Dans l’industrie de la micro-électronique, les efforts à fournir pour les nouvelles applications développées deviennent de plus en plus contraignants et difficiles à supporter en terme de coût. Les agressions provenant des décharges électrostatiques (ESD) générées par l’environnement direct sur les puces constituent un facteur important de la chute de rendement et donc des coûts. Ces difficultés s’ajoutent aux limites physiques plus strictes pour fabriquer des transistors lorsque l’on aborde des échelles nanométriques. La technologie Silicium sur Isolant (SOI) a été développée afin de contourner cette difficulté, mais l’intégration des protections ESD limite son émergence du fait de la complexité de la mise au point et du développement d’un réseau de protection pour la puce.L’objectif annoncé de ce travail de recherche, effectué en collaboration entre STMicroelectronics le CEA et l’IMEP est d’évaluer les caractéristiques principales de la technologie pour la protection contre les décharges et de proposer une stratégie innovante de protection adaptée au SOI. En effet, à partir de résultats expérimentaux, nous avons pu constater que l’oxyde enterré, le BOX, limite les performances en robustesse et diminue la fenêtre de conception pour le déclenchement des protections. Pour y remédier, une structure commandée bidirectionnelle a été développée sur PDSOI afin de faciliter la dissipation thermique et améliorer la robustesse. Pour prolonger cette solution sur technologie FDSOI, une étude approfondie sur le thyristor afin a été menée afin de porter cette solution. L’analyse de simulation 3D et de résultats silicium ont permis de proposer une stratégie de protections innovantes pour le thyristor sur FDSOI. / In the microelectronics industry, the fabrication process for advanced technological nodes becomes more and more cumbersome and limiting in terms of cost. The electrostatic discharges (ESD) generated by the direct environment affect the circuits and constitute an important factor for the decrease of the yield and thus result in an increase of the costs. Apart from these difficulties, there are also issues arising from the physical limits of transistor integration when reaching the nanoscale.The Silicon on Insulator (SOI) technology was developed in order to bypass this difficulty. However, the integration of ESD protections limits its emergence due to the development complexity and the protection circuit needed. The goal of this work which was a collaboration between STMicroelectronics, CEA and IMEP was to evaluate the principal characteristics of this technology for electrostatic discharge protection and propose a novel protection strategy adapted for SOI.In fact, we were able to confirm from experimental results that the buried oxide (BOX) limits the performances in terms of robustness and narrows the window of conception for the triggering of the protections. A commanded bidirectional structure was developed on PDSOI and proposed as a solution to facilitate the thermal dissipation and improve the robustness.In order to extend this solution on FDSOI technology, a detailed study on the thyristor was performed. Analysis of the 3D simulations and experimental results permitted to propose an innovative strategy for ESD protections on FDSOI.

The essence of this diploma thesis is the design and implementation of driver for Profinet bus coupler from Phoenix Contact, which use computer's network card for communication. The proposal builds on the knowledge gained from available literature and analysis of Profinet protocol.

The Io UV footprint (IFP) is one of the most spectacular signatures of the Io-Jupiter interaction. It consists of several auroral spots and an extended tail which are located close to the feet of the magnetic field lines passing through Io in each hemisphere. The purpose of the present study is to demonstrate that a careful analysis of the Io UV footprint based on observations acquired with the STIS and ACS high resolution and high sensitivity FUV cameras on board the Hubble Space Telescope can provide us with essential information on the ongoing physical processes. The thesis is organized around basic questions: What is the Io footprint?, Where is the Io footprint?, How high is the Io footprint?, How big is the Io footprint? and finally: How bright is the Io footprint? The answers to these questions have profound implications for the understanding of the phenomenon. Among the most important results of this work is the unexpected finding of a faint auroral spot appearing upstream of the main Io spot in one hemisphere while only downstream spots are seen in the opposite hemisphere. The detailed study of the evolution of the inter-spot distances puts previous models describing the footprint morphology under question. We propose a new interpretation which involves that some spots are caused by electrons accelerated away from the planet along the field lines in one hemisphere, crossing the equatorial plane in the form of electron beams and precipitating in the opposite hemisphere, creating the so-called Trans-hemispheric Electron Beam (TEB) spots. The information provided by the position of the satellite footprints is not restricted to the interaction between the moon and the Jovian magnetosphere. The analysis of the footpaths of Io, Europa and Ganymede helped us to further constrain the magnetic field models, notably through the identification