Efficient storage and retrieval of data is critical in today's computing environments and storage systems need to keep up with the pace of evolution of other system components like CPU, memory etc., for building an overall efficient system. With virtualization becoming pervasive in enterprise and cloud-based infrastructures, it becomes vital to build I/O systems that better account for the changes in scenario in virtualized systems. However, the evolution of storage systems have been limited significantly due to adherence to legacy interface standards between the operating system and storage subsystem. Even though storage systems have become more powerful in the recent times hosting large processors and memory, thin interface to file system leads to wastage of vital information contained in the storage system from being used by higher layers. Virtualization compounds this problem with addition of new indirection layers that makes underlying storage systems even more opaque to the operating system. This dissertation addresses the problem of inefficient use of disk information by identifying storage-level opportunities and developing pro-active techniques to storage management. We present a new class of storage systems called pro-active storage systems (PaSS), which in addition to being compatible with existing I/O interface, exerts a limit degree of control over the file system policies by leveraging it's internal information. In this dissertation, we present our PaSS framework that includes two new I/O interfaces called push and pull, both in the context of traditional systems and virtualized systems. We demonstrate the usefulness of our PaSS framework by a series of case studies that exploit the information available in underlying storage system layer, for overall improvement in IO performance. We also built a framework to evaluate performance and energy of modern storage systems by implementing a novel I/O trace replay tool and an analytical model for measuring performance and energy of complex storage systems. We believe that our PaSS framework and the suite of evaluation tools helps in better understanding of modern storage system behavior and thereby implement efficient policies in the higher layers for better performance, data reliability and energy efficiency by making use of the new interfaces in our framework.

Alors que le rat apprend à chercher de multiples sources de nourriture ou d'eau, des processus d'apprentissage de séquences spatiales et de rejeu ont lieu dans l'hippocampe et le cortex préfrontal.Des études récentes (De Jong et al. 2011; Carr, Jadhav, and Frank 2011) mettent en évidence que la navigation spatiale dans l'hippocampe de rat implique le rejeu de l'activation de cellules de lieu durant les étant de sommeil et d'éveil en générant des petites sous séquences contigues d'activation de cellules de lieu cohérentes entre elles. Ces fragments sont observés en particulier lors d'évènements sharp wave ripple (SPWR).Les phénomènes de rejeu lors du sommeil dans le contexte de la consolidation de la mémoire à long terme ont beaucoup attiré l'attention. Ici nous nous focalisons sur le rôle du rejeu pendant l'état d'éveil.Nous formulons l'hypothèse que ces fragments peuvent être utilisés par le cortex préfrontal pour réaliser une tâche d'apprentissage spatial comprenant plusieurs buts.Nous proposons de développer un modèle intégré d'hippocampe et de cortex préfrontal capable de générer des séquences d'activation de cellules de lieu.Le travail collaboratif proposé prolonge les travaux existants sur un modèle de cognition spatiale pour des tâches orientés but plus simples (Barrera and Weitzenfeld 2008; Barrera et al. 2015) avec un nouveau modèle basé sur le rejeu pour la formation de mémoire dans l'hippocampe et l'apprentissage et génération de séquences spatiales par le cortex préfrontal.En contraste avec les travaux existants d'apprentissage de séquence qui repose sur des règles d'apprentissage sophistiquées, nous proposons d'utiliser un paradigme calculatoire appelé calcul par réservoir (Dominey 1995) dans lequel des groupes importants de neurones artificiels dont la connectivité est fixe traitent dynamiquement l'information au travers de réverbérations. Ce modèle calculatoire par réservoir consolide les fragments de séquence d'activations de cellule de lieu en une plus grande séquence qui pourra être rappelée elle-même par des fragments de séquence.Le travail proposé est supposé contribuer à une nouvelle compréhension du rôle du phénomène de rejeu dans l'acquisition de la mémoire dans une tâche complexe liée à l'apprentissage de séquence.Cette compréhension opérationnelle sera mise à profit et testée dans l'architecture cognitive incarnée d'un robot mobile selon l'approche animat (Wilson 1991) [etc...] / As rats learn to search for multiple sources of food or water in a complex environment, processes of spatial sequence learning and recall in the hippocampus (HC) and prefrontal