Behavior-Based branch prediction by dynamically clustering branch instructions

Conditional branches frequently exhibit similar behavior (bias, time-varying behavior,...), a property that can be used to improve branch prediction accuracy. Branch clustering constructs groups or clusters of branches with similar behavior and applies different branch prediction techniques to each branch cluster. We revisit the topic of branch clustering with the aim of generalizing branch clustering. We investigate several methods to measure cluster information, with the most effective the storage of information in the branch target buffer. Also, we investigate alternative methods of using the branch cluster identification in the branch predictor. By these improvements we arrive at a branch clustering technique that obtains higher accuracy than previous approaches presented in the literature for the gshare predictor. Furthermore, we evaluate our branch clustering technique in a wide range of predictors to show the general applicability of the method. Branch clustering improves the accuracy of the local history (PAg) predictor, the path-based perceptron and the PPM-like predictor, one of the 2004 CBP finalists.

Unified: The Musical by Guy Le Jeune

A large-scale community play based on interviews with former workers at the Unifi textile plant in Letterkenny, Co. Donegal

The Holocene British and Irish ancient woodland fossil beetle fauna: implications for forest history, biodiversity and faunal colonisation

This paper presents a new review of our knowledge of the ancient forest beetle fauna from Holocene archaeological and palaeoecological sites in Great Britain and Ireland. It examines the colonisation, dispersal and decline of beetle species, highlighting the scale and nature of human activities in the shaping of the landscape of the British Isles. In particular, the paper discusses effects upon the insect fauna, and examines in detail the fossil record from the Humberhead Levels, eastern England. It discusses the local extirpation of up to 40 species in Britain and 15 species in Ireland. An evaluation of the timing of extirpations is made, suggesting that many species in Britain disappear from the fossil record between c. 3000 cal BC and 1000 cal BC (c. 5000-3000 cal BP), although some taxa may well have survived until considerably later. In Ireland, there are two distinct trends, with a group of species which seem to be absent after c. 2000 cal BC (c. 4000 cal BP) and a further group which survives until at least as late as the medieval period. The final clearance of the Irish landscape over the last few hundred years was so dramatic, however, that some species which are not especially unusual in a British context were decimated. Reasons behind the extirpation of taxa are examined in detail, and include a combination of forest clearance and human activities, isolation of populations, lack of temporal continuity of habitats, edaphic and competition factors affecting distribution of host trees (particularly pine), lack of forest fires and a decline in open forest systems. The role of climate change in extirpations is also evaluated. Consideration is given to the significance of these specialised ancient forest inhabitants in Ireland in the absence of an early Holocene land-bridge which suggests that colonisation was aided by other mechanisms, such as human activities and wood-rafting. Finally, the paper discusses the Continental origins of the British and Irish fauna and its hosts and the role played by European glacial refugia.

Holocene deposits from Neptune’s Cave, Nordland, Norway: Environmental interpretation and relation to the deglacial and emergence history of the Velfjord–Tosenfjord area

Neptune’s Cave in the Velfjord–Tosenfjord area of Nordland, Norway is described, together with its various organic deposits. Samples of attached barnacles, loose marine molluscs, animal bones and organic sediments were dated, with radiocarbon ages of 9840+/-90 and 9570+/-80 yr BP being derived for the barnacles and molluscs, based on the superseded but locally used marine reservoir age of 440 years. A growth temperature of c. 7.51C in undiluted seawater is deduced from the d13C and d18O values of both types of marine shell, which is consistent with their early Holocene age. From the dates, and an assessment of local Holocene uplift and Weichselian deglaciation, a scenario is constructed that could explain the situation and condition of the various deposits. The analysis uses assumed local isobases and sea-level curve to give results: that are consistent with previous data, that equate the demise of the barnacles to the collapse of a tidewater glacier in Tosenfjord, and that constrain the minimum extent of local Holocene uplift. An elk fell into the cave in the mid-Holocene at 510070 yr BP, after which a much later single ‘bog-burst’ event at 178070 yr BP could explain the transport of the various loose deposits further into the cave.