of a large magnetic anomaly in the northern hemisphere. Additionally, the study of the speed of the Io footprint along its reference contour suggests that a second anomaly regions may also exist in the North. In this work, we present a new and direct method to measure the altitude of the different footprint features. The main spot and the tail emissions have a peak altitude of 900 km while the peak altitude of the Trans-hemispheric Electron Beam spot is 700 km. These results suggest that the main spot and tail emissions are caused by the precipitation of electrons with a mean energy around 1 keV, far lower than the 55 keV value previously derived from spectral measurements. The vertical extent of these emissions is surprisingly broad (scale height ~400 km) and is best fitted with an incoming kappa electron energy distribution (spectral index ~2.3). This suggests that the electron acceleration is supplied by processes related to inertial Alfvén waves rather than by quasi-static potentials as proposed by some theoretical models. The size of the main footprint spot is carefully estimated on a much larger image sample than before: its length along the footpath is ~900 km while its width perpendicular to the footpath is <200 km. Larger lengths are sometimes observed but in that case, they are attributed to the mix of individual spots. The spot length is larger than the projected diameter of Io around the magnetic field lines but is consistent with recent simulations. As far as the Io footprint brightness is concerned, variations on two timescales have been studied. On timescales of minutes, systematic brightness fluctuation on the order of 30% (and going up to 50%) are observed. Additionally, cases of simultaneous variations of the main and the TEB spots are reported, which suggests that the process that triggers these fast variations is located close to the planet. Variations of the main spot brightness with the System III longitude of Io are also analyzed. Our new measurement method fully considering the multi-spot structure of the IFP and the real geometry of the observations provides more accurate estimates for the precipitating energy flux (between 100 and 500 mW/m for the main spot). The main spot brightness peaks at 110° and 290° longitude, which could be attributed either to an enhanced interaction strength when Io is near the dense torus center or to spots merging which is also observed to occur in these sector. Nevertheless, strong North-South asymmetries are also observed, which suggests that the surface magnetic field strength also influences the spots brightness. ---------------------------------------------------------------------- L'empreinte aurorale d'Io est l'une des signatures les plus impressionnantes de l'interaction Io-Jupiter. Présente dans chaque hémisphère, elle se situe à proximité des pieds des lignes de champ magnétique qui interceptent Io et se compose de plusieurs taches suivies d'une longue trainée. Cette étude vise à démontrer qu'une analyse minutieuse de l'empreinte UV d'Io basée sur les observations des instruments STIS et ACS du Télescope Spatial Hubble peut apporter des informations cruciales sur les processus physiques qui sont en jeu. Cette thèse est organisée autour de questions relativement basiques: Qu'est-ce que l'empreinte d'Io?, Où se trouve-t-elle?, A quelle altitude se trouve-t-elle?, Quelle est sa taille? et enfin Quelle est sa brillance?. Les réponses à ces questions ont de profondes implications pour la compréhension du phénomène. Parmi les résultats principaux de ce travail, il y a la découverte inattendue d'une faible tache aurorale apparaissant en amont de la tache principale dans un hémisphère alors que les seules taches observées dans l'hémisphère opposé sont situées en aval. L'étude détaillée de la distance inter-taches remet en question les précédents modèles décrivant la morphologie de l'empreinte. Nous proposons ici une nouvelle interprétation de certaines de ces taches: elles seraient causées par des électrons initialement accélérés le long des lignes de champ dans la direction opposée à Jupiter, qui ensuite traverseraient le plan équatorial sous la forme de faisceaux d'électrons et qui précipiteraient finalement dans l'hémisphère opposé en générant la tache du Faisceau d'Electrons Trans-hemisphérique (FET). Les informations fournies par la position des empreintes de satellites ne se limitent pas à l'interaction entre Io et la magnétosphère de Jupiter. L'analyse des contours parcourus par les empreintes d'Io, d'Europe et de Ganymède permet de mieux contraindre les modèles de champ magnétique joviens, entre autre à travers l'identification d'une importante anomalie magnétique dans l'hémisphère nord. De plus, l'étude de la vitesse de l'empreinte d'Io le long du contour de référence suggère qu'elle pourrait être accompagnée d'une deuxième anomalie dans cet hémisphère. Dans cette étude, nous présentons une méthode directe pour mesurer l'altitude des différentes sous-structures qui forment l'empreinte. Le pic d'émissions de la tache principale et de la trainée est situé à 900 km d'altitude alors que celui de la tache FET est à 700 km. Ces résultats suggèrent que la tache principale et la trainée sont la conséquence de la précipitation d'électrons ayant une énergie moyenne d'approximativement 1 keV, une valeur largement inférieure aux 55 keV déduits à partir de précédentes mesures spectrales. L'extension