cortex (PFC) are taking place. Recent studies (De Jong et al. 2011; Carr, Jadhav, and Frank 2011) show that spatial navigation in the rat hippocampus involves the replay of place-cell firing during awake and sleep states generating small contiguous subsequences of spatially related place-cell activations that we will call "snippets". These "snippets" occur primarily during sharp-wave-ripple (SPWR) events. Much attention has been paid to replay during sleep in the context of long-term memory consolidation. Here we focus on the role of replay during the awake state, as the animal is learning across multiple trials.We hypothesize that these "snippets" can be used by the PFC to achieve multi-goal spatial sequence learning.We propose to develop an integrated model of HC and PFC that is able to form place-cell activation sequences based on snippet replay. The proposed collaborative research will extend existing spatial cognition model for simpler goal-oriented tasks (Barrera and Weitzenfeld 2008; Barrera et al. 2015) with a new replay-driven model for memory formation in the hippocampus and spatial sequence learning and recall in PFC.In contrast to existing work on sequence learning that relies heavily on sophisticated learning algorithms and synaptic modification rules, we propose to use an alternative computational framework known as reservoir computing (Dominey 1995) in which large pools of prewired neural elements process information dynamically through reverberations. This reservoir computational model will consolidate snippets into larger place-cell activation sequences that may be later recalled by subsets of the original sequences.The proposed work is expected to generate a new understanding of the role of replay in memory acquisition in complex tasks such as sequence learning. That operational understanding will be leveraged and tested on a an embodied-cognitive real-time framework of a robot, related to the animat paradigm (Wilson 1991) [etc...]

L'auteur n'a pas fourni de résumé en français / High Performance Computing is preparing the era of the transition from Petascale to Exascale. Distributed computing systems are already facing new scalability problems due to the increasing number of computing resources to manage. It is now necessary to study in depth these systems and comprehend their behaviors, strengths and weaknesses to better build the next generation.The complexity of managing users applications on the resources conducted to the analysis of the workload the platform has to support, this to provide them an efficient service.The need for workload comprehension has lead to the collection of traces from production systems and to the proposal of a standard workload format. These contributions enabled the study of numerous of these traces. This also lead to the construction of several models, based on the statistical analysis of the different workloads from the collection.Until recently, existing workload traces did not enabled researchers to study the consumption of resources by the jobs in a temporal way. This is now changing with the need for characterization of jobs consumption patterns.In the first part of this thesis we propose a study of existing workload traces. Then we contribute with an observation of cluster workloads with the consideration of the jobs resource consumptions over time. This highlights specific and unattended patterns in the usage of resources from users.Finally, we propose an extension of the former standard workload format that enables to add such temporal consumptions without loosing the benefit of the existing works.Experimental approaches based on workload models have also served the goal of distributed systems evaluation. Existing models describe the average behavior of observed systems.However, although the study of average behaviors is essential for the understanding of distributed systems, the study of critical cases and particular scenarios is also necessary. This study would give a more complete view and understanding of the performance of the resources and jobs management. In the second part of this thesis we propose an experimental method for performance evaluation of distributed systems based on the replay of production workload trace extracts. These extracts, replaced in their original context, enable to experiment the change of configuration of the system in an online workload and observe the different configurations results. Our technical contribution in this experimental approach is twofold. We propose a first tool to construct the environment in which the experimentation will take place, then we propose a second set of tools that automatize the experiment setup and that replay the trace extract within its original context.Finally, these contributions conducted together, enable to gain a better knowledge of HPC platforms. As future works, the approach proposed in this thesis will serve as a basis to further study larger infrastructures.

