and heterogeneity. The rise of parallelism adds an additional dimension to the challenge of portability, as different processors support different notions of parallelism, whether vector parallelism executing in a few threads on multicore CPUs or large-scale thread hierarchies on GPUs. Thus, software experiences obstacles to portability and efficient execution beyond differences in instruction sets; rather, the underlying execution models of radically different architectures may not be compatible. Dynamic compilation applied to data-parallel heterogeneous architectures presents an abstraction layer decoupling program representations from optimized binaries, thus enabling portability without encumbering performance. This dissertation proposes several techniques that extend dynamic compilation to data-parallel execution models. These contributions include: - characterization of data-parallel workloads - machine-independent application metrics - framework for performance modeling and prediction - execution model translation for vector processors - region-based compilation and scheduling We evaluate these claims via the development of a novel dynamic compilation framework, GPU Ocelot, with which we execute real-world workloads from GPU computing. This enables the execution of GPU computing workloads to run efficiently on multicore CPUs, GPUs, and a functional simulator. We show data-parallel workloads exhibit performance scaling, take advantage of vector instruction set extensions, and effectively exploit data locality via scheduling which attempts to maximize control locality.

Heterogeneous parallel architectures like those comprised of CPUs and GPUs are a tantalizing compute fabric for performance-hungry developers. While these platforms enable order-of-magnitude performance increases for many data-parallel application domains, there remain several open challenges: (i) the distinct execution models inherent in the heterogeneous devices present on such platforms drives the need to dynamically match workload characteristics to the underlying resources, (ii) the complex architecture and programming models of such systems require substantial application knowledge and effort-intensive program tuning to achieve high performance, and (iii) as such platforms become prevalent, there is a need to extend their utility from running known regular data-parallel applications to the broader set of input-dependent, irregular applications common in enterprise settings. The key contribution of our research is to enable runtime specialization on such hybrid CPU-GPU platforms by matching application characteristics to the underlying heterogeneous resources for both regular and irregular workloads. Our approach enables profile-driven resource management and optimizations for such platforms, providing high application performance and system throughput. Towards this end, this research: (a) enables dynamic instrumentation for GPU-based parallel architectures, specifically targeting the complex Single-Instruction Multiple-Data (SIMD) execution model, to gain real-time introspection into application behavior; (b) leverages such dynamic performance data to support novel online resource management methods that improve application performance and system throughput, particularly for irregular, input-dependent applications; (c) automates some of the programmer effort required to exercise specialized architectural features of such platforms via instrumentation-driven dynamic code optimizations; and (d) proposes a specialized, affinity-aware work-stealing scheduling runtime for integrated CPU-GPU processors that efficiently distributes work across all CPU and GPU cores for improved load balance, taking into account both application characteristics and architectural differences of the underlying devices.

In this thesis, we present our approach to add dynamism support to simulation environments, which adopts DEVS-based modeling and simulation approach and builds upon previous work on SiMA, a DEVS-based simulation framework developed at TUBITAK UEKAE. Defining and executing simulation models of complex and adaptive systems is often a non-trivial task. One of the requirements of simulation software frameworks for such complex and adaptive systems is that supporting variable structure models, which can change their behavior and structure according to the changing conditions. In the relevant literature there are already proposed solutions to the dynamism support problem. One particular contribution offered in this study over previous approaches is the systematic and automatic framework support for poststructural-change state synchronization among models with related couplings, in a way that benefits from the strongly-typed execution environment SiMA provides. In this study, in addition to introducing theoretical extensions to classic SiMA, performance comparisons of dynamic version with classic version over a sample Wireless Sensor Network simulation is provided and possible effects of dynamism extensions to the performance are discussed.

Dans l'industrie automobile, les réductions de masse permettent des économies de matières premières et des gains importants en performance. Cet allégement ne peut cependant pas se faire au détriment des exigences en matière de confort et de sécurité. Dans ce contexte, l'objectif de l'optimisation topologique est de déterminer, en amont des projets, les caractéristiques générales des pièces mécaniques. Les logiciels commerciaux actuels ne permettant pas l'intégration de toutes les contraintes déclinées sur les projets véhicules, notamment les contraintes vibro-accoustiques, nous proposons dans cette thèse une méthodologie et une solution logicielle associée, permettant la prise en compte d'un cahier des charges conforme à ceux utilisés chez Renault. Nous combinons pour cela la méthode d'optimisation topologique SIMP et l'algorithme d'optimisation numérique FSQP. Après avoir évalué notre méthode sur des cas de géométrie simple, nous montrons son potentiel sur le cas d'une face accessoires.

