Performance Models of Interactive, Immersive Visualization for Scientific Applications

In this paper we develop a performance model for analyzing the end-to-end lag in a combined supercomputer/virtual environment. We first present a general model and then use this model to analyze the lag of an interactive, immersive visualization of a scientific application. This application consists of a finite element simulation executed on an IBM SP-2 parallel supercomputer and the results displayed in real-time in the CAVE Automatic Virtual Environment. Our model decouples the viewpoint lag (not involving the simulation) from the interaction lag (using the results of the simulations). This model allows one to understand the relative contributions to end-to-end lag of the following components: rendering, tracking, network latency, simulation time, and various types of synchronization lags. The results of the study indicate that the rendering and network latency are the major contributors of the end-to-end lag.