Cricket umpiring is demanding. In today’s world where so much technology is available to television viewers, they expect perfect umpiring, forgetting that what they see on their television screens is not available to the umpires standing in the middle of the field making the decisions. This study aimed to examine cricket umpires on-field decisions during the 2007 ICC Cricket World Cup. Examining leg before wicket (LBW) and caught behind decisions, in particular. The researcher made use of a notational analysis program known as Umpirestat to collect the necessary data on each umpire. The umpires were examined in two groups Elite and International umpires (groupings defined by the ICC) and were compared to a base line in the form of Hawk-Eye for LBW decisions and TV replays for caught behind decisions. The umpire groups were compared to each other and then to the base line, an individual umpire comparison was then done within each grouping. The data was statistically analysed using percentages, chi squared and modelling for the Elite Umpires. For LBW decisions there was a difference of 2.02% between Elite and International umpires, showing a similarity between the two groups. The difference between Elite Umpires and Hawk-Eye was 18.83% and between International umpires and Hawk-Eye was 16.81%, showing difference between the three groups. It was found that there was a difference of 3.63% for caught behind decisions between Elite and International umpires, showing a similarity between the two groups. The difference between Elite Umpires and TV replays was 2.99% and between International umpires and TV replays there was 0.64%, showing similarities between the three groups. In conclusion, for LBW appeals there is a similarity between Elite and International umpire groupings however a difference exists between Hawk-Eye and both groupings of umpires. For the individual umpires LBW comparison the following was seen for the International umpires; no real conclusions could be drawn due to limited amounts of data collected on them. However, variances in individual performances were seen within the International umpire grouping. For Elite Umpires, it is clearly seen in the model that the umpires expected number of dismissals got closer together and almost group together at around four appeals. This indicates a strong consensus amongst umpires when dealing with four, five and six appeals during a match. This consensus is more pronounced for four and five appeals. In conclusion, for caught behind appeals there is a similarity between Elite and International umpires as well as a similarity between replays and both groups of umpires. For the individual umpires caught behind comparison the following was seen for the International umpires; as with the LBW results there was limited amount of data available for the caught behind results. Thus no real conclusions could be drawn for International umpire caught behind decisions, although when examining the percentage data, differences can be seen. For the individual umpires caught behind comparison the following was seen for the Elite Umpires, no consensus is seen between the umpires - they all appear to have different expected dismissals to one another. Copyright / Dissertation (MA)--University of Pretoria, 2010. / Biokinetics, Sport and Leisure Sciences / unrestricted

Background. Instant replay allows an application to highlight events without exporting a video of the whole session. Hardware-accelerated video encoding allows replay footage to be encoded in real-time with less to no impact on the runtime performance of the actual simulation in the application. Hardware-accelerated video encoding, however, is not supported on all devices such as low-tier mobile devices, nor all platforms like web browsers. When hardware-acceleration is not supported, the replay has to be encoded using a software-implemented encoder instead. Objectives. To evaluate if recording rendering API commands is a suitable replacement of real-time encoding when hardware-accelerated video encoding is not supported. Method. An experimental research method is used to make quantitative measurements of the proposed approach, Reincore, and a real-time encoder. The measured metrics is frame time and memory consumption. The Godot game engine is modified with modules for real-time video encoding (H.264, H.265 and VP9 codecs) and rendering API command recording and replaying. The engine is also used to create test scenes to evaluate if object count, image motion, object loading/unloading, replay video resolution and replay video duration has any impact on the runtime overhead of frame time and memory consumption. Results. The implemented rendering API command replayer, Reincore, appears to have minimal to no impact on the frame time overhead in all scenarios, except for a spike in increased frame time when the replayer initializes synchronization. Reincore show to be overall inferior to real-time video