verticale de ces émissions est étonnamment large (hauteur d'échelle: ~400 km) et la distribution d'énergie des électrons incidents qui reproduit au mieux les observations est une distribution kappa d'indice spectral ~2.3. Cela suggère que l'accélération des électrons est liée à des ondes d'Alfvén inertielles plutôt qu'aux potentiels quasi-statiques proposés par certains modèles théoriques. La taille de la tache principale a été mesurée sur un ensemble d'images beaucoup plus étendu qu'auparavant: sa longueur le long du contour est de ~900 km alors que sa largeur telle que mesurée perpendiculairement à celui-ci est de <200 km. Des longueurs plus importants sont parfois observées mais elles résultent de la superposition partielle de plusieurs taches individuelles. La longueur des taches est plus grande que la projection du diamètre d'Io le long des lignes de champ, ce qui était prévu par des simulations récentes. En ce qui concerne la brillance des taches, deux échelles de temps ont été étudiées en particulier. A l'échelle de la minute, nous avons mis en évidence des fluctuations de l'ordre de 30% de la brillance moyenne et pouvant atteindre jusqu'à 50 % de celle-ci. Dans certains cas, on observe des variations corrélées de la tache principale et de la tache FET, ce qui suggère que le processus qui induit ces variations rapides se situe près de la surface de Jupiter. Les variations de la brillance de la tache principale en fonction de la longitude Système III d'Io ont également été analysées. Notre nouvelle méthode de mesure prend pleinement en compte la géométrie de l'observation ainsi que le fait que l'empreinte est composée de différentes taches, ce qui permet une estimation plus précise du flux d'énergie incident (entre 100 et 500 mW/m pour la tache principale). La brillance de la tache principale possède deux maxima, un à 110° et un autre à 290° de longitude. Ces augmentations de brillance peuvent avoir deux origines: soit elles sont dues à l'augmentation de l'intensité de l'interaction entre Io et le plasma quand Io est proche du centre du tore, soit elles sont liées à la superposition des taches principales et FET qui se produit également dans ces secteurs. Néanmoins, de fortes asymétries Nord-Sud sont aussi observées, ce qui semble indiquer que l'intensité du champ magnétique de surface joue aussi un rôle en ce qui concerne la brillance des spots.

Digital computers have changed human society in a profound way over the last 50 years. Key properties that contribute to the success of the computer are flexible programmability and fast access to large amounts of data and instructions. Effective access to algorithms and data is a fundamental property that limits the capabilities of computer systems. For PC computers, the main memory consists of dynamic random access memory (DRAM). Communication between memory and processor has traditionally been performed over a multi-drop bus. Signal frequencies on these buses have gradually increased in order to keep up with the progress in integrated circuit data processing capabilities. Increased signal frequencies have exposed the inherent signal degradation effects of a multidrop bus structure. As of today, the main approach to tackle these effects has been to reduce the number of endpoints of the bus structure. Though improvements in DRAM memory technology have increased the available memory size at each endpoint, the increase has not been able to fully fulfill the demand for larger system memory capacity. Different bus structural changes have been used to overcome this problem. All are different compromises between access latency, data transmission capacity, memory capacity, and implementation costs. In this thesis we focus on using the signal processing capabilities of a modern integrated circuit technology as an alternative to bus structural changes. This has the potential to give low latency, high memory capacity, and relatively high data transmission capacity at an additional cost limited to integrated circuit blocks. We first use information theory to estimate the unexplored potential of existing multi-drop bus structures. Hereby showing that reduction of the number of endpoints for multi-drop buses, is by no means based on the fundamental limit of the data transmission capacity of the bus structure. Two test-chips have been designed and fabricated to experimentally demonstrate the feasibility of several Gb/s data-rates over multidrop buses, with limited cost overhead and no latency penalty. The test-chips implement decision feedback equalization, adopted for high speed multi-drop use. The equalizers feature digital filter implementations which, in combination with high speed DACs, enable the use of long digital filters for high speed decision feedback equalization. Blind adaptation has also been implemented to demonstrate extraction of channel characteristics during data transmission. The use of single sided equalization has been proposed in order to limit the need for equalization implementation to the host side of a DRAM memory bus. Furthermore, we propose to utilize the reciprocal properties of the communication channel to ensure that single sided equalization can be performed without any channel characterization hardware on the memory chips. Finally, issues related to evaluation of high-speed channels are addressed and the on-chip structures used for channel evaluation in this project are presented.

