Search

In this paper, we provide a fine-grained parallel scheme for anisotropic mesh adaptation on NUMA 1 1 Non-Uniform Memory Access architectures. Data dependencies are expressed by a graph for each kernel, and concurrency is extracted through fine-grained graph coloring. Tasks are structured into bulk-synchronous steps to avoid data races and to aggregate shared-data accesses. To ensure performance prediction, time cost and load imbalance are theoretically characterized. The devised scheme was evaluated on a 4 NUMA node (2-socket) machine, and a mean efficiency of 70% was reached on 32 cores for 3 kernels out of 4. The impact of irregular degree distribution and data layout on scalability is highlighted. [ABSTRACT FROM AUTHOR]

Smallholder farmers in sub-Saharan Africa seek to improve their livelihoods by investing in new assets. These investments and their effectiveness are constrained by current capital and management practices. Therefore, to understand farm trajectories, the combined effects of different management practices and the timing of investments and losses must be considered. The present study aimed to determine, under 128 distinct scenarios, which ones enable a poorly endowed farm to develop and maintain a sustainable agropastoralism in southwestern Burkina Faso. For this purpose, we used the Ecological Discrete-Event Network (EDEN) modelling framework. This framework includes a formalism based on if-then rules describing economic and ecological events (e.g. investments and losses) that affect qualitative variables. The model rules were built from a literature review, expert interviews, and direct observations. Based on this model, the software then computes all trajectories the farm can take. Based on empirically-reported farm types and trajectories, we then attempted to falsify the modelled dynamics using model-checking techniques. Model predictions matched all observed farm types and trajectories, thus not falsifying the model. Results highlighted that, for this system, livelihood improvement relied on the ability of the farm to to increase its cultivated area, workforce, livestock and fodder resources, all while producing and applying organic inputs to maintain or recover soil fertility. Although qualitative, model predictions are consistent with available observations and provide explanations about farm trajectories in southwestern Burkina Faso. The EDEN modelling framework, through its qualitative — yet rigorous — exploration of all possible trajectories, can help the decision-making process by highlighting the far-reaching consequences of management actions. [Display omitted] • The trajectories of smallholder farms are driven by events triggered by social, environmental and economic conditions. • We studied in which management conditions a poorly endowed farm can develop and maintain a sustainable agropastoralism. • For that purpose, we used a discrete-event modelling framework to predict all possible trajectories of a generic farm. • Model predictions matched reported farm trajectories from southwestern Burkina Faso. • Access to arable land, workforce, soil fertility, livestock and equipment are key for building a sustainable agropastoralism. [ABSTRACT FROM AUTHOR]