encoding in terms of runtime memory overhead. Conclusions. Overall, real-time encoding using the H.264 or H.265 show a similar result in frame time as recording rendering commands. However, command recording implies a more significant overhead of memory usage than real-time encoding. The frame time of using the VP9 codec for real-time encoding is inferior to recording rendering API commands. / Bakgrund. Återspelning tillåter applikationer att visa upp händelser utan att exportera en video för hela sessionen. Hårdvaruaccelererad videokodning tillåter video av återspelning att kodas i realtid med minimal påverkan på applikationens prestanda för simulering. Hårdvaruaccelererad videokodning stöds dock inte alltid på alla enheter eller plattformar, så som lågt presterande mobila enheter eller webbläsare. När hårdvaruacceleration inte stöds, måste videokodning ske med en mjukvarubaserad implementering istället. Syfte. Att utvärdera om återspelning genom inspelade renderingskommandon som fördröjer arbetet för videokodning är ett lämpligt alternativ till videokodning i realtid, när hårdvaruacceleration inte stöds. Metod. En experimentel forskningsmetod används för att samla kvantitativ mätdata från den föreslagna tillvägagången, Reincore, and en realtidsvidekodare. Mätdatan består av bildtid och minnesanvändning. Genom att modifiera spelmotorn Godot skapas moduler för realtids-videokodning samt inspelning av renderingskommandon. Spelmotorn används också för att skapa testscener för att utvärdera om antal objekt, bildrörelse, skapande av objekt under körning, upplösning eller videolängd har någon inverkan på bildtid eller minnesanvändning. Resultat. Den implementerade renderingskommando-inspelaren, Reincore, visar minimal påverkan på bildtid, med undantag för en temporär ökning när återspelaren initierar synkronisering. Reincore visar sig vara underlägsen till realtids-videokodning när det gäller minnesanvändning. Slutsatser. Realtids-videokodning med H.264 eller H.265 som video-codec visar övergripande bättre resultat för återspelning än renderingskommandoinspelning, när det gäller både bildtid samt minnesanvändning. Bildtiden för VP9 video-codec för realtids-videokodning visar däremot sämre resultat än renderingskommandinspelning.

Background. One important part of creating quality games is to make sure thegames work as intended. It is done by testing the game, often by playtesters that play the game repeatedly to identify problems. It is a time-consuming task, and some of the testing can be automated. Objectives. This thesis explores two different automated testing approaches. The approaches selected are Record and Replay, and Script-Based Testing. The testing approaches are evaluated on a commercial FPS game. Methods. The scientific method used in this thesis was an experiment to collect data on the two testing approaches. The testing approaches were implemented and then tested on the game. The data collected was the time to create a test case andthe execution time of the test. Results. Creating tests with the Script-Based approach was significantly slower compared to Record and Replay. There was no significant difference in execution time between Record and Replay and Script-Based Testing, but the Script-Based approach is slightly faster in most tests. Conclusions. Record and Replay and Script-Based Testing have different strengths and weaknesses. Record and Replay is fast at creating tests but is not adaptable to changes in the 3D world. Script-Based Testing is slow at creating tests but is adaptable, and therefore less maintenance is required. / Bakgrund. En viktig del av att göra ett bra spel är att säkerställa att spelet fungerar som det ska. Det görs genom testning, oftast av speltestare som spelar spelet om och om igen för att hitta fel. Det är en tidskrävande uppgift och en del av testningen kan automatiseras. Syfte. Denna avhandling undersöker två olika automatiserade testmetoder. De valda metoderna är spela in och spela upp och skriptbaserad metod. Testmetoderna utvärderas på ett kommersiellt FPS-spel. Metod. Den vetenskapliga metoden som användes i denna avhandling var ett exper-iment för att samla in data om de två testmetoderna. Testmetoderna implementerades och testades sedan på spelet. Den insamlade datan var tiden för att skapa ett testfall och testets exekveringstid. Resultat. Att skapa tester med det skriptbaserade metoden var betydligt långsammare jämfört med spela in och spela upp. Det fanns ingen signifikant skillnad i exekveringstid mellan spela in och spela upp och skriptbaserad testning, men den skriptbaserade metoden är något snabbare i de flesta tester. Slutsatser. Spela in och spela upp och den skriptbaserad metoden har olika styrkor och svagheter. Spela in och spela upp är snabbt att skapa tester men kan inte anpassas till förändringar i 3D-världen. Skriptbaserade metoden är långsam när det gäller att skapa tester men är anpassningsbar och därför krävs mindre underhåll.

