Sexual size dimorphism (SSD) defined by differences in body size of a conspecific male and female are widespread phenomenon in the animal kingdom and ungulates belong among the most dimorphic mammals. In most species males are the larger sex which is often explained by differing sex-specific reproductive roles. While parental investment is predominantly left to females which are the selective sex, males have to fight for access to receptive mates in intensive combats where body size, strength, and condition are often critical. The relationship between male body size and reproductive success varies according to a mating system with the highest SSD being achieved by harem and promiscuous species. Even though several compilation studies of SSD have been done on ungulates it is rare that systematic research is closely concentrated on a well-defined specialised homogenous group where detailed knowledge on its life-history traits is also available. I have focused on subfamily Caprinae and Bovinae with the objective to conduct a detailed analysis of their SSD and its evolutionary traits. Using advanced phylogenetic methods I could reconstruct the ancestral state in wild goats and sheep that was characterised by medium SSD which then took two different routes of evolution depending on a type of habitat and...

ITC/USA 2014 Conference Proceedings / The Fiftieth Annual International Telemetering Conference and Technical Exhibition / October 20-23, 2014 / Town and Country Resort & Convention Center, San Diego, CA / The decision to buy or develop any system component is a difficult one. Clearly, the development time and cost is less when using an off-the-shelf component; however, these components attempt to meet the requirements of a main-stream market segment. Outside of those markets, a designer may need to add unique features to solve problems for specialty markets; one example is SSDs for telemetry markets where one critical component is the NAND flash controller. This paper will focus on some of the requirement differences and why it is so important for SSD designers to develop their own controller for these markets.

L'intérêt porté au domaine Terahertz (THz) ayant beau être en pleine expansion depuis les années 1990, un gros effort de recherche doit encore être effectué pour tirer la quintessence des applications actuelles ou potentielles que représente cette gamme du spectre électromagnétique dans des domaines aussi variés que la spectroscopie, la cosmologie, l'imagerie médicale, la sécurité ou les télécommunications. En effet les sources, les détecteurs mais également les outils qui permettent d'amplifier ou de moduler un signal – dispositifs très présents dans les régions voisines du spectre électromagnétique que sont l'infrarouge et les micro-ondes - sont encore particulièrement limités par des facteurs tels que la compacité, la température de fonctionnement, l'intégrabilité mais également la puissance, la sensibilité ou encore le coût.Cette thèse porte sur l'étude expérimentale de divers composants en nitrure de gallium (GaN) contenant un puits quantique avec pour objectif de déterminer leurs capacités d'émission, d'amplification ou de détection d'une radiation THz.Pour ce faire, trois différents dispositifs expérimentaux ont été utilisés, améliorés ou même créés dans le but de pouvoir faire varier des paramètres tels que la polarisation électrique, leur température de fonctionnement, les fréquences THz sondées et bien sûr les différentes géométries des échantillons.De plus amples détails sur le monde des THz, sur les dispositifs électroniques GaN utilisés ainsi que sur les montages expérimentaux mis en places sont développés dans ce manuscrit de thèse. Les principaux résultats expérimentaux obtenus montrent :- une émission vers 3 THz avec une fréquence accordable en fonction du champ électrique appliqué au puits quantique GaN,- un coefficient de transmission variable en fonction de la tension appliquée aux contacts en doigts interdigités de différentes structures GaN,- la détection hétérodyne de radiations avec une fréquence RF de 0,3 THz et IF pouvant monter jusqu'à 40 GHz. De plus, chaque type de résultats expérimentaux a été expliqué théoriquement à l'aide de modèles analytiques développés en collaboration avec des équipes internationales au cours de ces trois dernières années. / Even if the interest upon the Terahertz (THz) domain is increasing since the 1990s, a strong research effort still needs to be done to get the most of the current and potential applications that this area of the electromagnetic spectrum has to offer in the various domains of spectroscopy, cosmology, medical imaging, security and telecommunications. Indeed, sources, detectors and even the tools that permits to amplify or modulate a signal – these devices are well developed in the neighboring regions of infrared and microwaves – are still particularly limited by characteristics like compactness, operating temperature, integrability but also power, sensitivity or cost.This thesis focuses on the experimental study of different gallium nitride (GaN) devices containing a quantum well. The main objective was to determine their capacities in emission, amplification or detection of a THz radiation.To do so, three different experimental setups where used, improved or even created in order to be able to change parameters like the electric bias, their working temperature, the probed THz frequencies and of course the different geometries of the samples.More details about the THz domain, the studied GaN electronic devices and the used experimental setups are developed in this PhD thesis.The main obtained experimental results show:- an emission of radiation near 3 THz with a tunable frequency versus electric field applied to the GaN quantum well,- a transmission coefficient variable as a function of the voltage applied to the contacts of different GaN interdigitated fingers structures,- heterodyne detection of radiation with a RF frequency of 0.3 THz and an IF that can reach up 40 GHz.In addition, each type of experimental results has been investigated theoretically using analytical models developed in collaboration with international teams during the past three years.

