Visual Computing
An important aspect of scientific research is to communicate scientific results, putting numerical values into a proper context. The individual scientist must somehow inspect the results to build intuition on the problem domain, and the scientist's findings must be disseminated to a broader audience, letting non-experts gain insight into the problem at hand.
A visual computing application provides powerful means of building intuition into a problem domain. For such an application to provide new insight to the user it must extract useful features, which requires knowledge in the problem domain, that is, the mathematical model and the numerical schemes employed. Furthermore, knowledge in modern visualization techniques and available hardware capabilities are essential to create applications fast enough to allow an interative user experience.
Examining an existing simulation result is interesting in itself, but with today’s massively parallel processors, one can take this one step further and let the user interact directly with the numerical method, calculating new results on the fly.
The Heterogenenous Computing Group at SINTEF ICT Applied Mathematics combines a thorough theoretical mathematical and numerical foundation, expertise on GPU programming and modern visualization techniques, and experience with efficient implementations to focus on visual computing.