The key objective of this thesis is the outline of an alternative capital market modeling framework, the Volatility Asset Pricing Model, VAPM, inspired by the innovative dual approach of Mandelbrot and Hudson using the method based on synthesis of two seemingly antagonistic factors -- the volatility of market prices and their serial dependence determining the capital markets' dynamics. The pilot tests of this model in various periods using the market index as well as a portfolio of selected securities delivered generally satisfactory results. Firstly, the work delivers a brief recapitulation regarding the concepts of a consumer/investor choice under general conditions of hypothetical certainty. Secondly, this outline is then followed by a description of the "classical" methodologies in the risky environment of uncertainty, with assessment of their corresponding key models, i.e. the CAPM, SIM, MIM, APTM, etc., notwithstanding results of the related testing approaches. Thirdly, this assessment is based on evaluation of the underlying doctrine of Efficient Market Hypothesis in relation to the so called Random Walk Model. Fourthly, in this context the work also offers a brief exposure to a few selected tests of these contraversial concepts. Fifthly, the main points of conteporary approaches such as the Fractal Dimension and the Hurst Exponent in the dynamic framework of information entropy are subsequently described as the theoretical tools leading to development of the abovementioned model VAPM. The major contribution of this thesis is considered its attempt to apply the abovementioned concepts in practice, with the intention to possibly inspire a further analytical research.

In this dissertation, we discuss the Distributed Spacecraft Attitude Control System Simulator (DSACSS) testbed developed at Virginia Polytechnic Institute and State University for the purpose of investigating various control techniques for single and multiple spacecraft. DSACSS is comprised of two independent hardware-in-the-loop simulators and one software spacecraft simulator. The two hardware-in-the-loop spacecraft simulators have similar subsystems as flight-ready spacecraft (e.g. command and data handling; communications; attitude determination and control; power; payload; and guidance and navigation). The DSACSS framework is a flexible testbed for investigating a variety of spacecraft control techniques, especially control scenarios involving coupled attitude and orbital motion. The attitude hardware simulators along with numerical simulations assist in the development and evaluation of Lyapunov based asymptotically stable, nonlinear attitude controllers with three reaction wheels as the control device. The angular rate controller successfully tracks a time varying attitude trajectory. The Modified Rodrigues Parmater (MRP) attitude controller results in successfully tracking the angular rates and MRP attitude vector for a time-varying attitude trajectory. The attitude controllers successfully track the reference attitude in real-time with hardware similar to flight-ready spacecraft. Numerical simulations and the attitude hardware simulators assist in the development and evaluation of a robust, asymptotically stable, nonlinear attitude controller with three reaction wheels as the actuator for attitude control. The MRPs are chosen to represent the attitude in the development of the controller. The robust spacecraft attitude controller successfully tracks a time-varying reference attitude trajectory while bounding system uncertainties. The results of a Global Positioning System (GPS) hardware-in-the-loop simulation of two spacecraft flying in formation are presented. The simulations involve a chief spacecraft in a low Earth orbit (LEO), while a deputy spacecraft maintains an orbit position relative to the chief spacecraft. In order to maintain the formation an orbit correction maneuver (OCM) for the deputy spacecraft is required. The control of the OCM is accomplished using a classical orbital element (COE) feedback controller and simulating continual impulsive thrusting for the deputy spacecraft. The COE controller requires the relative position of the six orbital elements. The deputy communicates with the chief spacecraft to obtain the current orbit position of the chief spacecraft, which is determined by a numerical orbit propagator. The position of the deputy spacecraft is determined from a GPS receiver that is connected to a GPS hardware-in-the-loop simulator. The GPS simulator creates a radio frequency (RF) signal based on a simulated trajectory, which results in the GPS receiver calculating the navigation solution for the simulated trajectory. From the relative positions of the spacecraft the COE controller calculates the OCM for the deputy spacecraft. The formation flying simulation successfully demonstrates the closed-loop hardware-in-the-loop GPS simulator. This dissertation focuses on the development of the DSACSS facility including the development and implementation of a closed-loop GPS simulator and evaluation of nonlinear feedback attitude and orbit control laws using real-time hardware-in-the-loop simulators.<br>Ph. D.

