Hi Felix, The Intel embedded graphics chipset is going to be limited by different things than NVidia graphics card. The Intel chipset will be limited far more by transform and fill than the graphics card so optimizations in the way the data is stored and passed to OpenGL will have less impact. The NVidia graphics card performs far better in transform and fill so you find the way you pass the data to the graphics card far more critical - so VBO's and appropriate batching of the data will be far more critical.
Robert. On Tue, Nov 9, 2010 at 1:08 PM, Felix Engelmann <[email protected]> wrote: > Dear OSG Community, > > I'm developing visualization software, which involves rendering of large > point clouds. I optimized the scene graph in order to get better frame rates. > In particular I now use osg::Geometry instead of osg::ShapeDrawable. > > I tested it on 2 different machines and got different performance gains. > System 1 (Notebook): Intel Pentium-M, 1.86GHz, 1GB RAM, Intel GMA900 Chipset, > Windows XP > System 2 (Desktop PC): Intel Core 2 DuoE8400, 3 GHz, 2GB RAM, GeForce 9500 > GT, Windows XP > > On System 1 the rendering is about 40% faster with the optimizations. And on > System 2 I got a 7x higher framerate ! > > Now, my question is: what can be the reasons for that dramatically differnet > performance gain? I suppose, it is because of the different OpenGL > implementations and graphics hardware, but I don't have a clue whats really > going on here ... > Who can give me some comprehensible explanations? > > Thank you! > > Cheers, > Felix > > ------------------ > Read this topic online here: > http://forum.openscenegraph.org/viewtopic.php?p=33610#33610 > > > > > > _______________________________________________ > osg-users mailing list > [email protected] > http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org > _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