This dissertation presents novel approaches to the use of non-volatile memory devices in building storage systems. There are many types of non-volatile memory devices, and they usually have better performance than regular magnetic hard disks in terms of throughput and latency. This dissertation focused on two of them, NAND flash memory and Phase Change Memory (PCM). This work consisted of two parts. The first part was to design a high-performance hybrid storage system employing Solid State Drives that are build out of NAND flash memory and Hard Disk Drives. In this hybrid system, we proposed two different policies to improve its performance. One is to exploit the fact that the performances of Solid State Drive and Hard Disk Drive are asymmetric and the other is to exploit concurrency on multiple devices. We implemented prototypes in Linux and evaluate both policies in multiple workloads and multiple configurations. The results showed that the proposed approaches improve the performance significantly, and adapt to different configurations of the system under different workloads. The second part was to implement a file system on a special class of memory devices, Storage Class Memory (SCM), which is both byte addressable and also nonvolatile, e.g. PCM. We claimed that both the existing regular file systems and the memory based file systems are not suitable for SCM, and proposed a new file system, called SCMFS, which is implemented on the virtual address space. In SCMFS, we utilized the existing memory management module in the operating system to do the block management. Our design keeps address space within a file contiguous to reduce the block management software. The simplicity of SCMFS not only makes it easy to implement, but also improves the performance. We implemented a prototype of SCMFS in Linux and evaluated its performance through multiple benchmarks.

ITC/USA 2013 Conference Proceedings / The Forty-Ninth Annual International Telemetering Conference and Technical Exhibition / October 21-24, 2013 / Bally's Hotel & Convention Center, Las Vegas, NV / Solid-State Drives (SSDs) are an enabling technology for data recorders. SSDs can survive where Hard-Disk Drives (HDDs) cannot. SSDs deliver better performance with lower power consumption than HDDs. However, the end of Single-Level Cell (SLC) NAND flash may be near; Multi-Level Cell (MLC) NAND flash soon may be the only choice for industrial applications. System designers have two distinct concerns before implementing SSDs: 1. Cost: MLC NAND flash makes SSDs as affordable as HDDs 2. Endurance: SSDs are reliable and endurance assured with today's controller technology SSDs are leading the charge in transforming data storage in several applications, telemetry included.

Training prototxt file containing the hyperparameter settings for combinations 5 and 15 of optimized training runs.

Plots generated from training logs of combinations 5 and 15 of optimized training runs.

This thesis focuses on stochastic dominance in portfolio selection problems. The thesis recalls basic knowledge from the area of portfolio optimization with utility functions and first, second, $N$-th and infinite order of stochastic dominance. It sumarizes Post's, Kuosmanen's and Kopa's criteria for portfolio efficiency and necessary and sufficient conditions of stochastic dominance for discrete and continuous probability distributions. The thesis also contains formulations of optimization problems with second order stochastic dominance constraints derived for discrete and continuous probability distributions. A practical application is also a part of the thesis, where the optimization problems for monthly returns of Czech stocks are solved using optimization software GAMS.

