Search

Abstract The use of Supercomputers is currently very widespread, constituting an essential component in many fields of science. The interest in the use of high performance computation is increasing in a wider and more diverse population of higher education students, mainly senior undergraduates and postgraduates, because the use of these infrastructures allows learners to improve their skills and the results of their training. For this reason, the demand of courses related to supercomputing increases continuously. In this paper we propose, through a wide review of primary studies, several questions that have been considered as a way of knowing the most widely-used contents in Supercomputing training. We have focused on the factors considered for improving training in Supercomputing, in order to improve the results of researchers in higher education organizations, to identify the limitations of Supercomputing training, and to provide solutions for these limitations. During the search procedure for answering research questions, 1911 studies were considered in the first selection. Through the definition of inclusion and exclusion codes in the results of searching databases, 136 published articles were studied. Finally, using quality criteria, 34 studies were identified as relevant in answering the research questions. Several factors were described, such as the way in which courses related to Supercomputing are organized, the adaptations that are currently being applied in curricula related to the students of these techniques, the use of problem-solving training and the qualification of teachers, among the most relevant ones, as well as several limitations of this type of training and the identification of solutions for these limitations. Data was collected by searching keywords related to Supercomputing training and education in the most important databases used in Computational Science, finding empirical evidence to support the positive effect of High Performance Computers (HPC) on educators and researchers. The implications of this study are: first, it provides a summary of the most relevant factors in improving training, as well as the factors that improve the results through the use of a Supercomputer; and second, it provides the analysis of the limitations found for a better performance of learners and the solutions for these limitations. [ABSTRACT FROM AUTHOR]

In this paper, the authors describe a mechanism for the introduction of small variations in the original learning design process defined in a particular Unit of Learning (UoL). The objective is to increase the UoL reusability by offering the designers an alternative to introduce slight variations on the original design instead of creating a new one each time they want to reuse it. No changes or extensions to the Learning Design definition are required to perform these modifications. The use of design patterns to include the adaptations offers the possibility to easily introduce new operations, such as tracing the activity progress, for instance. The structure of a Learning Design player that is able to process the desired adaptation information and to apply it at runtime will be outlined. The player will be part of an architecture for the automatic adaptation of UoLs to their actual context of execution. [ABSTRACT FROM AUTHOR]