Io, the innermost Galilean satellite of Jupiter, exhibits a wide variety of complex phenomena such as interaction with Jupiter’s magnetosphere, volcanic activity, and a rarefied multi-species sublimating and condensing atmosphere with an ionosphere. Io’s orbital resonance with Jupiter and the other Galilean satellites produces intense tidal heating. This makes Io the most volcanically active body in the solar system with plumes that rise hundreds of kilometers above the surface. In the present work, the interaction of Io’s atmosphere with the Jovian plasma torus is simulated via the Direct Simulation Monte Carlo (DSMC) method and the aurora produced via electron-neutral excitation collisions is examined using electron transport Monte Carlo simulation. The electron-transport Monte Carlo simulation models the electron collisions with the neutral atmosphere and their transport along field lines as they sweep past Io, using a pre-computed steady atmosphere and magnetic field. As input to the Monte Carlo simulation, the neutral atmosphere was first modeled using prior 2D sunlit continuum simulations of Io’s atmosphere produced by others. In order to justify the use of a sunlit atmosphere for eclipse, 1D two-species (SO2 and a non-condensable) DSMC simulations of Io’s atmospheric dynamics during and immediately after eclipse were performed. It was found that the inclusion of a non-condensable species (SO or O2) leads to the formation of a diffusion layer which prevents rapid collapse. The degree to which the diffusion layer slowed the atmospheric collapse was found to be extremely sensitive to both the initial non-condensable mole fraction and the reaction (or sticking) probability on the surface of the “non-condensable”. Furthermore, upon egress, vertical stratification of the atmosphere occurred with the non-condensable species being lifted to higher altitudes by the rapid sublimation of SO2 as the surface warms. Simulated aurorae (specifically the [OI] 6300 Å and the S2, SO, and SO2 molecular band emission in the middle ultraviolet) show good agreement with observations of Io in eclipse and an attempt was made to use the simulations to constrain the upstream torus electron temperature and Io’s atmospheric composition, structure, and volcanic activity. It is found that the position of the bright [OI] 6300 Å wake spot relative to Io’s equator depends on the position of Io relative to the plasma torus’ equator and the asymmetric electron number flux that results. Using HST/STIS UV-Vis spectra, the upstream electron temperature is weakly constrained to be between 3 eV and 8 eV depending on the flux of a low energy (35 eV), non-thermal component of the plasma (more non-thermal flux requires lower thermal plasma temperatures to fit the spectrum). Furthermore, an upper limit of 5% of the thermal torus density (or 180 cm−3 based on the Galileo J0 plasma density at Io) is obtained for the low energy non-thermal component of the plasma. These limits are consistent with Galileo observations of the upstream torus temperature and estimates for the the non-thermal component. Finally, plume activity and S2 content during eclipse observations with HST/STIS were constrained by examining the emission intensity along the spatial axis of the aperture. During the August 1999 UV-Vis observations, the auroral simulations indicate that the large volcanoes Pele and Surt were inactive whereas Tvashtar was active and that Dazhbog and possibly Loki were also actively venting gas. The S2 content inferred for the large Pele-type plumes was between 5% (Tvashtar) and 30% (Loki, if active), consistent with prior observations (Spencer et al., 2000; Jessup et al., 2007). A 3D DSMC simulation of Io’s sublimation and sputtered atmosphere including photo- and plasma-chemistry was developed. In future work these atmospheric simulations will replace the continuum target atmosphere in the auroral model and thus enable a better match to the observed high altitude auroral emission. In the present work, the plasma interaction is modeled by a flux of ions and electrons which flow around and through Io’s atmosphere along pre-computed fields and interact with the neutral gas. A 3D DSMC simulation of Io’s atmosphere assuming a simple thermal model for the surface just prior to ingress