L'omniprésence des ordinateurs et la demande de toujours plus de puissance poussent les architectes processeur à chercher des moyens d'augmenter les performances de ces processeurs. La tendance actuelle est de répliquer sur une même puce plusieurs cœurs d'exécution pour paralléliser l'exécution. Si elle se poursuit, les processeurs deviendront massivement multicoeurs avec plusieurs centaines voire un millier de cœurs disponibles. Cependant, la loi d'Amdahl nous rappelle que l'augmentation de la performance séquentielle sera toujours nécessaire pour améliorer les performances globales. Une voie essentielle pour accroître la performance séquentielle est de perfectionner le traitement des branchements, ceux-ci limitant le parallélisme d'instructions. La prédiction de branchements est la solution la plus étudiée, dont l'intérêt dépend essentiellement de la précision du prédicteur. Au cours des dernières années, cette précision a été continuellement améliorée et a atteint un seuil qu'il semble difficile de dépasser. Une autre solution est d'éliminer les branchements et de les remplacer par une construction reposant sur des instructions prédiquées. L'exécution des instructions prédiquées pose cependant plusieurs problèmes dans les processeurs à exécution dans le désordre, en particulier celui des définitions multiples. Les travaux présentés dans cette thèse explorent ces deux aspects du traitement des branchements. La première partie s'intéresse à la prédiction de branchements. Une solution pour améliorer celle-ci sans augmenter la précision est de réduire le coût d'une mauvaise prédiction. Cela est possible en exploitant la reconvergence de flot de contrôle et l'indépendance de contrôle pour récupérer une partie du travail fait par le processeur sur le mauvais chemin sur les instructions communes aux deux chemins pour éviter de le refaire sur le bon chemin. La deuxième partie s'intéresse aux instructions prédiquées. Nous proposons une solution au problème des définitions multiples qui passe par la prédiction sélective de la valeur des prédicats. Un mécanisme de rejeu sélectif est utilisé pour réduire le coût d'une mauvaise prédiction de prédicat. / Computers are everywhere and the need for always more computation power has pushed the processor architects to find new ways to increase performance. The today's tendency is to replicate execution core on the same die to parallelize the execution. If it goes on, processors will become manycores featuring hundred to a thousand cores. However, Amdahl's law reminds us that increasing the sequential performance will always be vital to increase global performance. A perfect way to increase sequential performance is to improve how branches are executed because they limit instruction level parallelism. The branch prediction is the most studied solution, its interest greatly depending on its accuracy. In the last years, this accuracy has been continuously improved up to reach a hardly exceeding limit. An other solution is to suppress the branches by replacing them with a construct based on predicated instructions. However, the execution of predicated instructions on out-of-order processors comes up with several problems like the multiple definition problem. This study investigates these two aspects of the branch treatment. The first part is about branch prediction. A way to improve it without increasing the accuracy is to reduce the coast of a branch misprediction. This is possible by exploiting control flow reconvergence and control independence. The work done on the wrong path on instructions common to the two paths is saved to be reused on the correct path. The second part is about predicated instructions. We propose a solution to the multiple definition problem by selectively predicting the predicate values. A selective replay mechanism is used to reduce the cost of a predicate misprediction.

