Performance and Usability Improvements for Massive Data Grids using Silverlight

Abstract: TRIMMA Affärsutveckling AB is developing and marketing a business intelligence solution called INSIGHT. INSIGHT presents tables showing possibly very large data sets and the performance and user experience is sometimes suffering. The main objective of this thesis is an evaluation of the pros and cons of replacing the existing ASP.NET/HTML table component in INSIGHT with a component developed in Silverlight.This thesis examines two techniques to speed up a Silverlight application showing a lot of data: UI- and data virtualization. UI virtualization intends to render only the user interface elements that appear on the screen and are visible to the user, while data virtualization intends to fetch (from the data source) only the section of the data that is visible to the user. The result of the project is a fully working prototype integrated into a test version of INSIGHT. Performance testing results indicate that the prototype performs approximately the same as the ASP.NET/HTML version of INSIGHT for small tables but significantly better for large data sets.  The prototype also contains a few extra features, not available in INSIGHT, exemplifying the possibilities to create highly responsive user interfaces in Silverlight