into eclipse and uniform frost coverage has been performed in order to understand how Io’s general atmospheric dynamics are affected by the new plasma model with chemistry and sputtering. Sputtering was found to supply most of the nightside atmosphere (producing an SO2 column of ~5×1013 cm−2); however, the dense dayside sublimation atmosphere was found to block sputtering of the surface. The influence of the dynamic plasma pressure on the day-to-night circumplanetary flow was found to be quite substantial causing the day-to-night wind across the dawn terminator to flow slightly towards the equator. This results in a region of high density near the equator that extends far (~2000 km for the condensable species) onto the nightside across the dawn terminator. Thus, even without thermal lag due to rotation or variable surface frost, highly asymmetric equatorial column densities relative to the subsolar point are obtained. The non-condensable O2, which is a trace species on the dayside, is the dominant species on the nightside despite increased SO2 sputtering because the loss rate of O2 is slow. Finally, a very intriguing O2 flow feature was observed near the dusk terminator where the flow from the leading hemisphere (pushed by the plasma) meets the flow from the dayside trailing hemisphere. Since the O2 does not condense on the surface, it slowly convects towards the poles and then back onto the nightside, eventually to be dissociated or stripped away by the plasma. / text

Over the last fifty years, our knowledge of the Solar System has increased exponentially. Many planetary surfaces were seen for the first time through spacecraft observations. Yet the interiors of most planetary bodies remain poorly studied. This dissertation focuses on two main topics: the formation of central pit craters and what this reveals about the subsurface volatile content of the target material, and the mantle dynamics of Io and how they relate to the extensive volcanism on its surface. Central pit craters are seen on icy satellites, Mars, the Moon, and Mercury. They have terraced rims, flat floors, and a pit at or near their center. Several formation mechanisms have been suggested. This dissertation assesses the feasibility of central pit crater formation via drainage of impact melt through impact-generated fractures. For impacts on Ganymede, the expected volume of melt and volume of fracture space generated during the impact and the volume of melt able to drain before fractures freeze shut all exceed the observed central pit volumes on Ganymede. This suggests that drainage of impact melt could contribute to central pit crater formation on Ganymede. Molten rock draining through solid rock fractures will freeze shut more rapidly, so this work suggests that impact melt drainage is unlikely to be a significant factor in the formation of central pit craters on rocky bodies unless a significant amount of volatiles are present in the target. Io is the most volcanically active body in the Solar System. While volcanoes are most often associated with plate tectonics on Earth, Io shows no signs of plate tectonics. Previous work has suggested that Io could lose a significant fraction of its internal heat through volcanic eruptions. In this dissertation, I investigate the relationship between mantle convection and magma generation, migration by porous flow, and eruptions on Io. I couple convective scaling laws to a model solving the two-phase flow equations applied to a rising column of mantle. I show that Io has a partially molten upper mantle and loses the majority of its internal heat through volcanic eruption. Next, I present two-dimensional numerical simulations that self-consistently solve the two-phase flow equations including mantle convection and magma generation, migration by porous flow, and eruption. These simulations produce a high heat flux due to volcanic eruption, a thick lithosphere, a partially molten upper mantle, and a high eruption rate—all consistent with observations of Io. This model also reveals the eruption rate oscillates around the statistical steady state average eruption rate suggesting that the eruption rate and total heat flux measurements from the past 35 years may not be representative of Io's long term behavior.

No description available.

