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.

Autonomic scheduling of tasks from data parallel patterns to CPU/GPU core mixes

We propose a methodology for optimizing the execution of data parallel (sub-)tasks on CPU and GPU cores of the same heterogeneous architecture. The methodology is based on two main components: i) an analytical performance model for scheduling tasks among CPU and GPU cores, such that the global execution time of the overall data parallel pattern is optimized; and ii) an autonomic module which uses the analytical performance model to implement the data parallel computations in a completely autonomic way, requiring no programmer intervention to optimize the computation across CPU and GPU cores. The analytical performance model uses a small set of simple parameters to devise a partitioning-between CPU and GPU cores-of the tasks derived from structured data parallel patterns/algorithmic skeletons. The model takes into account both hardware related and application dependent parameters. It computes the percentage of tasks to be executed on CPU and GPU cores such that both kinds of cores are exploited and performance figures are optimized. The autonomic module, implemented in FastFlow, executes a generic map (reduce) data parallel pattern scheduling part of the tasks to the GPU and part to CPU cores so as to achieve optimal execution time. Experimental results on state-of-the-art CPU/GPU architectures are shown that assess both performance model properties and autonomic module effectiveness. © 2013 IEEE.

The impacts of management reform on irrigation water use efficiency in the Guanzhong plain, China

This paper examines the efficiency of the 1998 irrigation management reform in the Guanzhong Plain, Shaanxi, China, at farm and canal level. Stochastic frontier analysis is applied to estimate irrigation water use efficiency, based on panel data for 800 farmers, spread over 80 irrigation canals, for the period 1999–2005. Analysis of determinants of water use efficiency shows that at farm level, water price and disclosure are important factors. Compared to the base case of unreformed, management reform has a positive impact with water user association having the largest effect, followed by joint-stock co-operative and private company. The canal model is in line with the farm level model, although estimates are less significant.

Technical and allocative efficiency of irrigation water use in the Guanzhong Plain, China

Due to increasing water scarcity, accelerating industrialization and urbanization, efficiency of irrigation water use in Northern China needs urgent improvement. Based on a sample of 347 wheat growers in the Guanzhong Plain, this paper simultaneously estimates a production function, and its corresponding first-order conditions for cost minimization, to analyze efficiency of irrigation water use. The main findings are that average technical, allocative, and overall economic efficiency are 0.35, 0.86 and 0.80, respectively. In a second stage analysis, we find that farmers’ perception of water scarcity, water price and irrigation infrastructure increase irrigation water allocative efficiency, while land fragmentation decreases it. We also show that farmers’ income loss due to higher water prices can be offset by increasing irrigation water use efficiency.

Compiler and Runtime Support for Hybrid Static/Dynamic Scheduling

One of the outstanding issues in parallel computing is the selection of task granularity. This work proposes a solution to the task granularity problem by lowering the overhead of the task scheduler and as such supporting very fine-grain tasks. Using a combination of static (compile-time) scheduling and dynamic (run-time) scheduling, we aim to make scheduling decisions as fast as with static scheduling while retaining the dynamic load- balancing properties of fully dynamic scheduling. We present an example application and discuss the requirements on the compiler and runtime system to realize hybrid static/dynamic scheduling.

Multiuser Scheduling for Cognitive MIMO with Channel Estimation Errors and Feedback Delay

A multiuser scheduling multiple-input multiple-output (MIMO) cognitive radio network (CRN) with space-time block coding (STBC) is considered in this paper, where one secondary base station (BS) communicates with one secondary user (SU) selected from K candidates. The joint impact of imperfect channel state information (CSI) in BS → SUs and BS → PU due to channel estimation errors and feedback delay on the outage performance is firstly investigated. We obtain the exact outage probability expressions for the considered network under the peak interference power IP at PU and maximum transmit power Pm at BS which cover perfect/imperfect CSI scenarios in BS → SUs and BS → PU. In addition, asymptotic expressions of outage probability in high SNR region are also derived from which we obtain several important insights into the system design. For example, only with perfect CSIs in BS → SUs, i.e., without channel estimation errors and feedback delay, the multiuser diversity can be exploited. Finally, simulation results confirm the correctness of our analysis.

Sprinkler irrigation of rice fields reduces grain arsenic but enhances cadmium

Previous studies have demonstrated that rice cultivated under flooded conditions has higher concentrations of arsenic (As) but lower cadmium (Cd) compared to rice grown in unsaturated soils. To validate such effects over long terms under Mediterranean conditions a field experiment, conducted over 7 successive years was established in SW Spain. The impact of water management on rice production and grain arsenic (As) and cadmium (Cd) was measured, and As speciation was determined to inform toxicity evaluation. Sprinkler irrigation was compared to traditional flooding.

Both irrigation techniques resulted in similar grain yields (similar to 3000 kg grain ha(-1)). Successive sprinkler irrigation over 7 years decreased grain total As to one-sixth its initial concentration in the flooded system (0.55 to 0.09 mg As kg(-1)), while one cycle of sprinkler irrigation also reduced grain total As by one-third (0.20 mg kg(-1)). Grain inorganic As concentration increased up to 2 folds under flooded conditions compared to sprinkler irrigated fields while organic As was also lower in sprinkler system treatments, but to a lesser extent. This suggests that methylation is favored under water logging. However, sprinkler irrigation increased Cd transfer to grain by a factor of 10, reaching 0.05 mg Cd kg(-1) in 7 years. Sprinlder systems in paddy fields seem particularly suited for Mediterranean climates and are able to mitigate against excessive As accumulation, but our evidence shows that an increased Cd load in rice grain may result.

