Analyzing the workload of scientific visualization tools: a preliminary study on TIPSY

In recent years, a lot of research has focused on the study of scientific applications running on high-performance systems. So far, however, little concern has been expressed over performance issues related to interactive visualization and analysis. This is due to the lack of standardization of scientific visualization software, as many tools are designed and developed on an ad-hoc basis, coping with specific applications. On the other hand, general-purpose packages for data analysis and visualization are of a proprietary nature, making instrumentation hard. Furthermore, analyzing performance behavior on networked environments in the presence of resource contention from many users is a new and complex field of experimental computer science. In this paper, we study the workload characteristics of TIPSY (Theoretical Image Processing SYstem), which is a portable, client-server, interactive package for visualization and analysis of astrophysics and astronomy simulations. The purpose of our research is to identify workload characteristics of interactive visualization tools and understand performance bottlenecks that arise when running TIPSY on networked environments