Late Pleistocene history of turbidite sedimentation in a submarine canyon off the northern Great Barrier Reef, Australia

Cores from slopes east of the Great Barrier Reef (GBR) challenge traditional models for sedimentation on tropical mixed siliciclastic-carbonate margins. However, satisfactory explanations of sediment accumulation on this archetypal margin that include both hemipelagic and turbidite sedimentation remain elusive, as submarine canyons and their role in delivering coarse-grained turbidite deposits, are poorly understood. Towards addressing this problem we investigated the shelf and canyon system bordering the northern Ribbon Reefs and reconstructed the history of turbidite deposition since the Late Pleistocene. High-resolution bathymetric and seismic data show a large paleo-channel system that crosses the shelf before connecting with the canyons via the inter-reef passages between the Ribbon Reefs. High-resolution bathymetry of the canyon axis reveals a complex and active system of channels, sand waves, and local submarine landslides. Multi-proxy examination of three cores from down the axis of the canyon system reveals 18 turbidites and debrites, interlayered with hemipelagic muds, that are derived from a mix of shallow and deep sources. Twenty radiocarbon ages indicate that siliciclastic-dominated and mixed turbidites only occur prior to 31 ka during Marine Isotope Stage (MIS) 3, while carbonate-dominated turbidites are well established by 11 ka in MIS1 until as recently as 1.2 ka. The apparent lack of siliciclastic-dominated turbidites and presence of only a few carbonate-dominated turbidites during the MIS2 lowstand are not consistent with generic models of margin sedimentation but might also reflect a gap in the turbidite record. These data suggest that turbidite sedimentation in the Ribbon Reef canyons, probably reflects the complex relationship between the prolonged period (> 25 ka) of MIS3 millennial sea level changes and local factors such as the shelf, inter-reef passage depth, canyon morphology and different sediment sources. On this basis we predict that the spatial and temporal patterns of turbidite sedimentation could vary considerably along the length of the GBR margin.

Range expansion in an invasive small mammal: influence of life-history and habitat quality.

Invasive species pose a major threat to biodiversity but provide an opportunity to describe the processes that lead to changes in a species’ range. The bank vole (Myodes glareolus) is an invasive rodent that was introduced to Ireland in the early twentieth century. Given its continuing range expansion, the substantial empirical data on its spread thus far, and the absence of any eradication program, the bank vole in Ireland represents a unique model system for studying the mechanisms influencing the rate of range expansion in invasive small mammals. We described the invasion using a reaction–diffusion model informed by empirical data on life history traits and demographic parameters. We subsequently modelled the processes involved in its range expansion using a rule-based spatially explicit simulation. Habitat suitability interacted with density-dependent parameters to influence dispersal, most notably the density at which local populations started to donate emigrating individuals, the number of dispersing individuals and the direction of dispersal. Whilst local habitat variability influenced the rate of spread, on a larger scale the invasion resembled a simple reaction–diffusion process. Our results suggest a Type 1 range expansion where the rate of expansion is generally constant over time, but with some evidence for a lag period following introduction. We demonstrate that a two-parameter empirical model and a rule-based spatially explicit simulation are sufficient to accurately describe the invasion history of a species that exhibits a complex, density-dependent pattern of dispersal.

Opening the Frontier: The Gubbio-Perugia Frontier in the Course of History.

The frontier between Gubbio (ancient Umbria) and Perugia (ancient Etruria), in the northeast part of the modern region of Umbria, was founded in the late sixth century BC. The frontier endured in different forms, most notably in the late antique and medieval periods, as well as fleetingly in 1944, and is fossilized today in the local government boundaries. Archaeological, documentary and
philological evidence are brought together to investigate different scales of time that vary from millennia to single days in the representation of a frontier that captured a watershed of geological origins. The foundation of the frontier appears to have been a product of the active agency of the Etruscans, who projected new settlements across the Tiber in the course of the sixth century BC,
protected at the outer limit of their territory by the naturally defended farmstead of Col di Marzo. The immediate environs of the ancient abbey of Montelabate have been studied intensively by targeted, systematic and geophysical survey in conjunction with excavation, work that is still in progress. An overview of the development of the frontier is presented here, employing the data currently available.

Spatially-resolved mapping of history-dependent coupled electrochemical and electronical behaviors of electroresistive NiO

Bias-induced oxygen ion dynamics underpins a broad spectrum of electroresistive and memristive phenomena in oxide materials. Although widely studied by device-level and local voltage-current spectroscopies, the relationship between electroresistive phenomena, local electrochemical behaviors, and microstructures remains elusive. Here, the interplay between history-dependent electronic transport and electrochemical phenomena in a NiO single crystalline thin film with a number of well-defined defect types is explored on the nanometer scale using an atomic force microscopy-based technique. A variety of electrochemically-active regions were observed and spatially resolved relationship between the electronic and electrochemical phenomena was revealed. The regions with pronounced electroresistive activity were further correlated with defects identified by scanning transmission electron microscopy. Using fully coupled mechanical-electrochemical modeling, we illustrate that the spatial distribution of strain plays an important role in electrochemical and electroresistive phenomena. These studies illustrate an approach for simultaneous mapping of the electronic and ionic transport on a single defective structure level such as dislocations or interfaces, and pave the way for creating libraries of defect-specific electrochemical responses.