Les processeurs multi-coeurs sont maintenant largement utilisés presque partout en informatique: ordinateurs de bureau, ordinateurs portables et accélérateurs tels que les GPGPU (General Purpose Graphics Processing Units). La difficulté de la programmation des systèmes parallèles est considérée comme un problème majeur qui va empêcher l'exploitation de leurs capacités dans le futur. Pour exploiter la puissance des processeurs multi-coeurs et les hiérarchies complexes de mémoire, il y a une grande nécessité pour utiliser des outils de parallélisation et d'optimisation automatique de code. L'optimisation polyédrique est un axe de recherche qui a comme but de résoudre ces problèmes. C'est est une représentation algébrique du programme et un ensemble d'analyses, de transformations et d'algorithmes de génération de code qui permettent à un compilateur de raisonner sur des transformations avancées de nids de boucle. Dans cette thèse, nous abordons certaines des limites du modèle polyédrique. Nous nous intéréssons particulièrement à trois problèmes et nous proposons des solutions pratiques à ces trois problèmes. Le premier problème est lié à la capacité d'appliquer l'optimisation de tuilage sur un code qui contient des fausses dépendances. Nous proposons une téchnique qui permet d'ignorer certaines fausses dépendences et donc qui permet d'appliquer l'optimisation de tuilage qui n'est pas possible sinon. Le second problème est lié au temps de compilation qui peut être trés long pour certains programmes. Nous proposons une téchnique qui transforme la représentation originale du programme à une nouvelle representation dans laquelle il y a moins d'instructions. L'optimisation de cette nouvelle représentation du programme est moins couteuse en terme de temps de compilation en comparaison avec l'optimisation de la représentation originale du programme. Le troisième problème est lié à deux limites: la première limite concerne la possibilité d'utiliser la compilation polyédrique sur des programmes qui ne resepectent pas les restrictions classiques du modèle polyédrique (un programme peut être représenté de façon précise dans le modèle polyédrique s'il ne contient pas des conditionnelles non-affines, des bornes de boucles non-affines et des accés non-affines). La seconde limite est liée à l'aptitude des outils à générer un code performant dans les performances se rapprochent des performances du code écrit à la main. Pour éviter ces deux limites, nous proposons un language de programmation que l'on appelle PENCIL, c'est un sous-ensemble de GNU C99 avec des règles de programmation spécifiques et quelques extensions. L'utilisation de ce sous-ensemble et l'utilisation de ces extensions permettent aux compilateurs de mieux exploiter le parallélisme et de mieux optimiser le code. / Multi-core processors are now in widespread use in almost all areas of computing: desktops, laptops and accelerators such as GPGPUs (General Purpose Graphics Processing Units). To harness the power of multi-core processors and complex memory hierarchies, the need for powerful compiler optimizations and especially loop nest transformations is now in high demand. The polyhedral optimization framework is showing promising results in addressing such a problem. It's an algebraic program representation and a set of analyses, transformations and code generation algorithms that enable a compiler to reason about advanced loop nest transformations addressing most of the parallelism and locality-enhancing challenges.In this thesis we address some of the limitations of the polyhedral framework. We address three problems and propose practical solutions to these three problems.The first problem is related to the ability to apply tiling on code that has false dependences (loop nest tiling is an optimization that changes the order of execution of statements in a loop nest in order to enhance data locality; false dependences are induced by the reuse of a single memory location to store multiple values during the life of the program). To preserve the validity of loop nest transformations and parallelization, data-dependences need to be analyzed. Memory dependences come in two varieties: true dependences (a.k.a. flow dependences) and false dependences (a.k.a. output and anti dependences). While true dependences must be satisfied in order to preserve the correct order of computations. False dependences reduce the degrees of freedom for loop transformations. In particular, loop tiling is severely limited in the presence of these dependences. While array expansion, a transformation that transforms scalars into arrays and arrays into higher dimensional arrays, removes all false dependences, the overhead of this transformation on memory and the detrimental impact on register-level reuse can be catastrophic. We propose and evaluate a compilation technique to safely ignore a large number of false dependences in order to enable loop nest tiling in the polyhedral model. It is based on the precise characterization of interferences between live range intervals, and it does not incur any scalar or array expansion.The second problem is related to the long compilation time that one may experience when using polyhedral tools to optimize a program. Particularly, the long execution time of the Pluto affine scheduling algorithm. The Pluto affine scheduling algorithm is the algorithm that is responsible for changing the schedule (order of execution) of statements in order to optimize the code (maximize parallelism and data locality). Reducing the execution time of this affine scheduling algorithm enhances the overall compilation time. We introduce and evaluate a technique called offline statement clustering. It is a practical technique designed to reduce the execution time of the Pluto affine scheduling algorithm without much loss in optimization opportunities. Using this technique, the statements of the program are clustered into macro-statements, the Pluto affine scheduling algorithm is then used to schedule the macro-statements instead of scheduling the original statements of the program. Since the number of macro-statements is less than the number of statements in the original program, scheduling the macro-statements is in general faster than scheduling the original statements of the program. We present the statement clustering algorithm, we show how offline statement clustering integrates transparently with the work-flow of a state-of-the-art polyhedral compiler and present two heuristics for choosing how statements should be clustered together. We show experimentally that statement clustering can reduce the scheduling time by a factor of 8x (in median) without a significant loss in optimization opportunities...

