Recent trends of mass character and individualization of education impose new requirements on technology-enhanced learning environments. One of the main problems in maintaining high-quality education in knowledge-intensive areas such as Computational Science is rapid obsolescence of learning materials due to the accelerating rate of an appearance of novel research and technological achievements. In this paper, we propose an approach to a fast design of hybrid learning resources which combine different forms of representations of novel scientific results (including hypertext and software) with out-of-the-box support of learning objects’ execution in cloud computing platforms. The experimental part of the study considers the preparation of hybrid learning resources for Master students in High-Performance Computing Department of ITMO University, Russia. We show how the implementation of proposed approach on the basis of CLAVIRE platform speeds up the translation of scientific results into reusable learning materials and thus contributes to the improvement of the quality of training professionals in Computational Science.
