Recent increases in hard fault rates in modern chip multi-processors have led to a variety of approaches to try and save manufacturing yield. Among these are: fine-grain fault tolerance (such as error correction coding, redundant cache lines, and redundant functional units), and large-grain fault tolerance (such as disabling of faulty cores, adding extra cores, and core salvaging techniques). This paper considers the case of core salvaging techniques and the heterogeneous per- formance introduced when these techniques have some salvaged and some non-faulty cores. It proposes a hypervisor-based hardware thread scheduler, triggered by detection of spin locks and thread imbalance, that mitigates the loss of throughput resulting from this het- erogeneity. Specifically, a new algorithm, called Most ProgressMade algorithm, reduces the number of synchronization locks held on a salvaged core and balances the time each thread in an application spends running on that core. For some benchmarks, the results show as much as a 2.68x increase in performance over a salvaged chip multi-processor without this technique.

No description available.

Particulate contamination of the optical surfaces of solar collectors, often called "soiling", can have a significant deteriorating impact on energy yield due to the absorption and scattering of incident light. Soiling has more destructive effect on concentrated solar systems than on flat-plate photovoltaic panels, as the former are incapable of converting scattered sunlight. The first part of this thesis deals with the soiling losses of flat-plate photovoltaic (PV), concentrated solar power (CSP), and concentrated photovoltaic (CPV) systems in operation in several regions of the world. Influential parameters in dust accumulation losses, as well as different cleaning mechanisms in pursuit of restoring the efficiency of soiled systems, have been thoroughly investigated. In lieu of the most commonly-practiced manual cleaning method of using high-pressure water jets, the concept of automatic dust removal using the electrostatic forces of electrodynamic screen (EDS) technology is in a developmental stage and on its way toward commercialization. This thesis provides comprehensive analytical solutions for the electric potential and electric field distribution in EDS devices having different configurations. Numerical simulations developed using finite element analysis (FEA) software have corroborated the analytical solutions which can easily be embedded into software programs for particle trajectory simulations while also providing flexibility and generality in the study on the effect of different parameters of the EDS on the electric field and ensuing dust-removal performance. Evaluation and comparison of different repelling and attracting forces exerted on dust particles is of utmost importance to a detailed analysis of EDS performance in dust removal. Hence, the balance of electrostatic and adhesion forces, including van der Waals and capillary forces, have received significant attention in this dissertation. Furthermore, different numerical analyses have been conducted to investigate the potential causes of observed failures of EDS prototypes that functioned well in a laboratory environment but failed after outdoor exposure. Experimental studies form the last two chapters of this dissertation. Different tests have been conducted on an EDS sample integrated with a PV cell to restore the efficiency of the cell after dust deposition. In order to evaluate the performance of the EDS in dust-particle removal, we have studied the particle size distribution on the EDS surface after each dust deposition and EDS cleaning cycle using a custom-built dust-deposition analyzer. Furthermore, we have pursued several experiments to examine how the geometric and operational EDS parameters affect particle charge via charge-to-mass-ratio measurements.