Depurar sistemas distribuídos continua uma tarefa difícil, mesmo após 30 anos de pesquisa intensa. Embora essa situação possa ser parcialmente atribuída à complexidade das execuções concorrentes, o rápido passo de desenvolvimento das plataformas e tecnologias para computação distribuída também carrega a sua parcela de culpa, por encurtar a vida de muitas ferramentas potencialmente úteis. Neste trabalho, apresentamos uma análise dos principais problemas, técnicas e ferramentas ligados à depuração de sistemas concorrentes e discutidos na literatura. Baseados nessa análise, desenvolvemos e apresentamos uma nova técnica, simples e portátil, que pode ser aplicada a sistemas distribuídos que utilizam chamadas síncronas e bloqueantes. Essa técnica, concebida para sobreviver à heterogeneidade, é validada por meio da implementação de um arcabouço escrito para plataforma Eclipse e instanciado para sistemas de objetos distribuídos baseados em Java/CORBA. / After over thirty years of intense research, debugging distributed systems is still regarded as a difficult task. While this situation could be partially blamed on the fact that concurrent executions are complex, the fast pace of evolution witnessed with distributed computing technologies have also played its by shortening the lifespan of many potentially useful debugging tools. This work presents an analysis of the main issues, techniques and tools in the field of parallel, distributed, and concurrent debugging in general. Based on this analysis, we develop and present a simple and portable technique targeted at synchronous-call-based distributed systems. This technique, designed for portability, is validated through the implementation of an Eclipse-based framework that is instantiated for Java/CORBA distributed object systems.

Les systèmes embarqués peuvent contenir des données sensibles. Elles sont généralement échangées en clair entre le système sur puces et la mémoire, mais aussi en interne. Cela constitue un point faible: un attaquant peut observer cet échange et récupérer des informations ou insérer du code malveillant. L'objectif de la thèse est de fournir une solution dédiée et adaptée à ces problèmes en considérant l'intégralité de la durée de vie du système embarqué (démarrage, mises à jour et exécution) et l'intégralité des données (bitstream du FPGA, noyau du système d'exploitation, code et données critiques). En outre, il est nécessaire d'optimiser les performances des mécanismes matériels de sécurité introduits afin de correspondre aux attentes des systèmes embarqués. Cette thèse se distingue en proposant des solutions innovantes et adaptées au monde des FPGAs. / Embedded systems may contain sensitive data. They are usually exchanged in plaintext between the system on chips and the memory, but also internally. This is a weakness: an attacker can spy this exchange and retrieve information or insert malicious code. The aim of the thesis is to provide a dedicated and suitable solution for these problems by considering the entire lifecycle of the embedded system (boot, updates and execution) and all the data (FPGA bitstream, operating system kernel, critical data and code). Furthermore, it is necessary to optimize the performance of hardware security mechanisms introduced to match the expectations of embedded systems. This thesis is distinguished by offering innovative and suitable solutions for the world of FPGAs.

Depurar sistemas distribuídos continua uma tarefa difícil, mesmo após 30 anos de pesquisa intensa. Embora essa situação possa ser parcialmente atribuída à complexidade das execuções concorrentes, o rápido passo de desenvolvimento das plataformas e tecnologias para computação distribuída também carrega a sua parcela de culpa, por encurtar a vida de muitas ferramentas potencialmente úteis. Neste trabalho, apresentamos uma análise dos principais problemas, técnicas e ferramentas ligados à depuração de sistemas concorrentes e discutidos na literatura. Baseados nessa análise, desenvolvemos e apresentamos uma nova técnica, simples e portátil, que pode ser aplicada a sistemas distribuídos que utilizam chamadas síncronas e bloqueantes. Essa técnica, concebida para sobreviver à heterogeneidade, é validada por meio da implementação de um arcabouço escrito para plataforma Eclipse e instanciado para sistemas de objetos distribuídos baseados em Java/CORBA. / After over thirty years of intense research, debugging distributed systems is still regarded as a difficult task. While this situation could be partially blamed on the fact that concurrent executions are complex, the fast pace of evolution witnessed with distributed computing technologies have also played its by shortening the lifespan of many potentially useful debugging tools. This work presents an analysis of the main issues, techniques and tools in the field of parallel, distributed, and concurrent debugging in general. Based on this analysis, we develop and present a simple and portable technique targeted at synchronous-call-based distributed systems. This technique, designed for portability, is validated through the implementation of an Eclipse-based framework that is instantiated for Java/CORBA distributed object systems.