La distribution de deformation a l'interface entre une couche epitaxique d'un compose tertiaire (arseniure de al et ga par exemple) et un support semiconducteur iii-v a pu etre mise en evidence et mesuree a l'aide, essentiellement, de la diffraction d'une onde rx plane ou pseudo-plane. Etude de la sensibilite de la methode a un etalement de l'interface en fonction de l'epaisseur de la couche et de son desaccord avec le support. Simulation sur ordinateur du profil de reflexion des jonctions abruptes et etalees. Application a divers echantillons et au cas des structures multicouches et des superreseaux

碩士<br>國立成功大學<br>電腦與通信工程研究所<br>103<br>SIMD divergence is one of the critical causes that decrease the parallel computing efficiency in contemporary GPGPU (General Purpose Graphic Processor Unit) architecture. In this thesis, we evaluate a cycle accurate GPU simulator platform based on HSAIL under OpenCL framework by offloading the kernel programs into simulator. A wavefront (“wavefront” and “warp” in AMD and NVIDIA terminology respectively) is the gathering of multiple threads that execute the same instruction in SIMD fashion. When a wavefront or a warp executes a conditional branch instruction, threads in the warp may go to distinct PCs if the threads have different branch targets, and it’s called SIMD control divergence. Re-convergence mechanisms are applied to help divergent wavefront to execute instructions properly. We develop a new dynamic stack-based re-convergence scheme that can be implemented with or without finalizer generated re-convergence instructions. Using the scheme we propose, the divergent warp re-converges dynamically and get a 13.36% activity factor improvement on average from opportunistic early re-convergence in the unstructured control flow, and the performance is better in the way that warp re-convergence without finalier generated hint instructions.

碩士<br>國立交通大學<br>資訊科學與工程研究所<br>102<br>Modern processors are increasingly enhanced with SIMD instructions. For examples, the MMX, SSE, and AVX instructions in the x86 architecture, and the Neon instruction set in the ARM architecture are all SIMD instructions. Using these SIMD instructions could significantly increase the performance of applications, hence application binaries are likely to have a greater fraction of instructions that are SIMD instructions. However, SIMD instruction translation has not attacked much attention in Dynamic Binary Translation (DBT). For example, in the popular QEMU system emulator, guest SIMD instructions are often emulated with a sequence of scalar instructions even when the host machines do have SIMD instructions to support such parallel computation, leaving a large potential for performance enhancement. In this thesis, we propose two approaches, one to leverage the existing helper function implementation in QEMU, and the other to use a newly introduced vector IR (Intermediate Representation) to enhance the performance of SIMD instructions translation in DBT of QEMU. The two approaches have been implemented in the QEMU with ARM frontend and x86-64 backend. In our experiment, the vector IR QEMU is 1.01 to 5.55 times faster than original QEMU with benchmark SPEC2006 CFP and 7.61 times faster than original QEMU with benchmark Linpack.

碩士<br>國立臺灣大學<br>資訊工程學研究所<br>105<br>Processor manufacturers have embraced single instruction multiple data (SIMD) for decades because of its superior performance and power efficiency. The configurations of SIMD registers (i.e., the number and width) have evolved and diverged rapidly through various ISA extensions on different architectures. However, migrating legacy or proprietary applications optimized for one guest ISA to another host ISA that has fewer but longer SIMD registers through binary translation raises the issues of asymmetric SIMD register configurations. To date, these issues have been overlooked. As a result, only a small fraction of the potential performance gain is realized due to underutilization of the host&apos;&apos;s SIMD parallelism and register capacity. In this paper, we present a novel dynamic binary translation technique called spill-aware SLP (saSLP), which transforms binary loops optimized for a guest ISA to exploit longer host registers in terms of both data parallelism and register capacity. Proposed saSLP combines short guest SIMD instructions and registers to fully utilize the host&apos;&apos;s parallelism as well as minimize register spilling. Experiment results show that saSLP improves the performance by 1.6X (2.3X) across a number of benchmarks, and reduces spilling by 97% (99%) for ARMv8 NEON to x86 AVX2 (AVX-512) translation.

