This is just amazing! Another great use case of this approach could be computational fluid dynamics (CFD) or other simulation data post processing. As to my knowledge the current approach for large models is to run the post processing in batch mode on a large clusters. The approach used by Nanite could allow post processing on any device with a large and fast enough hard drive. As I understand it you can more or less choose the memory budget. Does anybody know if someone is working on an open source implementation of something similar? Could be a great addition to VTK (as it powers ParaView).
This is just amazing! Another great use case of this approach could be computational fluid dynamics (CFD) or other simulation data post processing. As to my knowledge the current approach for large models is to run the post processing in batch mode on a large clusters. The approach used by Nanite could allow post processing on any device with a large and fast enough hard drive. As I understand it you can more or less choose the memory budget. Does anybody know if someone is working on an open source implementation of something similar? Could be a great addition to VTK (as it powers ParaView).