ITER is an experimental tokamak device that has many collaborators from different countries and aims to produce energy from nuclear fusion inside a reactor with magnetic confinement. Corresponding scientific data and structure visualization are important factors in connecting qualitative information from human intuition and quantitative content of physical-based datasets. Therefore, building an integrated platform with the interaction of scientific data visualization is inevitable for engineers and researchers in geographically distributed groups to effectively promote consistent design work, and it is also a key part in research and training processes for operators and students to better understand the ITER structure with analysis results. In this context, a digital ITER-type mock-up has been developed at the Institute of Modern Physics, Chinese Academy of Sciences based on an immersive virtual reality platform that is designed using client/server hybrid heterogeneous cloud architecture with many user-friendly and flexible features. Detailed ITER-type models have been developed and optimized in low-poly style to increase rendering speed, and scientific data in the scalar field have been represented by means of the view-dependent ray-casting method with topological texture images, which can fulfill real-time roaming and inspecting tasks for research purposes and provide an efficient and effective tool for fusion technology education at the University of Chinese Academy of Sciences.