The study of randomly excited nonlinear dynamical systems forms the focus of this thesis. We discuss two classes of problems: first, the characterization of nonlinear random response of the system before it comes into existence and, the second, assimilation of measured responses into the mathematical model of the system after the system comes into existence. The first class of problems constitutes forward problems while the latter belongs to the class of inverse problems. An outstanding feature of these problems is that they are almost always not amenable for exact solutions. We tackle in the present study these two classes of problems using Monte Carlo simulation tools in conjunction with Markov process theory, Bayesian model updating strategies, and particle filtering based dynamic state estimation methods. It is well recognized in literature that any successful application of Monte Carlo simulation methods to practical problems requires the simulation methods to be reinforced with effective means of controlling sampling variance. This can be achieved by incorporating any problem specific qualitative and (or) quantitative information that one might have about system behavior in formulating estimators for response quantities of interest. In the present thesis we outline two such approaches for variance reduction. The first of these approaches employs a substructuring scheme, which partitions the system states into two sets such that the probability distribution of the states in one of the sets conditioned on the other set become amenable for exact analytical solution. In the second approach, results from data based asymptotic extreme value analysis are employed to tackle problems of time variant reliability analysis and updating of this reliability. We exemplify in this thesis the proposed approaches for response estimation and model updating by considering wide ranging problems of interest in structural engineering, namely, nonlinear response and reliability analyses under stationary and (or) nonstationary random excitations, response sensitivity model updating, force identification, residual displacement analysis in instrumented inelastic structures under transient excitations, problems of dynamic state estimation in systems with local nonlinearities, and time variant reliability analysis and reliability model updating. We have organized the thesis into eight chapters and three appendices. A resume of contents of these chapters and appendices follows. In the first chapter we aim to provide an overview of mathematical tools which form the basis for investigations reported in the thesis. The starting point of the study is taken to be a set of coupled stochastic differential equations, which are obtained after discretizing spatial variables, typically, based on application of finite element methods. Accordingly, we provide a summary of the following topics: (a) Markov vector approach for characterizing time evolution of transition probability density functions, which includes the forward and backward Kolmogorov equations, (b) the equations governing the time evolution of response moments and first passage times, (c) numerical discretization of governing stochastic differential equation using Ito-Taylor’s expansion, (d) the partial differential equation governing the time evolution of transition probability density functions conditioned on measurements for the study of existing instrumented structures, (e) the time evolution of response moments conditioned on measurements based on governing equations in (d), and (f) functional recursions for evolution of multidimensional posterior probability density function and posterior filtering density function, when the time variable is also discretized. The objective of the description here is to provide an outline of the theoretical formulations within which the problems of response estimation and model updating are formulated in the subsequent chapters of the present thesis. We briefly state the class of problems, which are amenable for exact solutions. We also list in this chapter major text books, research monographs, and review papers relevant to the topics of nonlinear random vibration analysis and dynamic state estimation. In Chapter 2 we provide a review of literature on solutions of problems of response analysis and model updating in nonlinear dynamical systems. The main focus of the review is on Monte Carlo simulation based methods for tackling these problems. The review accordingly covers numerical methods for approximate solutions of Kolmogorov equations and associated moment equations, variance reduction in simulation based analysis of Markovian systems, dynamic state estimation methods based on Kalman filter and its variants, particle filtering, and variance reduction based on Rao-Blackwellization. In this review we chiefly cover papers that have contributed to the growth of the methodology. We also cover briefly, the efforts made in applying the ideas to structural engineering problems. Based on this review, we identify the problems of variance reduction using substructuring schemes and data based extreme value analysis and, their incorporation into response estimation and model updating strategies, as problems requiring further research attention. We also identify a range of problems where these tools could be applied. We consider the development of a sequential Monte Carlo scheme, which incorporates a substructuring strategy, for the analysis of nonlinear dynamical systems under random excitations in Chapter 3. The proposed substructuring ensures that a part of the system states conditioned on the remaining states becomes Gaussian distributed and is amenable for an exact analytical solution. The use of Monte Carlo simulations is subsequently limited for the analysis of the remaining system states. This clearly results in reduction in sampling variance since a part of the problem is tackled analytically in an exact manner. The successful performance of the proposed approach is illustrated by considering response analysis of a single degree of freedom nonlinear oscillator under random excitations. Arguments based on variance decomposition result and Rao-Blackwell theorems are presented to demonstrate that the proposed variance reduction indeed is effective. In Chapter 4, we modify the sequential Monte Carlo simulation strategy outlined in the preceding chapter to incorporate questions of dynamic state estimation when data on measured responses become available. Here too, the system states are partitioned into two groups such that the states in one group become Gaussian distributed when conditioned on the states in the other group. The conditioned Gaussian states are subsequently analyzed exactly using the Kalman filter and, this is interfaced with the analysis of the remaining states using sequential importance sampling based filtering strategy. The development of this combined Kalman and sequential importance sampling filtering method constitutes one of the novel elements of this study. The proposed strategy is validated by considering the problem of dynamic state estimation in linear single and multi-degree of freedom systems for which exact analytical solutions exist. In Chapter 5, we consider the application of the tools developed in Chapter 4 for a class of wide ranging problems in nonlinear random vibrations of existing systems. The nonlinear systems considered include single and multi-degree of freedom systems, systems with memoryless and hereditary nonlinearities, and stationary and nonstationary random excitations. The specific applications considered include nonlinear dynamic state estimation in systems with local nonlinearities, estimation of residual displacement in instrumented inelastic dynamical system under transient random excitations, response sensitivity model updating, and identification of transient seismic base motions based on measured responses in inelastic systems. Comparisons of solutions from the proposed substructuring scheme with corresponding results from direct application of particle filtering are made and a satisfactory mutual agreement is demonstrated. We consider next questions on time variant reliability analysis and corresponding model updating in Chapters 6 and 7, respectively. The research effort in these studies is focused on exploring the application of data based asymptotic extreme value analysis for problems on hand. Accordingly, we investigate reliability of nonlinear vibrating systems under stochastic excitations in Chapter 6 using a two-stage Monte Carlo simulation strategy. For systems with white noise excitation, the governing equations of motion are interpreted as a set of Ito stochastic differential equations. It is assumed that the probability distribution of the maximum over a specified time duration in the steady state response belongs to the basin of attraction of one of the classical asymptotic extreme value distributions. The first stage of the solution strategy consists of selection of the form of the extreme value distribution based on hypothesis testing, and, the next stage involves the estimation of parameters of the relevant extreme value distribution. Both these stages are implemented using data from limited Monte Carlo simulations of the system response. The proposed procedure is illustrated with examples of linear/nonlinear systems with single/multiple degrees of freedom driven by random excitations. The predictions from the proposed method are compared with the results from large scale Monte Carlo simulations, and also with the classical analytical results, when available, from the theory of out-crossing statistics. Applications of the proposed method for vibration data obtained from laboratory conditions are also discussed. In Chapter 7 we consider the problem of time variant reliability analysis of existing structures subjected to stationary random dynamic excitations. Here we assume that samples of dynamic response of the structure, under the action of external excitations, have been measured at a set of sparse points on the structure. The utilization of these measurements in updating reliability models, postulated prior to making any measurements, is considered. This is achieved by using dynamic state estimation methods which combine results from Markov process theory and Bayes’ theorem. The uncertainties present in measurements as well as in the postulated model for the structural behaviour are accounted for. The samples of external excitations are taken to emanate from known stochastic models and allowance is made for ability (or lack of it) to measure the applied excitations. The future reliability of the structure is modeled using expected structural response conditioned on all the measurements made. This expected response is shown to have a time varying mean and a random component that can be treated as being weakly stationary. For linear systems, an approximate analytical solution for the problem of reliability model updating is obtained by combining theories of discrete Kalman filter and level crossing statistics. For the case of nonlinear systems, the problem is tackled by combining particle filtering strategies with data based extreme value analysis. The possibility of using conditional simulation strategies, when applied external actions are measured, is also considered. The proposed procedures are exemplified by considering the reliability analysis of a few low dimensional dynamical systems based on synthetically generated measurement data. The performance of the procedures developed is also assessed based on limited amount of pertinent Monte Carlo simulations. A summary of the contributions made and a few suggestions for future work are presented in Chapter 8. The thesis also contains three appendices. Appendix A provides details of the order 1.5 strong Taylor scheme that is extensively employed at several places in the thesis. The formulary pertaining to the bootstrap and sequential importance sampling particle filters is provided in Appendix B. Some of the results on characterizing conditional probability density functions that have been used in the development of the combined Kalman and sequential importance sampling filter in Chapter 4 are elaborated in Appendix C.