No description available.

Compared to other instruments for chemical analyses, electrochemical analyzers are relatively simple, inexpensive, easy to miniaturize and require little-to-no maintenance. However, like all commercially available instruments, commercial electrochemical analyzers like potentiostats primarily operate as black boxes with manufacturers providing little or no information about internal circuitry and programming. This practice can limit a researcher’s ability to develop new techniques or adapt an instrument for applications outside its typical use. In contrast, creators of open-source instruments release all the necessary information for reproduction of the hardware and software – minimizing such barriers to innovation in chemical analyses. Here, we report a low-cost, compact potentiostat based on an open-source Arduino microcontroller capable of performing electrochemical analyses such as cyclic and linear sweep voltammetry with an operating range of ± 208 𝜇A and ± 2.5 V. Performance of the potentiostat is investigated with low-cost pencil graphite electrodes and compared to a commercial potentiostat.

Microfluidic paper based analytical devices (microPADs) are a novel platform for point of care (POC) diagnostics. Limitations of reagent shelf life have been overcome with the introduction of reagent pencils as a method for solid-based reagent deposition. While useful, little work has been reported on the characterization and optimization of reagent pencils. Herein, an investigation on reagent pencil composition and efficiency is conducted via colorimetric release profile tests utilizing an erioglaucine disodium salt that yields a quantifiable blue colored product in the presence of water. Within this work, an investigation on the molecular weight dependence, polymer chain end functionality, and polymer-graphite ratio was conducted to determine the most desirable parameters in reagent pencil composition. Further, the effects of enzyme stability in the presence of poly(ethylene glycol) (PEG) is investigated. To show the versatility of reagent pencils, a novel reagent pencil incorporating a stimuli responsive polymer, poly(N-isporopylacrylamide) (PNIPAM) was developed. In this work, PNIPAM’s lower critical solution temperature (LCST) was manipulated with various salt solutions to control fluid flow both laterally and vertically through various microPAD designs. It was found that, while PNIPAM successfully blocked or retarded fluid flow in microPADs, the effect was limited when DI H2O wash solutions were run prior to salt solutions. To counteract this, PNIPAM was successfully covalently bound to alkene modified chromatography paper via thiolene click chemistry to reinforce solution wash tolerance.

Spaceborne scatterometer instruments are important tools for the remote sensing of the Earth's environment. In addition to the primary goal of measuring ocean winds, data from scatterometers have proven useful in the study of a variety of land and cryopshere processes as well. Several satellites carrying scatterometers have flown in the last two decades. These previous systems have been "fan-beam" scatterometers, where multiple antennas placed in fixed positions are used. The fan-beam scatterometer approach, however, has disadvantages which limit its utility for future missions. An alternate approach, the conically-scanning "pencil-beam" scatterometer technique, alleviates many of the problems encountered with earlier systems and provides additional measurement capability. Due to these advantages, the pencil-beam approach has been selected by NASA as the basis for future scatterometer missions. Whereas the fan-beam approach is mature and well understood, there is need for a fundamental study of the unique aspects of the pencil-beam technique. In this dissertation, a comprehensive treatment of the design issues associated with pencil-beam scatterometers is presented. A new methodology is established for evaluating and optimizing the performance of conically-scanning radar systems. Employing this methodology, key results are developed and used in the design of the SeaWinds instrument - NASA's first pencil-beam scatterometer. Further, the theoretical framework presented in this study is used to propose new scatterometer techniques which will significantly improve the spatial resolution and measurement accuracy of future instruments.