Le domaine des fréquences térahertz (THz, 1 THz = 10^12 Hz) fait partie du spectre électromagnétique qui n'est pas encore sous contrôle. Il possède des propriétés physiques qui peuvent être exploitées dans des domaines très différents comme la spectroscopie, les télécommunications, l'imagerie, la sécurité ...L'objectif de ce projet de thèse était d'étudier et de réaliser un nouveau type de capteur THz basé sur des nanocanaux asymétriques de gaz d'électrons bidimensionnel également appelés Self-Switching Devices (SSD), c'est à dire dispositifs auto-commutants. Ces nouveaux dispositifs ont des caractéristiques I(V) non-linéaires leur permettant de fonctionner en tant que détecteurs quadratiques tout comme les diodes Schottky classiques.Les SSDs qui sont fabriqués par la création - dans une hétérojonction contenant un gaz d'électrons bidimensionnel - deux tranchées symétriques isolantes en forme de L disposées en tête bêchent, possèdent une haute mobilité d'électrons qui leur permet de travailler à des fréquences très élevées. Les contacts sont simplement réalisés par un dépôt métallique de chaque côté du dispositif et l'architecture complètement planaire permet facilement la réduction de leurs tailles et leur parallélisation. Ceci est en contraste avec la diode traditionnelle, et évidemment conduit à une réduction significative du coût de production. En outre, l'architecture extrêmement simple permet une très faible capacité parasite et donc une très grande vitesse de fonctionnement. À partir des simulations Monte Carlo, le dispositif devrait fonctionner également dans la gamme de fréquences THz, dans laquelle de très larges champs d'applications ont été démontrés.Nous avons développé deux bancs expérimentaux et démontré, dans une configuration quasi-optique, dans un espace libre et à température ambiante que les SSDs en nitrure de gallium (GaN) travaillent en tant que détecteurs directs par redressement et aussi en tant que détecteurs hétérodynes et ceux-ci jusqu'à 0,69 THz. Des réponses de 2 V/W et de 0,3 V/W avec des bandes passantes supérieures à 40 GHz et à 13 GHz ont été obtenues dans les gammes de fréquences de 0,30 et de 0,69 THz, respectivement. La caractérisation des SSDs en tant que mélangeurs ne montre aucune déviation de la linéarité entre la puissance de la porteuse THz (signal d'entrée RF) et celle du signal de sortie (IF) ayant une fréquence intermédiaire. Les simulations de Monte Carlo, utilisées pour estimer les pertes de conversion des nano-dispositifs de 27 dB à 0,69 THz, ont confirmé ces résultats.En conséquence, la mise en œuvre pratique des SSDs en tant que mélangeurs d'ondes submillimétriques de hautes puissances semble faisable. En outre, des études plus récentes de nanodispositifs similaires à base de GaN, ont démontré leurs possibilités d'agir comme oscillateurs Gunn qui pourraient être des éléments actifs dans des émetteurs THz. Par conséquence, les bonnes performances des nanocanaux en GaN démontrées dans cette thèse permettent non seulement l'intégration facile des antennes pour un meilleur couplage en espace libre, mais aussi (i) une flexibilité dans la conception pour une dissipation thermique optimale et la réduction des effets parasites et (ii) la possibilité de développer un système émetteur/détecteur d'ondes submillimétriques complètement intégré fonctionnant à température ambiante. / Terahertz frequency domain (THz, 1 THz = 10^12 Hz) is part of the electromagnetic spectrum that is not yet under control. It does have physical properties that concern very different fields such as spectroscopy, telecommunications, imaging, security...The project goal of this thesis was to study and to realize a new type of THz sensor based on asymmetrical nanochannels of two-dimensional electron gas (2DEG) also called Self‐Switching‐Devices (SSDs). These new devices have a non-linear I-V characteristics allowing them to operate as quadratic detectors like conventional Schottky diodes.SSDs manufactured by creating – into a heterojunction containing a two-dimensional electron gas — two symmetrical L-shaped insulating trenches arranged in head to tail, present high electron mobility that allows them to work at very high frequencies. Contacts are simply made by metal deposit on each side of the device and the completely planar architecture allows easy downscaling and parallelization. This is a great contrast with the traditional diode, and obviously leads to a significant reduction of the production cost. Moreover, the extremely simple architecture enables a very low parasitic capacitance and therefore a very high operation speed. From Monte Carlo simulations, the device is expected to operate also in the THz frequency range, in which very broad ranges of applications have been visualized.We have developed two experimental benches and demonstrated, in a quasi-optical configuration, free space and at room temperature that gallium nitride (GaN) based SSDs act as rectifying direct detectors and heterodyne detectors up to 0.69 THz. Responsivities of 2 V/W and 0.3 V/W with bandwidths greater than 40 GHz and 13 GHz were obtained in the 0.30 and 0.69 THz ranges, respectively. The characterization of the SSDs as mixers didn't show any deviation from linearity between the THz carrier (RF) input power and the power of the intermediate frequency (IF) signal output. Monte Carlo simulations, used to estimate nano-device intrinsic conversion losses of 27 dB at 0.69 THz, have confirmed these results. As a consequence, the practical implementation of SSDs as mixers of high-power sub-millimetre waves seems feasible. Moreover recent studies of similar GaN-based nanodevices, have demonstrated their possibilities to act as Gunn oscillators that could be active elements in THz emitters. Therefore, the good performances of GaN nanochannels demonstrated in this thesis allow not only the easy integration of antennas for a better free space coupling but also (i) a flexibility in the design of an optimum thermal dissipation and reduction of parasitic effects, and (ii) the possibility to develop a fully integrated emitter/detector submillimeter-wave system working at room temperature.