Electrostatic Discharges (ESD) is a significant hazard to electronic components and systems. Based on a specific process technology, a given circuit application requires a customized ESD consideration that meets all the requirements such as the core circuit's operating condition, maximum accepted leakage current, breakdown conditions for the process and overall device sizes. In every several years, there will be a new process technology becomes mature, and most of those new technology requires custom design of effective ESD protection solution. And usually the design window will shrinks due to the evolving of the technology becomes smaller and smaller. The ESD related failure is a major IC reliability concern and results in a loss of millions dollars each year in the semiconductor industry. To emulate the real word stress condition, several ESD stress models and test methods have been developed. The basic ESD models are Human Body model (HBM), Machine Mode (MM), and Charge Device Model (CDM). For the system-level ESD robustness, it is defined by different standards and specifications than component-level ESD requirements. International Electrotechnical Commission (IEC) 61000-4-2 has been used for the product and the Human Metal Model (HMM) has been used for the system at the wafer level. Increasingly stringent design specifications are forcing original equipment manufacturers (OEMs) to minimize the number of off-chip components. This is the case in emerging multifunction mobile, industrial, automotive and healthcare applications. It requires a high level of ESD robustness and the integrated circuit (IC) level, while finding ways to streamline the ESD characterization during early development cycle. To enable predicting the ESD performance of IC's pins that are directly exposed to a system-level stress condition, a new the human metal model (HMM) test model has been introduced. In this work, a new testing methodology for product-level HMM characterization is introduced. This testing framework allows for consistently identifying ESD-induced failures in a product, substantially simplifying the testing process, and significantly reducing the product evaluation time during development cycle. It helps eliminates the potential inaccuracy provided by the conventional characterization methodology. For verification purposes, this method has been applied to detect the failures of two different products. Addition to the exploration of new characterization methodology that provides better accuracy, we also have looked into the protection devices itself. ICs for emerging high performance precision data acquisition and transceivers in industrial, automotive and wireless infrastructure applications require effective and ESD protection solutions. These circuits, with relatively high operating voltages at the Input/Output (I/O) pins, are increasingly being designed in low voltage Complementary Metal-Oxide-Semiconductor (CMOS) technologies to meet the requirements of low cost and large scale integration. A new dual-polarity SCR optimized for high bidirectional blocking voltages, high trigger current and low capacitance is realized in a sub 3-V, 180-nm CMOS process. This ESD device is designed for a specific application where the operating voltage at the I/O is larger than that of the core circuit. For instance, protecting high voltage swing I/Os in CMOS data acquisition system (DAS) applications. In this reference application, an array of thin film resistors voltage divider is directly connected to the interface pin, reducing the maximum voltage that is obtained at the core device input down to ± 1-5 V. Its ESD characteristics, including the trigger voltage and failure current, are compared against those of a typical CMOS-based SCR. Then, we have looked into the ESD protection designs into more advanced technology, the 28-nm CMOS. An ESD protection design builds on the multiple discharge-paths ESD cell concept and focuses the attention on the detailed design, optimization and realization of the in-situ ESD protection cell for IO pins with variable operation voltages. By introducing different device configurations fabricated in a 28-nm CMOS process, a greater flexibility in the design options and design trade-offs can be obtained in the proposed topology, thus achieving a higher integration and smaller cell size definition for multi-voltage compatibility interface ESD protection applications. This device is optimized for low capacitance and synthesized with the circuit IO components for in-situ ESD protection in communication interface applications developed in a 28-nm, high-k, and metal-gate CMOS technology. ESD devices have been used in different types of applications and also at different environment conditions, such as high temperature. At the last section of this research work, we have performed an investigation of several different ESD devices' performance under various temperature conditions. And it has been shown that the variations of the device structure can results different ESD performance, and some devices can be used at the high temperature and some cannot. And this investigation also brings up a potential threat to the current ESD protection devices that they might be very vulnerable to the latch-up issue at the higher temperature range.

