Analysis of Dependence Tracking Algorithms for Task Dataflow Execution

Processor architectures has taken a turn towards many-core processors, which integrate multiple processing cores on a single chip to increase overall performance, and there are no signs that this trend will stop in the near future. Many-core processors are harder to program than multi-core and single-core processors due to the need of writing parallel or concurrent programs with high degrees of parallelism. Moreover, many-cores have to operate in a mode of strong scaling because of memory bandwidth constraints. In strong scaling increasingly finer-grain parallelism must be extracted in order to keep all processing cores busy.

Task dataflow programming models have a high potential to simplify parallel program- ming because they alleviate the programmer from identifying precisely all inter-task de- pendences when writing programs. Instead, the task dataflow runtime system detects and enforces inter-task dependences during execution based on the description of memory each task accesses. The runtime constructs a task dataflow graph that captures all tasks and their dependences. Tasks are scheduled to execute in parallel taking into account dependences specified in the task graph.

Several papers report important overheads for task dataflow systems, which severely limits the scalability and usability of such systems. In this paper we study efficient schemes to manage task graphs and analyze their scalability. We assume a programming model that supports input, output and in/out annotations on task arguments, as well as commutative in/out and reductions. We analyze the structure of task graphs and identify versions and generations as key concepts for efficient management of task graphs. Then, we present three schemes to manage task graphs building on graph representations, hypergraphs and lists. We also consider a fourth edge-less scheme that synchronizes tasks using integers. Analysis using micro-benchmarks shows that the graph representation is not always scalable and that the edge-less scheme introduces least overhead in nearly all situations.

On the potential of significance-driven execution for energy-aware HPC

Dynamic Voltage and Frequency Scaling (DVFS) exhibits fundamental limitations as a method to reduce energy consumption in computing systems. In the HPC domain, where performance is of highest priority and codes are heavily optimized to minimize idle time, DVFS has limited opportunity to achieve substantial energy savings. This paper explores if operating processors Near the transistor Threshold Volt- age (NTV) is a better alternative to DVFS for break- ing the power wall in HPC. NTV presents challenges, since it compromises both performance and reliability to reduce power consumption. We present a first of its kind study of a significance-driven execution paradigm that selectively uses NTV and algorithmic error tolerance to reduce energy consumption in performance- constrained HPC environments. Using an iterative algorithm as a use case, we present an adaptive execution scheme that switches between near-threshold execution on many cores and above-threshold execution on one core, as the computational significance of iterations in the algorithm evolves over time. Using this scheme on state-of-the-art hardware, we demonstrate energy savings ranging between 35% to 67%, while compromising neither correctness nor performance.

SHEPARD: Scheduling on heterogeneous platforms using application resource demands

Heterogeneous computing technologies, such as multi-core CPUs, GPUs and FPGAs can provide significant performance improvements. However, developing applications for these technologies often results in coupling applications to specific devices, typically through the use of proprietary tools. This paper presents SHEPARD, a compile time and run-time framework that decouples application development from the target platform and enables run-time allocation of tasks to heterogeneous computing devices. Through the use of special annotated functions, called managed tasks, SHEPARD approximates a task's performance on available devices, and coupled with the approximation of current device demand, decides which device can satisfy the task with the lowest overall execution time. Experiments using a task parallel application, based on an in-memory database, demonstrate the opportunity for automatic run-time task allocation to achieve speed-up over a static allocation to a single specific device. © 2014 IEEE.

On Being Let Loose in the Field: The Execution of Professional Ethics

In recent years concerns over litigation and the trend towards close monitoring of academic activity has seen the effective hijacking of research ethics by university managers and bureaucrats. This can effectively curtail cutting edge research as perceived ‘safe’ research strategies are encouraged. However, ethics is about more than research governance. Ultimately, it seeks to avoid harm and to increase benefits to society. Rural development debate is fairly quiet on the question of ethics, leaving guidance to professional bodies. This study draws on empirical research that examined the lives of migrant communities in Northern Ireland. This context of increasingly diverse rural development actors provides a backdrop for the way in which the researcher navigates through ethical issues as they unfold in the field. The analysis seeks to relocate ethics from being an annoying bureaucratic requirement to one where it is inherent to rigorous and professional research and practice. It reveals how attention to professional ethics can contribute to effective, situated and reflexive practice, thus transforming ethics to become an asset to professional researchers.