碩士<br>國立臺灣大學<br>材料科學與工程學研究所<br>99<br>Using pulsed primary cluster ions, especially for C60+ cluster ion, time-of-flight secondary ion mass spectrometry (ToF-SIMS) has been shown to be a promising technique for analyzing biological specimens. With molecular secondary ions of high mass, multiple molecules can be identified at the same time without prior separation or isotope labeling. While current reports are based on static-SIMS that makes depth profile more complicated, a dynamic-SIMS based technique is reported in this work. Mixed trehalose and peptides were used as a model for evaluating the parameters that lead to parallel detection and quantification of biomaterials. Trehalose is mixed with different peptides separately with varied concentrations of peptides. It is found that the normalized secondary ion intensity of peptide as respect to trehalose is direct proportional to its concentration in the matrix. Therefore, by plotting the percentages of peptides exist in trehalose versus their normalized SIMS intensities, calibration curves of each peptide are obtained. Using these curves, it is shown that parallel detection, identification, and quantification of multiple peptides in the matrix can be achieved. To suppress the associated carbon deposition with high energy C60+ bombardment that leads to suppressed ion intensity in prolonged profiling, a low energy Ar+ is used to co-sputter the peptide-doped trehalose thin film. It is shown that the co-sputtering technique yields more steady molecular ion intensity than single C60+ beam. In other words, the co-sputtering is more suitable for analyzing thick specimens. Furthermore, because the C60+ is responsible for generating the molecular ions, it is found that the does of the auxiliary Ar+ does not change calibration curve for quantification.