Computational scheduling methods for integrating plug-in electric vehicles with power systems: a review

Traditional internal combustion engine vehicles are a major contributor to global greenhouse gas emissions and other air pollutants, such as particulate matter and nitrogen oxides. If the tail pipe point emissions could be managed centrally without reducing the commercial and personal user functionalities, then one of the most attractive solutions for achieving a significant reduction of emissions in the transport sector would be the mass deployment of electric vehicles. Though electric vehicle sales are still hindered by battery performance, cost and a few other technological bottlenecks, focused commercialisation and support from government policies are encouraging large scale electric vehicle adoptions. The mass proliferation of plug-in electric vehicles is likely to bring a significant additional electric load onto the grid creating a highly complex operational problem for power system operators. Electric vehicle batteries also have the ability to act as energy storage points on the distribution system. This double charge and storage impact of many uncontrollable small kW loads, as consumers will want maximum flexibility, on a distribution system which was originally not designed for such operations has the potential to be detrimental to grid balancing. Intelligent scheduling methods if established correctly could smoothly integrate electric vehicles onto the grid. Intelligent scheduling methods will help to avoid cycling of large combustion plants, using expensive fossil fuel peaking plant, match renewable generation to electric vehicle charging and not overload the distribution system causing a reduction in power quality. In this paper, a state-of-the-art review of scheduling methods to integrate plug-in electric vehicles are reviewed, examined and categorised based on their computational techniques. Thus, in addition to various existing approaches covering analytical scheduling, conventional optimisation methods (e.g. linear, non-linear mixed integer programming and dynamic programming), and game theory, meta-heuristic algorithms including genetic algorithm and particle swarm optimisation, are all comprehensively surveyed, offering a systematic reference for grid scheduling considering intelligent electric vehicle integration.

Efficiently Scheduling Task Dataflow Parallelism: A Comparison Between Swan and QUARK

Increased system variability and irregularity of parallelism in applications put increasing demands on the ef- ficiency of dynamic task schedulers. This paper presents a new design for a work-stealing scheduler supporting both Cilk- style recursively parallel code and parallelism deduced from dataflow dependences. Initial evaluation on a set of linear algebra kernels demonstrates that our scheduler outperforms PLASMA’s QUARK scheduler by up to 12% on a 16-thread Intel Xeon and by up to 50% on a 32-thread AMD Bulldozer.

Optimal Scheduling Methods to Integrate Plug-In Electric Vehicles with the Power System: A Review

The introduction of the Tesla in 2008 has demonstrated to the public of the potential of electric vehicles in terms of reducing fuel consumption and green-house gas from the transport sector. It has brought electric vehicles back into the spotlight worldwide at a moment when fossil fuel prices were reaching unexpected high due to increased demand and strong economic growth. The energy storage capabilities from of fleets of electric vehicles as well as the potentially random discharging and charging offers challenges to the grid in terms of operation and control. Optimal scheduling strategies are key to integrating large numbers of electric vehicles and the smart grid. In this paper, state-of-the-art optimization methods are reviewed on scheduling strategies for the grid integration with electric vehicles. The paper starts with a concise introduction to analytical charging strategies, followed by a review of a number of classical numerical optimization methods, including linear programming, non-linear programming, dynamic programming as well as some other means such as queuing theory. Meta-heuristic techniques are then discussed to deal with the complex, high-dimensional and multi-objective scheduling problem associated with stochastic charging and discharging of electric vehicles. Finally, future research directions are suggested.

Effects of arsenic-contaminated irrigation water on growth, yield, and nutrient concentration in rice

A study was undertaken to determine the effects of different concentrations of arsenic (As) in irrigation water on Boro (dry-season) rice (Oryza sativa) and their residual effects on the following Aman (wet-season) rice. There were six treatments, with 0, 0.1, 0.25, 0.5, 1, and 2 mg As L-1 applied as disodium hydrogen arsenate. All the growth and yield parameters of Boro rice responded positively at lower concentrations of up to 0.25 mg As L-1 in irrigation water but decreased sharply at concentrations more than 0.5 mg As L-1. Arsenic concentrations in grain and straw of Boro rice increased significantly with increasing concentration of As in irrigation water. The grain As concentration was in the range of 0.25 to 0.97 μg g-1 and its concentration in rice straw varied from 2.4 to 9.6 μg g-1 over the treatments. Residual As from previous Boro rice showed a very similar pattern in the following Aman rice, although As concentration in Aman rice grain and straw over the treatments was almost half of the As levels in Boro rice grain. Arsenic concentrations in both grain and straw of Boro and Aman rice were found to correlate with iron and be antagonistic with phosphorus. Copyright © Taylor & Francis Group, LLC.

Latent vs. Observed Variables: Analysis of Irrigation Water Efficiency Using SEM and SUR

In this paper we compare conceptualising single factor technical and allocative efficiency as indicators of a single latent variable, or as separate observed variables. In the former case, the impacts on both efficiency types are analysed by means of structural equation modeling (SEM), in the latter by seemingly unrelated regression (SUR). We compare estimation results of the two approaches based on a dataset on single factor irrigation water use efficiency obtained from a survey of 360 farmers in the Guanzhong Plain, China. The main methodological findings are that SEM allows identification of the most important dimension of irrigation water efficiency (technical efficiency) via comparison of their factor scores and reliability. Moreover, it reduces multicollinearity and attenuation bias. It thus is preferable to SUR. The SEM estimates show that perception of water scarcity is the most important positive determinant of both types of efficiency, followed by irrigation infrastructure, income and water price. Furthermore, there is a strong negative reverse effect from efficiency on perception.