Electrochemical measurements are utilized in various fields, including healthcare (e.g., potentiometric measurements for electrolytes in blood and blood gas, amperometric biosensing of glucose as in blood glucose meters), water quality (e.g., pH measurement, voltammetric analyses for heavy metals), and energy. Much of the appeal of electrochemical analyses can be attributed to the relative simplicity, low cost and lack of maintenance associated with electrochemical instruments, along with techniques that can exhibit high sensitivity and selectivity, wide linear dynamic range, and low limits of detection for many analytes. While commercial electrochemical analyzers are less expensive than many other instruments for chemical analyses and are available from various manufacturers, versatility and performance often coincide with added expense. Recently, the development of low-cost, adaptable, open-source chemical instruments, including electrochemical analyzers, has emerged as a topic of great interest in the scientific community. In contrast to commercial instruments, for which schematics and underlying operation details are often obscured – severely limiting modifications and improvements, creators of open-source instruments release all the necessary information for reproduction of the hardware and software. As a result, open-source instruments not only serve as excellent teaching tools for novices to gain experience in electronics and programming, but also present opportunity to design and develop low-cost, portable instruments, which have particular significance for point-of-care sensing applications, use in resource-limited settings, and the rapidly developing field of on-body sensors. In this work, we report the design of a low-cost, compact electrochemical analyzer based on an open-source Arduino microcontroller. The instrument is capable of performing electrochemical analyses such as cyclic and linear sweep voltammetry with an operating range of ± 138 ��A and ± 1.65 V. Performance of the platform is investigated with low-cost pencil graphite electrodes and results compared to commercial potentiostats.

Abstract Electrochemical sensors are simple, fast, accurate, and low-cost analytical devices. They are especially important to the field of healthcare since they enable measurement of important indicators of patient health such as electrolytes and glucose in blood. Continued development and improvements in electrochemical sensors can result in more accessible, affordable, and effective diagnoses and treatment strategies. Electrochemical sensors employ electrodes, usually modified with a recognition agent specific for the analyte (the biomolecule of interest). The presence of the analyte at the electrode surface is typically measured through an electrochemical reaction that generates a signal in the form of an electric current or difference in electric potential. As an alternative, electrochemiluminescence, a phenomenon whereby an electrochemical reaction generates a product in an electronically excited state that is capable of emitting light, has great benefits due to its high sensitivity, selectivity, and extremely low background signal. Here we employ a camera equipped with a complementary metal-oxide semiconductor (CMOS) detector that is interfaced with a simple zoom lens to measure ECL generated at low-cost pencil graphite electrodes and small electrode arrays using tris(2,2′- bipyridyl) dichlororuthenium(II) hexahydrate ([Ru(bpy)3]2+) with tri-n-propylamine (TPA) as the coreactant. ECL signals produced at pencil graphite working electrodes were linear with respect to [Ru(bpy)3]2+ concentrations for 45–450μM [Ru(bpy)3]2+. The detection limit was found to be 2µM using the CMOS camera with exposure time set at 10s. This proof-of-concept work suggests the pencil graphite electrode with simple imaging system platform can be applied for ECL-based biosensing strategies.

3-D paper-based microfluidic devices (micoPADs) are small and portable devices made out of paper that offer a promising platform for diagnostic applications outside of a laboratory. These devices are easy to use, low cost, require no power source, and capable of detecting multiple targets simultaneously. The work in this thesis demonstrated the ability of a 3-D paper-microfluidic platform to simultaneously detect 5 targets. Rubber cord stock was used in conjunction with an acrylic housing unit to apply pressure along the edge of the channel. The indirect pressure application was successful in promoting vertical fluid flow between layers. Average channel development times were recorded between 110 seconds and 150 seconds. The implementation of the 3-D paper-microfluidic platform as a diagnostic device was validated with a colorimetric glucose assay. In a novel application, reagents were deposited onto the 3-D platform via a glucose reagent pencil created by Martinez et al. A visual signal was observed for the successful detection of glucose at a concentration of 1.2 mM. These results offer promise for future work in combing new reagent deposition techniques with a multi-layer paper-microfluidic platform. Overall, this research made advancements in the design of a paper-microfluidic platform capable of the simultaneous detection of 5 targets.

Electrochemiluminescence (ECL) is a phenomenon whereby electrochemical reactions generate a product that is capable of emitting light. ECL’s high sensitivity, selectivity, extremely low background, and relatively simple instrumentation make it particularly well-suited for chemical sensing and biosensing strategies. Here we report a simple ECL imaging system based on a camera interfaced with a zoom lens to compare pencil graphite electrode (PGE) and screen-printed carbon electrode (SPCE) arrays as ECL platforms. With this system, ECL signals generated from tris(2,2′- bipyridine)ruthenium(II) chloride hexahydrate ([Ru(bpy)3]2+) using co-reactant tri-n-propylamine (TPA) were linear with respect to [Ru(bpy)3]2+ concentrations from 9 to 450 μM. Detection limits for [Ru(bpy)3]2+ were found to be 1.8 μM with PGEs and 0.9 μM with SPCEs. Immobilization of a thin polyvinylpyridine (PVP) film ECL reporter [Ru(bpy)2(PVP)10]2+ on SPCEs was also investigated. Overall, the combination of PGEs or SPCEs with the simple ECL imaging system offers a cost-effective approach to ECL-based sensing and biosensing.

No description available.