High-performance computing for surface modelling and analysis

Real-time response and updating, and concurrent use of complex models are p roblems which require a high computational effort and might thus benefit fr om the use of parallel processing. These problems typically occur in surfac e analysis, a key tool for shape understanding and processing. We survey di fferent applications,of parallel processing at different stages of the shap e-understanding process. The emphasis is on parallel program portability an d reuse of sequential software. Results are provided for different architec tures like workstation networks and massively parallel machines (Cray T3D).