sorry last mail got to wrong direction, argghh :(
Hi Robert,
> I have used of even read detailed docs on CUDA yet, but my
> inclination
> would be to map CUDA to a graphics context rather than have
> a
> singleton,
>
Of
course, this is also what I was thought about. The problem is that it
requires some bigger osg introvention, therefor I decided first to try
out to implement it externally by osgPPU and later maybe include it
into the core of osg. However of course one has to think about if we
should provide CUDA to core osg at all, since it is more or less not
scene graph dependent, so it should be somehow moved to an external
nodekit.
> or if a single CUDA context can map to multiple
> contexts
>
I think it cannot, or just cannot in the current version. It requires to set a
certain device before it can start any computation. Hence it is one context-one
device based.
> then for it to be assigned to the GraphicsContext that you
> want to
> associate it with, but his would require that CUDA context
> to be
> thread safe. Now if a single CUDA context can work with
> multiple
> context, on multiple GPU's and work multi-thread then
> perhaps a
> Singleton might be justified.
>
We have to see. I am currently working on the implementation at this moment. I
hope I could say more about this later.
Art
_______________________________________________
osg-users mailing list
[email protected]
http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
