TProf: An energy profiler for task-parallel programs

We present TProf, an energy profiling tool for OpenMP-like task-parallel programs. To compute the energy consumed by each task in a parallel application, TProf dynamically traces the parallel execution and uses a novel technique to estimate the per-task energy consumption. To achieve this estimation, TProf apportions the total processor energy among cores and overcomes the limitation of current works which would otherwise make parallel accounting impossible to achieve. We demonstrate the value of TProf by characterizing a set of task parallel programs, where we find that data locality, memory access patterns and task working sets are responsible for significant variance in energy consumption between seemingly homogeneous tasks. In addition, we identify opportunities for fine-grain energy optimization by applying per-task Dynamic Voltage and Frequency Scaling (DVFS).

Current and Turbulence Measurement with Collocated ADP and Turbulence Profiler Data

This paper presents a current and turbulence measurement campaign conducted at a test site in an energetic tidal channel known as Strangford Narrows, Northern Ireland. The data was collected as part of the MaRINET project funded by the EU under their FP7 framework. It was a collaborative effort between Queen’s University Belfast, SCHOTTEL and Fraunhofer IWES. The site is highly turbulent with a strong shear flow. Longer term measurements of the flow regime were made using a bottom mounted Acoustic Doppler Profiler (ADP). During a specific turbulence measurement campaign, two collocated in- struments were used to measure incoming flow characteristics: an ADP (Aquadopp, Nortek) and a turbulence profiler (MicroRider, Rockland Scientific International). The instruments recorded the same incoming flow, so that direct comparisons between the data can be made. In this study the methodology adopted to deploy the instruments is presented. The resulting turbulence measurements using the different types of instrumentation are compared and the usefulness of each instrument for the relevant range of applications is discussed. The paper shows the ranges of the frequency spectra obtained using the different instruments, with the combined measurements providing insight into the structure of the turbulence across a wide range of scales.

A review of ADP processing software

Data processing is an essential part of Acoustic Doppler Profiler (ADP) surveys, which have become the standard tool in assessing flow characteristics at tidal power development sites. In most cases, further processing beyond the capabilities of the manufacturer provided software tools is required. These additional tasks are often implemented by every user in mathematical toolboxes like MATLAB, Octave or Python. This requires the transfer of the data from one system to another and thus increases the possibility of errors. The application of dedicated tools for visualisation of flow or geographic data is also often beneficial and a wide range of tools are freely available, though again problems arise from the necessity of transferring the data. Furthermore, almost exclusively PCs are supported directly by the ADP manufacturers, whereas small computing solutions like tablet computers, often running Android or Linux operating systems, seem better suited for online monitoring or data acquisition in field conditions. While many manufacturers offer support for developers, any solution is limited to a single device of a single manufacturer. A common data format for all ADP data would allow development of applications and quicker distribution of new post processing methodologies across the industry.

A sedimentologic and 14C dating study of five eastern Australian upper continental slope submarine landslides

Sedimentologic and AMS 14C age data are reported for calcareous hemipelagic mud samples taken from gravity cores collected at sites within, or adjacent to five submarine landslides identified with multibeam bathymetry data on the Nerrang Plateau segment and surrounding canyons of eastern Australia's continental slope (Bribie Bowl, Coolangatta-2, Coolangatta-1, Cudgen and Byron). Sediments are comprised of mixtures of calcareous and terrigenous clay (10-20%), silt (50-65%) and sand (15-40%) and are generally uniform in appearance. Their carbonate contents vary between and 17% and 22% by weight while organic carbon contents are less than 10% by weight. Dating of conformably deposited material identified in ten of the twelve cores indicates a range of sediment accumulation rates between 0.017mka-1 and 0.2 mka-1 which are consistent with previous estimates reported for this area. One slide-adjacent core, and four within-landslide cores present depositional hiatus surfaces located at depths of 0.8 to 2.2 meters below the present-day seafloor and identified by a sharp, colour-change boundary; discernable but small increases in sediment stiffness; and a slight increase in sediment bulk density of 0.1 gcm-3. Distinct gaps in AMS 14C age of at least 20ka are recorded across these boundary surfaces. Examination of sub-bottom profiler records of transects through three of the within-slide core-sites and their nearby landslide scarps available for the Coolangatta-1 and Cudgen slides indicate that: 1) the youngest identifiable sediment layer reflectors upslope of these slides, terminate on and are truncated by slide rupture surfaces; and 2) there is no obvious evidence in the sub-bottom profiles for a post-slide sediment layer draped over or otherwise burying slide ruptures or exposed slide detachment surfaces. This suggests that both these submarine landslides are geologically recent and suggests that the hiatus surfaces identified in Coolangatta-1's and Cudgen's within-slide cores are either: a) erosional features that developed after the occurrence of the landslide in which case the hiatus surface age provides a minimum age for landslide occurrence or b) detachment surfaces from which slabs of near-surface sediment were removed during landsliding in which case the post-hiatus sediment dates indicates approximately when landsliding occurred. In either case, it is reasonable to suggest that these two spatially adjacent slides occurred penecontemporaneously approximately 20,000 years ago.