[vtkusers] Scalability

Berk Geveci berk.geveci at gmail.com
Mon Nov 8 08:53:21 EST 2004


VTK is being used more and more for large scale data. We have put a
lot of effort in updating the architecture of VTK so that scalable
pipelines can be created. This does require some effort however. You
have to experiment and find the right filters, the right parallel
rendering algorithm. This is what ParaView tries to do. It also has
additional code, such as IceT from Sandia Labs to provide better
scalability when rendering on a distributed system. With some effort,
all of these tools can be used through VTK alone but I would recommend
starting with ParaView. As of this release (1.8), ParaView provides
access to it's backend engine through C++ (and any other interpreted
language VTK supports). This should allow developers to create
customized user interfaces.


On Mon, 8 Nov 2004 09:43:46 +0100, Roland Schulz <r2s2 at gmx.de> wrote:
> Hey,
> 
> how does VTK scale to Terabyte data and over 200 render processes? Is VTK
> apropriate for that?
> The goal is to visual isoplanes for finite element data. We have no export
> filter yet, so we can't just test it. Has someone a tool to create
> parallel/partitioned sample data of that size?
> Is the scalability different when using Paraview ontop of VTK?
> 
> regards
> Roland Schulz
> 
> _______________________________________________
> This is the private VTK discussion list.
> Please keep messages on-topic. Check the FAQ at: &lt;http://www.vtk.org/Wiki/VTK_FAQ&gt;
> Follow this link to subscribe/unsubscribe:
> http://www.vtk.org/mailman/listinfo/vtkusers
>



More information about the vtkusers mailing list