30 likes | 160 Vues
Managing large geometric models often exceeds the RAM capacity of typical PCs, necessitating External Memory (EM) techniques. This poster presents the Objects Management in Secondary Memory (OMSM) framework, designed to be dimension-independent and extensible. OMSM accommodates user needs via dynamic plugins, enabling seamless integration of various techniques into its storage architecture. The framework features independent aspects like spatial indexing, clustering policy, and transfer management, all facilitated through a unified interface and dynamic loading design patterns.
E N D
A Dimension-Independent and Extensible Framework for Huge Geometric Models David CANINO – canino@disi.unige.it
Brief Description • The dimension of huge models often exceeds the RAM size in a common PC workstation. Thus, using an External Memory (EM) technique is mandatory. • In this poster, we define a dimension-independent and extensible framework, called Objects Management in Secondary Memory (OMSM). • The OMSM framework can be easily adapted to the users needs through dynamic plugins, integrating a lot of techniques in a storing architecture. David CANINO – canino@disi.unige.it – Dept. of Computer Science, Genoa, IT
OMSM Organization • Independent aspects: • Spatial indexing • Clustering policy • Clusters transfer • A layer foreach aspect with a Common Interface • Independent data model for each layer • Interface and Dynamic Loading design patterns David CANINO – canino@disi.unige.it – Dept. of Computer Science, Genoa, IT