Search

We evaluated the use of coordinated fallowing as a means to control salmon lice Lepeophtheirus salmonis infestation in farmed Atlantic salmon Salmo salar. In discrete management zones, aquaculture operations such as stocking, fallowing, treatments and harvesting are synchronized at all sites in coordinated areas within the zones. The expected benefit of synchronized generations is to reduce the presence of salmon lice larvae after a period of fallowing, as well as to minimize external infestation pressure from surrounding aquaculture sites. A regression analysis was used to evaluate the effectiveness of coordinated fallowing on the progression of external salmon lice infestation pressure and abundance in Atlantic salmon farming sites in 2 areas (zones) in Norway. The overall results show that external infestation pressure was higher inside than outside the management zones, and the external infestation pressure increased with increasing biomass throughout the production cycle. However, within the zones, the external infestation pressure at the beginning of a production cycle was high and in many cases even higher than the general external infestation pressure in the non-coordinated areas. This suggests that external infestation pressure from the neighboring areas has a considerable effect on the fallowed area. Higher numbers of salmon lice were recorded within the zones than outside and, as the production cycle progressed, this phenomenon became more evident. We conclude that the high infestation pressure from salmon lice at the beginning of the grow-out period after fallowing raises severe doubts about the effectiveness of coordinated fallowing practices.

Both cloud and GRID are computing paradigms that manage large distributed resources, and currently there is a lot of interest in their integration. An integrated architecture cloud-GRID requires fine-grained access control and identity federation among untrusted distributed domains in the cloud. This paper deals with the trade-off between security and performance in such architectures, comparing the overhead introduced by cloud services with different security levels. The quantitative results obtained in PerfCloud, an existing cloudgrid infrastructure, are presented and discussed.

International audience; Due to the complexity of modern data parallel applications such as image processing applications, automatic approach to infer suitable and efficient hardware realizations are more and more required. Typically, the optimization of data transfer and storage micro-architecture has a key role for the data parallelism. In this paper, we propose a comprehensive method to explore the mapping of a high-level representation of an application into a customizable hardware accelerator. The high-level representation is in a language called Array-OL. The customizable architecture uses FIFO queues and double buffering mechanism to mask the latency of data transfers and external memory access. The mapping of a high-level representation onto the given architecture is performed by applying a set of loop transformations in Array-OL. A method based on integer partition is used to reduce the space of explored solutions.