Due to the explosive growth in the size of scientific data sets, data-intensive computing is an emerging trend in computational science. HPC applications are generating and processing large amount of data ranging from terabytes (TB) to petabytes (PB). This new trend of growth in data for HPC applications has imposed challenges as to what is an appropriate parallel programming framework to efficiently process large data sets. In this work, we study the applicability of two programming models (MPI/MPI-IO and MapReduce) to a variety of I/O-intensive HPC applications ranging from simulations to analytics. We identify several performance and programmer productivity related limitations of these existing programming models, if used for I/O-intensive applications. We propose new frameworks which will improve both performance and programmer productivity for the emerging I/O-intensive applications. Message Passing Interface (MPI) is widely used for writing HPC applications. MPI/MPI- IO allows a fine-grained control of assigning data and task distribution. At the programming frameworks level, various optimizations have been proposed to improve the performance of MPI/MPI-IO function calls. These performance optimizations are provided as various function options to the programmers. In order to write an efficient code, they are required to know the exact usage of the optimization functions, hence programmer productivity is limited. We propose an abstraction called Reduced Function Set Abstraction (RFSA) for MPI-IO to reduce the number of I/O functions and provide methods to automate the selection of appropriate I/O function for writing HPC simulation applications. The purpose of RFSA is to hide the performance optimization functions from the application developer, and relieve the application developer from deciding on a specific function. The proposed set of functions relies on a selection algorithm to decide among the most common optimizations provided by MPI-IO. Additionally, many application scientists are looking to integrate data-intensive computing into computational-intensive High Performance Computing facilities, particularly for data analytics. We have observed several scientific applications which must migrate their data from an HPC storage system to a data-intensive one. There is a gap between the data semantics of HPC storage and data-intensive system, hence, once migrated, the data must be further refined and reorganized. This reorganization must be performed before existing data-intensive tools such as MapReduce can be effectively used to analyze data. This reorganization requires at least two complete scans through the data set and then at least one MapReduce program to prepare the data before analyzing it. Running multiple MapReduce phases causes significant overhead for the application, in the form of excessive I/O operations. For every MapReduce application that must be run in order to complete the desired data analysis, a distributed read and write operation on the file system must be performed. Our contribution is to extend Map-Reduce to eliminate the multiple scans and also reduce the number of pre-processing MapReduce programs. We have added additional expressiveness to the MapReduce language in our novel framework called MapReduce with Access Patterns (MRAP), which allows users to specify the logical semantics of their data such that 1) the data can be analyzed without running multiple data pre-processing MapReduce programs, and 2) the data can be simultaneously reorganized as it is migrated to the data-intensive file system. We also provide a scheduling mechanism to further improve the performance of these applications. The main contributions of this thesis are, 1) We implement a selection algorithm for I/O functions like read/write, merge a set of functions for data types and file views and optimize the atomicity function by automating the locking mechanism in RFSA. By running different parallel I/O benchmarks on both medium-scale clusters and NERSC supercomputers, we show an improved programmer productivity (35.7% on average). This approach incurs an overhead of 2-5% for one particular optimization, and shows performance improvement of 17% when a combination of different optimizations is required by an application. 2) We provide an augmented Map-Reduce system (MRAP), which consist of an API and corresponding optimizations i.e. data restructuring and scheduling. We have demonstrated up to 33% throughput improvement in one real application (read-mapping in bioinformatics), and up to 70% in an I/O kernel of another application (halo catalogs analytics). Our scheduling scheme shows performance improvement of 18% for an I/O kernel of another application (QCD analytics).

The structure of the plasma in Jupiter’s vast magnetosphereis complicated and not fully understood. One way to study the plasma is to look at auroral emissions from the moonIo as it moves through different regions of the plasma torus that surrounds Jupiter. In this paper, the correlation between aurorabrightness on Io and the plasma density at the position of the moon is investigated. If a correlation exists, auroral emissonson Io could be used as a diagnostic for the current state ofJupiter’s plasma environment. For this purpose, a model of the Io plasma torus is developed, combining ideas from different existing models. The model is compared with observations of aurorae on Io made by the Hubble Space Telescope. Io’s position at the time of the observations is obtained with SPICE, a software developed by NASA. A moderate correlation is found when using the whole data set of observations. However, a strong correlation is found for observations on the dusk side of Jupiter. Strong correlations are also found when studying individual years and epochs. / Strukturen på plasman i Jupiters vidsträckta magnetosfär är komplicerad och inte fullständigt känd. Ett sätt att studera plasman är att undersöka ljuset från polarsken på månen Io då den passerar genom olika regioner av det torusformade plasmamolnet som omsluter Jupiter. I denna artikel undersöks korrelationen mellan polarskenets ljusstyrka och plasmans densitiet kring månens position. Om ett sådant samband finns skulle ljusstyrkan hos månens polarsken kunna användas som diagnostik för plasmans aktuella tillstånd. För detta ändamål utvecklas en modell av plasmatorusen genom att kombinera ide´er från flera tidigare modeller. Modellen jämförs sedan med observationer av polarskenet på Io genomförda med rymdteleskopet Hubble. Månens position vid de olika tidpunkterna bestäms med hjälp av SPICE, en mjukvara utvecklad av NASA. En måttligt stark korrelation uppnås när hela datamängden används. När däremot endast data från Jupiters gryningssida används uppnås en korrelation. Det hittas även starka samband när enskilda år studeras. / Kandidatexjobb i elektroteknik 2022, KTH, Stockholm