Key consideration factors of adopting cloud computing for science

The scientific projects are characterized of requiring high performance computing resources, complex configuration environment, and massive storage. Compared with the traditional grid computing that has been employed by scientific projects for decades, the emerging cloud computing technology offers scientists an alternative of acceptable computing performance and massive storage required with lower infrastructure construction cost and more elastic and flexible system extendibility. Despite these enticing benefits and growing interest of exercising science cloud, there exist uncertainties and challenges that barricade the wide adoption of cloud computing in the scientific field. The goal of this paper is to identify the key consideration factors of adopting cloud computing for science and explore the interrelationship among them. We first developed a comprehensive hierarchy of consideration factors comprising four dimensions and ten underlying factors. Subsequently, we conducted an in-depth survey and interview of 22 domain experts who were actively involved in constructing Taiwan national science clouds. The surveyed data was finally analyzed by employing Decision Making Trial and Evaluation Laboratory (DEMATEL) technique to explore the influential network relationship of these consideration factors. Our findings suggest that Cost, Networking, and Elasticity are the top three dominant factors influencing the adoption of cloud computing for science.