Hi, that does seem to be the type of thing that's being done. Definately cool stuff, but not what I'm looking for.
The plan now is to see if the computation and visualization segments can be split in such away that visualization takes place on the local workstation with spiffy GL hardware and computation can take place on the compute nodes of the cluster. we'll see how it goes, I'm not familiar enough with the code base in question to say if this is feasable, but it's the next thing we're looking at. Thanks for this interesting discussions, -Jon