The current diverse and wide range of computing moves towards the cloud and de- mands high performance in low latency and high throughput. Facebook reported that 3.3 billion people monthly and 2.6 billion people daily use their data centers over the network. Many emerging user-facing applications require strict control over the stor- age latency’s tail to provide a quality user experience. The low-latency requirement triggers the ongoing replacement of hard drives (HDDs) by solid-state drives (SSDs) in the enterprise, enabling much higher performance and lower end-to-end storage latencies. It becomes more challenging to ensure low latency while maintaining the device’s endurance ratings. We address this challenge in the following ways: 1. Enhance the overall storage system’s performance and maintain the SSD endurance using emerging Non-volatile memory (ENVM) technology. 2. Implement deterministic la- tency in the storage path for latency-sensitive applications. 3. Provide low-latency and differentiated services when write-intensive workloads are present in a shared environment. We have proposed the performance and endurance-centric mechanisms to evaluate the tradeoffs between performance and endurance. In the first approach, our goal is to achieve low storage latency and a long lifetime of the SSD simultane- ously, even for a write-heavy workload. Incorporating a significantly smaller amount of ENVM with SSD as a cache helps to achieve the said goal.SSDs using the NVMe (Non-Volatile Memory Express) interface can achieve low latency as the interface provides several advanced features. The second approach has iii explored such features to control the storage tail latency in a distributed environment. The ”predictable latency mode (PLM)” advanced feature helps to achieve determinis- tic storage latency. SSDs need to perform many background management operations to deal with the underlying flash technology traits, the most time-consuming ones be- ing garbage collection and wear leveling. The latency requirement of latency-sensitive applications violates when the I/O requests fall behind such management activities. PLM leverages SSD controllers to perform the background operations during a win- dow, called a ”non-deterministic window (NDWin)”. Whereas during the ”determin- istic window (DTWin)”, applications will experience no such operations. We have extended this feature in the distributed environment and showed how it helps achieve low storage latency when the proposed ”PLM coordinator (PLMC)” is incorporated. In a shared environment with write-intensive workloads present, result in latency peak for Read IO. Moreover, it is required to provide differentiated services with multiple QoS classes present in the workload mixture. We have extended the PLM concept on hybrid storage to realize the deterministic latency for tight tail-controlled appli- cations and assure differentiated services among multiple QoS applications. Since nearly all of the storage access in a data center is over the network, an end-to-end path consists of three components: The host component, Network component, and Storage Component. For latency-sensitive applications, the overall tail latency needs to consider all these components. In a NAS (Network Attached Storage) architecture, it is worth studying the QoS class aware services present at the different components to provide an overall low request-response latency. Therefore, it helps future research to embrace the gaps that have not been considered yet. / Computer and Information Science