A chromium based viewer for cumulvs

Abstract—As cluster computing becomes increas-ingly more accessible the need for tools to simplify the development and use of parallel simulations grows proportionally. One class of tools, steering and visualization middle-ware, promises to assist in code development as well as increase the value of the final simulation produced. We are interested in employing a steering and visualization middle-ware package known as CU-MULVS. By creating a distributed visualization client, or viewer, we have been able to produce images of the running application at a much higher data rate than a similarly configured single processor viewer. In this paper, we describe the design of such a viewer and present preliminary performance data. I