Organizations act in very dynamic and rapidly changing environments, wherein they face multiple exogenous shocks that threaten their sustainability and competitive advantage. Therefore, it is of utmost importance that firms continuously adapt and change their strategies to find a fit with the changing environment in order to guarantee survival and remain competitive. Scholars have developed management literature on the importance of strategic change on firm’s alignment with unstable and dynamic environments. Moreover, the Dynamic Capability view, which illustrates how firms can achieve sustainable competitive advantage and cope with the changing ecosystems, has gained increased awareness over the years. In order to demonstrate a real-life example of Strategic Change and the dynamic capabilities' theory in action, this dissertation examines the case of SIBS's digital transformation. The teaching case focuses on SIBS' changing process, strategic reorientation and also examines how, by introducing MB WAY, SIBS coped with the payments industry's dynamism. This case clearly illustrates the firm’s imminent need for strategic change and the importance of the company’s use of dynamic capabilities to successfully adapt to the external environment and enact superior performance.<br>As empresas atuam em ecossistemas bastante dinâmicos e em rápida mudança, onde enfrentam vários choques exógenos que tendem a ameaçar a sua sustentabilidade e vantagem competitiva. Portanto, é de extrema importância que as empresas tentem continuamente adaptar e alterar as suas estratégias de forma a encontrar um posicionamento no ecossistema em mudança, garantindo, assim, a sua sobrevivência e competitividade. Académicos têm desenvolvido, cada vez mais, literatura relativa à importância e necessidade de mudança estratégica procurando a adaptação ao ecossistema instável e dinâmico em que as empresas se encontram. Além disto, a teoria das dynamic capabilities ganhou maior preponderância ao longo dos anos. Estas duas correntes literárias vêm demonstrar que as empresas podem obter vantagem competitiva sustentável e lidar com as mudanças nos ecossistemas. Para ilustrar com um exemplo real da mudança estratégica e a teoria das dynamic capabilities, esta dissertação incide sobre a transformação digital da SIBS. O caso de estudo concentra-se no processo de mudança da SIBS, na sua reorientação estratégica e em como, ao introduzir o MB WAY, a empresa conseguiu responder às mudanças no setor de pagamentos. Este caso ilustra claramente a necessidade iminente das empresas recorrerem a mudanças estratégicas e, em particular, a importância do uso de dynamic capabilities por parte da SIBS para se adaptar com sucesso ao ambiente externo e obter um desempenho superior.

O desenvolvimento da tecnologia financeira tem tido um papel marcante na vida dos consumidores e vendedores, pois permite simplificar cada vez mais a forma como hoje em dia realizamos transações financeiras. Como exemplo, temos os meios de pagamento que cada vez mais têm sido alvo de inovações constantes, em que a facilidade com que hoje em dia se faz um pagamento, tem sido proporcionada pelo avanço da tecnologia. Entre outras vantagens para os utilizadores, temos a facilidade com que é executada a operação, custos baixos, burocracia reduzida e um maior controle sobre as operações financeiras. Nesta dissertação pretende fazer-se um estudo de caso sobre a SIBS - Sociedade Interbancária de Serviços, em que as empresas que estão incluídas neste grupo têm como objetivo procurar soluções que permitam inovar nos serviços financeiros nomeadamente na área dos pagamentos. Deste modo, o objetivo foca-se em mostrar como a SIBS é inovadora usando um conjunto de indicadores, tendo em conta as suas competências dinâmicas.<br>The development of financial technology has had a significant role in the life of consumers and sellers, because allows to simplify even more the way as we carry out financial transactions nowadays. As an example, we have the payment ways that has been increasingly targeted by constant innovations, in which the facility with which a payment is made today, has been provided by the advancement of technology. Among other advantages for users, we have the facility with which to perform an operation, low costs, reduced bureaucracy and greater control over how financial transactions. This dissertation intends to make a case study about a SIBS - Sociedade Interbancária de Serviços, in which the companies included in this group, aim to look as goal searching to solutions that allow innovating in financial services, especially in the area of payments. Thus, the goal is to show how a SIBS is innovative using a set of indicators, taking into account their dynamic competencies.

