Interoperability problems

[Aleksandar]

Recently I have carried out many experiments with OpenGL/CUDA interoperability, and instead of being more and more self confident, I became totally confused. :(

On every particular machine I have tried something, it works differently. Does anyone here have experience with CUDA/GL interoperability?

Also, I have a suggestion for the moderators of this site, to open another forum called Interoperability. There are a lot of interesting topics that can be discussed on such forum.

[M. Bially]

There are already forums for each of the different operating systems which make some of the largest differences, other wise this forum is for vendor(usually?) specific drivers, which is probably where you would post...

[Aleksandar]

Well, CUDA/OpenGL interoperability is very vendor (driver) specific topic. On the other hand, it is not likely that people not involved in GL programming will ever encounter similar problems. That's why I have posted my question here.

The most disturbing problem is several orders of magnitude slower frame rate when using CUDA with resident buffers (Bindless extensions) on Vista with 260.99 drivers. On XP everything works perfectly.

Also, the time required to accomplish some tasks is much longer than I have assumed.