Hi Art, On Sun, Nov 23, 2008 at 4:15 PM, Art Tevs <[EMAIL PROTECTED]> wrote: >> 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.
Is it a state machine like OpenGL then? Is so then likely the current thread has a CUDA context active, just like in OpenGL. W.r.t suitability of integration with core OSG, this is something that will probably be required in the future - but done in a way that can easily be remapped to using OpenCL once that is released. In the long term we may well be just bring up an OpenGL graphics context for the frame buffer and texturing and almost everything else w.r.t rendering done in CUDA/OpenCL. Prior to this point one may well be pushing high level scene graph operations down to the GPU as well, and CUDA/OpenCL are potentially good fits for this. Robert. _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