Software-Managed Energy-Efficient Hybrid DRAM/NVM Main Memory

This paper evaluates the viability of user-level software management of a hybrid DRAM/NVM main memory system. We propose an operating system (OS) and programming interface to place data from within the user application. We present a profiling tool to help programmers decide on the placement of application data in hybrid memory systems. Cycle-accurate simulation of modified applications confirms that our approach is more energy-efficient than state-of-the- art hardware or OS approaches at equivalent performance. Moreover, our results are validated on several candidate NVM technologies and a wide set of 14 benchmarks.
The key observation behind this work is that, for the work- loads we evaluated, application objects are too short-lived to motivate migration. Utilizing this property significantly reduces the hardware complexity of hybrid memory systems.

A sting in the spit: widespread cross-infection of multiple RNA viruses across wild and managed bees

Declining populations of bee pollinators are a cause of concern, with major repercussions for biodiversity loss and food security. RNA viruses associated with honeybees represent a potential threat to other insect pollinators, but the extent of this threat is poorly understood. This study aims to attain a detailed understanding of the current and ongoing risk of emerging infectious disease (EID) transmission between managed and wild pollinator species across a wide range of RNA viruses. Within a structured large-scale national survey across 26 independent sites, we quantify the prevalence and pathogen loads of multiple RNA viruses in co-occurring managed honeybee (Apis mellifera) and wild bumblebee (Bombus spp.) populations. We then construct models that compare virus prevalence between wild and managed pollinators. Multiple RNA viruses associated with honeybees are widespread in sympatric wild bumblebee populations. Virus prevalence in honeybees is a significant predictor of virus prevalence in bumblebees, but we remain cautious in speculating over the principle direction of pathogen transmission. We demonstrate species-specific differences in prevalence, indicating significant variation in disease susceptibility or tolerance. Pathogen loads within individual bumblebees may be high and in the case of at least one RNA virus, prevalence is higher in wild bumblebees than in managed honeybee populations. Our findings indicate widespread transmission of RNA viruses between managed and wild bee pollinators, pointing to an interconnected network of potential disease pressures within and among pollinator species. In the context of the biodiversity crisis, our study emphasizes the importance of targeting a wide range of pathogens and defining host associations when considering potential drivers of population decline.

PAlliative Care in chronic Kidney diSease: the PACKS study—quality of life, decision making, costs and impact on carers in people managed without dialysis

BACKGROUND: The number of patients with advanced chronic kidney disease opting for conservative management rather than dialysis is unknown but likely to be growing as increasingly frail patients with advanced renal disease present to renal services. Conservative kidney management includes ongoing medical input and support from a multidisciplinary team. There is limited evidence concerning patient and carer experience of this choice. This study will explore quality of life, symptoms, cognition, frailty, performance decision making, costs and impact on carers in people with advanced chronic kidney disease managed without dialysis and is funded by the National Institute of Health Research in the UK.

METHODS: In this prospective, multicentre, longitudinal study, patients will be recruited in the UK, by renal research nurses, once they have made the decision not to embark on dialysis. Carers will be asked to 'opt-in' with consent from patients. The approach includes longitudinal quantitative surveys of quality of life, symptoms, decision making and costs for patients and quality of life and costs for carers, with questionnaires administered quarterly over 12 months. Additionally, the decision making process will be explored via qualitative interviews with renal physicians/clinical nurse specialists.

DISCUSSION: The study is designed to capture patient and carer profiles when conservative kidney management is implemented, and understand trajectories of care-receiving and care-giving with the aim of optimising palliative care for this population. It will explore the interactions that lead to clinical care decisions and the impact of these decisions on informal carers with the intention of improving clinical outcomes for patients and the experiences of care givers.