Distributed computing of all-to-all comparison problems in heterogeneous systems

The requirement of distributed computing of all-to-all comparison (ATAC) problems in heterogeneous systems is increasingly important in various domains. Though Hadoop-based solutions are widely used, they are inefficient for the ATAC pattern, which is fundamentally different from the MapReduce pattern for which Hadoop is designed. They exhibit poor data locality and unbalanced allocation of comparison tasks, particularly in heterogeneous systems. The results in massive data movement at runtime and ineffective utilization of computing resources, affecting the overall computing performance significantly. To address these problems, a scalable and efficient data and task distribution strategy is presented in this paper for processing large-scale ATAC problems in heterogeneous systems. It not only saves storage space but also achieves load balancing and good data locality for all comparison tasks. Experiments of bioinformatics examples show that about 89\% of the ideal performance capacity of the multiple machines have be achieved through using the approach presented in this paper.

Experimental studies on the performance of rail joints with modified wheel/railhead contact

Rail joints are provided with a gap to account for thermal movement and to maintain electrical insulation for the control of signals and/or broken rail detection circuits. The gap in the rail joint is regarded as a source of significant problems for the rail industry since it leads to a very short rail service life compared with other track components due to the various, and difficult to predict, failure modes – thus increasing the risk for train operations. Many attempts to improve the life of rail joints have led to a large number of patents around the world; notable attempts include strengthening through larger-sized joint bars, an increased number of bolts and the use of high yield materials. Unfortunately, no design to date has shown the ability to prolong the life of the rail joints to values close to those for continuously welded rail (CWR). This paper reports the results of a fundamental study that has revealed that the wheel contact at the free edge of the railhead is a major problem since it generates a singularity in the contact pressure and railhead stresses. A design was therefore developed using an optimisation framework that prevents wheel contact at the railhead edge. Finite element modelling of the design has shown that the contact pressure and railhead stress singularities are eliminated, thus increasing the potential to work as effectively as a CWR that does not have a geometric gap. An experimental validation of the finite element results is presented through an innovative non-contact measurement of strains. Some practical issues related to grinding rails to the optimal design are also discussed.

Micro inverter with a front-end current-fed converter

The reliability of micro inverters is an important factor as it would be necessary to reduce cost and maintenance of the small and medium scale distributed PV power conversion systems. Electrolytic capacitors and active power decouple circuits can be avoided in micro inverters with the use of medium voltage DC-link. Such a DC-link based micro inverter is proposed with a front-end dual inductor current-fed push-pull converter. The primary side power switches of the front-end converter have reduced switching losses due to multi-resonant operation. In addition, the voltage and current stresses on the diodes of the secondary diode voltage doubler rectifier are reduced due to the presence of a series resonant circuit in the front-end converter. The operation of the proposed micro inverter is explained using an in-depth analysis of the switching characteristics of the power semiconductor devices. The theoretical analysis of the proposed micro inverter is validated using simulation result.

Targeting glucocorticoid receptors that promote resilience in the treatment of addiction

There is strong evidence to suggest that the combination of alcohol and chronic repetitive stress leads to long-lasting effects on brain function, specifically areas associated with stress, motivation and decision-making such as the amygdala, nucleus accumbens and prefrontal cortex. Alcohol and stress together facilitate the imprinting of long-lasting memories. The molecular mechanisms and circuits involved are being studied but are not fully understood. Current evidence suggests that corticosterone (animals) or cortisol (humans), in addition to direct transcriptional effects on the genome, can directly regulate pre- and postsynaptic synaptic transmission through membrane bound glucocorticoid receptors (GR). Indeed, corticosterone-sensitive synaptic receptors may be critical sites for stress regulation of synaptic responses. Direct modulation of synaptic transmission by corticosterone may contribute to the regulation of synaptic plasticity and memory during stress (Johnson et al., 2005; Prager et al., 2010). Specifically, previous data has shown that long term alcohol (1) increases the expression of NR2Bcontaining NMDA receptors at glutamate synapses, (2) changes receptor density, and (3) changes morphology of dendritic spines (Prendergast and Mulholland; 2012). During alcohol withdrawal these changes are associated with increased glucocorticoid signalling and increased neuronal excitability. It has therefore been proposed that these synapse changes lead to the anxiety and alcohol craving associated with withdrawal (Prendergast and Mulholland; 2012). My lab is targeting this receptor system and the amygdala in order to understand the effect of combining alcohol and stress on these pathways. Lastly, we are testing GR specific compounds as potential new medications to promote the development of resilience to developing addiction.

Improving factory performance using an integrated sugar factory model

A whole of factory model of a raw sugar factory was developed in SysCAD software to assess and improve factory operations. The integrated sugar factory model ‘Sugar-SysCAD’ includes individual models for milling, heating and clarification, evaporation, crystallisation, steam cycle, sugar dryer and process and injection water circuits. These individual unit operation models can be either used as standalone models to optimise the unit operation or in the integrated mode to provide more accurate prediction of the effects of changes in any part of the process on the outputs of the whole factory process. Using the integrated sugar factory model, the effect of specific process operations can be understood and practical solutions can be determined to address process problems. The paper presents two factory scenarios to show the capabilities of the whole of factory model.

Implementation of parallel tridiagonal solvers for a heterogeneous computing environment

Tridiagonal diagonally dominant linear systems arise in many scientific and engineering applications. The standard Thomas algorithm for solving such systems is inherently serial forming a bottleneck in computation. Algorithms such as cyclic reduction and SPIKE reduce a single large tridiagonal system into multiple small independent systems which can be solved in parallel. We have developed portable cyclic reduction and SPIKE algorithm OpenCL implementations with the intent to target a range of co-processors in a heterogeneous computing environment including Field Programmable Gate Arrays (FPGAs), Graphics Processing Units (GPUs) and other multi-core processors. In this paper, we evaluate these designs in the context of solver performance, resource efficiency and numerical accuracy.

The interaction of quantum dots with plasmons supported by metal waveguides

Plasmonics is a recently emerged technology that enables the compression of electromagnetic waves into miniscule metallic structures, thus enabling the focusing and routing of light on the nanoscale. Plasmonic waveguides can be used to miniaturise the size of integrated chip circuits while increasing the data transmission speed. Plasmonic waveguides are used to route the plasmons around a circuit and are a major focus of this thesis. Also, plasmons are highly sensitive to the surrounding dielectric environment. Using this property we have experimentally realised a refractive index sensor to detect refractive index change in solutions.