In situ visualization is clearly a promising solution for ultrascale simulations. We've seen some success in realizing this solution and in ongoing efforts to add support for in situ visualization to open source visualization toolkits such as ParaView and Visit. However, for others to adopt this approach, we need further research and experimental studies to derive a set of guidelines and usable visualization software components. If this research is successful, it will lead to a new visualization and data-understanding infrastructure, potentially change how scientists work, and accelerate scientific discovery. This paper discusses critical issues in realizing in situ visualization and suggest important research directions.
Published in:
Computer Graphics and Applications, IEEE
(Volume:29
,
Issue:
6
)
Date of Publication: Nov.-Dec. 2009