O estudo da economia da inovação tem se destacado nos últimos anos no seio das empresas de modo a que estas consigam ultrapassar os desafios constantes a que estão sujeitas. A adaptação dos serviços financeiros a esta nova realidade, constitui uma problemática que necessita de uma contínua evolução por parte empresas de modo a corresponder às preferências, cada vez mais diversificadas, e necessidades dos consumidores. Assim, é necessário que as organizações se mostrem dinâmicas e tenham políticas de inovação de modo a superarem os obstáculos existentes, proporcionando à sociedade vários serviços com novas soluções para satisfazer e simplificar os hábitos dos consumidores. Deste modo, a presente dissertação propõe-se a analisar a importância e a necessidade do dinamismo e inovação numa infraestrutura e, também, a influência da marca e patentes no reconhecimento dos serviços, sendo que se irá utilizar uma metodologia de estudo de caso da empresa SIBS. Através de relatórios da própria empresa, assim como uma análise a certos indicadores de inovação, pretende-se aprofundar este tema, no sentido de perceber qual o impacto da inovação de serviços na sociedade, revelado através do caráter inovador da SIBS a importância de um contínuo processo de inovação e preocupação constante, por parte da empresa, em propor novas soluções que colmatam as necessidades da sociedade.<br>The study of the economics of innovation has been outstanding in the past few years within companies so that they can overcome the constant challenges they face. The adaptation of financial services to this new reality is a problem which continually needs to be addressed by companies in order to respond to the increasingly diverse preferences and needs of consumers. In this sense, it’s required by the environment that the organizations portray themselves as dynamics and support innovations politics in ways to overstep the existing boundaries, providing society with numerous services that represent new solutions to satisfy and simplify the consumers habits. Thus, this thesis purposes itself to analyze the meaning and need of dynamism and innovation on an infra-structure and also the influence of the brand in the recognition of the services, building it from a case study research about SIBS enterprise. Through reports of the company itself together with the analysis of selected innovation indicators it is intended to deepen this subject in effort to understand what is the real impact of services innovation in society, revealed throw the innovative character of SIBS, the importance of a continuous innovation process and constant worry, by the company, in proposing new solutions to fulfill the society’s necessities.

Lattice-Boltzmann method(LBM), a promising new particle-based simulation technique for complex and multiscale fluid flows, has seen tremendous adoption in recent years in computational fluid dynamics. Even with a state-of-the-art LBM solver such as Palabos, a user still has to manually write his program using the library-supplied primitives. We propose an automated code generator for a class of LBM computations with the objective to achieve high performance on modern architectures. Tiling is a very important loop transformation used to improve the performance of stencil computations by exploiting locality and parallelism. In the first part of the work, we explore diamond tiling, a new tiling technique to exploit the inherent ability of most stencils to allow tile-wise concurrent start. This enables perfect load-balance during execution and reduces the frequency of synchronization required. Few studies have looked at time tiling for LBM codes. We exploit a key similarity between stencils and LBM to enable polyhedral optimizations and in turn time tiling for LBM. Besides polyhedral transformations, we also describe a number of other complementary transformations and post processing necessary to obtain good parallel and SIMD performance on modern architectures. We also characterize the performance of LBM with the Roofline performance model. Experimental results for standard LBM simulations like Lid Driven Cavity, Flow Past Cylinder, and Poiseuille Flow show that our scheme consistently outperforms Palabos–on average by3 x while running on 16 cores of a n Intel Xeon Sandy bridge system. We also obtain a very significant improvement of 2.47 x over the native production compiler on the SPECLBM benchmark.