Computational Science

As technology becomes more and more advanced, and concepts become more complex and abstract, the need to visualize all of these concepts and how they interact with certain parameters becomes harder as well. More variables come into play as the scope of researching or studying things like molecules widen to include not just overall structure but maybe even each individual structure as nucleus and electrons. Computational Science's goal is to use mathematical models and quantitative analysis to expand that scope and simulate on computers such interactions (Source). 

There are many branches of Computational Science, such as Biological Computation which as the name suggests focuses on biology systems. Defining what biological systems compute and the even bigger question -- how it is computed, are the fundamental principles to uncover from such research. As such, things like DNA and cells will need to be programmed, expanding our visual scope the two of the smallest units we know of in the body. Being the smallest unit also means there are going to be millions, maybe billions, of objects in the simulation, which is why, supposedly, computational science projects often require supercomputers according to wikipedia; and it wouldn't be surprising with so many complex objects. 

If such projects depend on supercomputers, it would seem they are bound to have hardware problems eventually. Or at the very least they will become limited by it until hardware issues like RAM processing speed can catch up with other hardware.

References:

http://research.microsoft.com/en-us/groups/science/