A case for visualization-integrated system-level design space exploration

Abstract

Design space exploration plays an essential role in the system-level design of embedded systems. It is imperative therefore to have efficient and effective exploration tools in the early stages of design, where the design space is largest. System-level simulation frameworks that aim for early design space exploration create large volumes of simulation data in exploring alternative architectural solutions. Interpreting and drawing conclusions from these copious simulation results can be extremely cumbersome. In other domains that also struggle with interpreting large volumes of data, such as scientific computing, data visualization is an invaluable tool. Such visualization is often domain specific and has not become widely used in evaluating the results of computer architecture simulations. Surprisingly little research has been undertaken in the dynamic use of visualization to guide architectural design space exploration. In this paper, we plead for the study and development of generic methods and techniques for run-time visualization of system-level computer architecture simulations. We further explain that these techniques must be scalable and interactive, allowing designers to better explore complex (embedded system) architectures. © Springer-Verlag Berlin Heidelberg 2005.