Infiniband wird zunehmend als Verbindungsnetzwerk für Cluster eingesetzt. Dadurch wird es nötig existierende Bibliotheken für parallele Programmiersprachen an das neue Netzwerk bestmöglich anzupassen. Ein wichtiger Bestandteil paralleler Programmiersprachen sind dabei kollektive Operationen, die es erfordern, eine Nachricht von einem Knoten zu vielen anderen oder auch von vielen Knoten an einen einzelnen zu senden. Um herauszufinden, welche Verbindungsarten und Operationen am besten für diese kollektiven Operationen geeignet sind, wurde ein Benchmark entwickelt. Ziel dieser Studienarbeit ist es, dieses Programm zu erweitern, auf einem Cluster zu testen und die Ergebnisse auszuwerten.

This report describes PlaceMe, a client side, mobile application built on the Android platform that provides personal location-based services such as location reminders, bookmarking, mapping and search nearby. The reminder system allows creating location based reminders, and alerts the user what he needs to do, when he is in the right place to do them. Bookmarking allows the user to virtually “save” places of interest while he is on the move and obtain driving directions. Mapping enables the user to visualize his relative geographic location in real time, and map the location reminders and bookmarks. Finally, search nearby exploits Google’s powerful local search engine to allow finding and bookmarking nearby places such as gas stations, restaurants, etc, and retrieving map-based directions. We first discuss the requirements and use-cases for PlaceMe, followed an introduction to the Android software stack. Next, we describe our design architecture, implementation model, test strategy and key performance enhancements. Then, we evaluate and compare the performance of the Android platform across a set of standard micro-benchmarks. Finally, we conclude with a discussion of future development ideas and present our thoughts on prospects of app-based mobile computing. / text

The current trends in computer architecture are shifting towards smaller byte/flop ratios, while available parallelism is increasing at all levels of granularity – vector length, core count, and MPI process. Intel’s Xeon Phi coprocessor, NVIDIA’s Kepler GPU, and IBM’s BlueGene/Q all have a Byte/flop ratio close to 0.2, which makes it very difficult for most algorithms to extract a high percentage of the theoretical peak flop/s from these architectures. Popular algorithms in scientific computing such as FFT are continuously evolving to keep up with this trend in hardware. In the meantime it is also necessary to invest in novel algorithms that are more suitable for computer architectures of the future. The fast multipole method (FMM) was originally developed as a fast algorithm for ap- proximating the N-body interactions that appear in astrophysics, molecular dynamics, and vortex based fluid dynamics simulations. The FMM possesses have a unique combination of being an efficient O(N) algorithm, while having an operational intensity that is higher than a matrix-matrix multiplication. In fact, the FMM can reduce the requirement of Byte/flop to around 0.01, which means that it will remain compute bound until 2020 even if the cur- rent trend in microprocessors continues. Despite these advantages, there have not been any benchmarks of FMM codes on modern architectures such as Xeon Phi, Kepler, and Blue- Gene/Q. This study aims to provide a comprehensive benchmark of a state of the art FMM code “exaFMM” on the latest architectures, in hopes of providing a useful reference for deciding when the FMM will become useful as the computational engine in a given application code. It may also serve as a warning to certain problem size domains areas where the FMM will exhibit insignificant performance improvements. Such issues depend strongly on the asymptotic constants rather than the asymptotics themselves, and therefore are strongly implementation and hardware dependent. The primary objective of this study is to provide these constants on various computer architectures.

This thesis examines the extent to which benchmark beating by Australian firms around the earnings level and earnings changes thresholds can be reliably interpreted as evidence of earnings management. A number of recent academic papers challenge the earnings management explanation for the observed kinks in the distribution of net Income. In response to this criticisms, this thesis is motivated to conduct tests of earnings management with a refined methodology of selecting a subset of firms immediately above the threshold that have a priori incentives to achieve the benchmark. This approach allows for investigations to focus on benchmark beating observations where earnings manipulations would be more prevalent and thereby provide a powerful test for the existence of opportunistic reporting. The paper uses a number of unexpected accruals measures including the Kothari et al. (2005) performance matched models. In testing the hypotheses, this thesis utilises two approaches which were; the regression approach and the test of difference of means approach. Based on a broad sample drawn from all listed Australian firms for the years 1995-2007, small profit firms and small increase firms with high price-to-sales ratio were found to have evidence consistent with opportunistic benchmark beating behaviour. Similar results are also documented for benchmark beating firms with low book-to-market (high market-to-book) ratio. This thesis also finds that firms with equity offering incentives who reported improvement in earnings display unexpected accruals consistent with earnings management. In addition, the accounting behaviour of firms which previously incurred a loss is consistent with earnings management explanation. Firms with long strings of earnings increases also appear to use accounting discretion in order to avoid earnings deterioration. Similarly, evidence of earnings management are also displayed by small profit firms which have consistently reported negative earnings. Finally, this thesis provides evidence that resolves the apparent paradox that benchmark beating is evidence of earnings management which is devoid of the statistical artefact argument posited by Durtschi and Easton (2005) and Durtschi and Easton (2008).

Thesis (doctoral)--Universität Karlsruhe, 2007.

Zugl.: Hannover, Univ., Diss., 2006

Zugl.: Rostock, Univ., Diss., 2009

Zugl.: Duisburg, Essen, Univ., Diss., 2009

Techn. Hochsch., Diss., 2004--Aachen.

There is an emerging demand on efficiently archiving and (temporal) querying different versions of evolving semantic Web data. As novel archiving systems are starting to address this challenge, foundations/standards for benchmarking RDF archives are needed to evaluate its storage space efficiency and the performance of different retrieval operations. To this end, we provide theoretical foundations on the design of data and queries to evaluate emerging RDF archiving systems. Then, we instantiate these foundations along a concrete set of queries on the basis of a real-world evolving dataset. Finally, we perform an empirical evaluation of various current archiving techniques and querying strategies on this data that is meant to serve as a baseline of future developments on querying archives of evolving RDF data.