Create demo of billion polygon rendering

[allenb]

OpenSG can rendering a billion polygons. Let's show it off and make a little demo of this that we can show other people.

[praveen@…]

Can we add the target platform/architecure for this? It would be good to get an idea for this since the billion polygon target would be platform dependent IMO. Can we also add the type of data set that we will use for this, i.e. acquired (medical), synthetic, tessellated from higher-order surfaces (CAD), etc. This might help in narrowing the target problem further and help identify a target market which will give us maximum visibility.

[dirk]

He's refering to the David test case that we did for the EGPGV 2006 paper. That was scanned polygonal data rendered on a 48 node cluster. Not exactly easy to reproduce, but somewhat impressive anyway. ;)

[allenb]

As far as target platform, I would suggest platform neutral if possible. I don't know the details of how this code works, but it we could get some platform generic code that just provided a basic viewer for clustered rendering as a base, that would be a good start.

Data set type: Various. We could start with synthetic, maybe something like a fractal landscape or something else algorithmic so we can generate whatever size we need on the fly. We coulad also do something artificial like loading the same model many times in different locations in the scene. Long term it would be nice to have more real-world data sets but I don't know of any this large that are available for public use.

[dirk]

I see this one being important, but I want to keep it simple if we can. Therefore I would promote deriving it from the cluster tutorials and creating a very simple polygonal/fractal/sinusoidal landscape for testing.