Tesla Cards share the same architecture with some of the supported Cards, thus it cannot be a deeper issue.
How so? Sorry, you are being naive! Anything that does not register as a normal graphics card cannot be used by OpenGL, plain and simple. That's all there is to say. AE does not use CUDA or any of that, so even remotely pondering upon using Tesla cards/ bricks is a waste of time. That aside, these things can only be addressed by including the proper NVidia libraries in your code and that at this point is very much up to whoever is using CUDA in his plug-ins and programs, which are not that many, either.
Well, thanks, but I am not sure about the fact that Tesla Cards cannot be used by OpenGL.
I recently saw how the new Arion Render registered a Tesla Card for use by OpenGL very explicitly,
possibly this was a wrong claim of that software.
Those renderers use CUDA, not OpenGL, which are completely different things.
OK, you are probably much deeper into these things than I will ever be.
Thus,I apologize, but I do not really understand your first statement saying that
only "normal graphics card" could be used. From what I was told Tesla Cards are
graphic cards with the only difference that they do not have a video output.....how can
OpenGL recognize this difference ???
Hope yuu are not (yet) annoyed....
A response to the general question:
If you want the After Effects team to add a piece of hardware to its development and testing plan so that the hardware can be officially supported, file a feature request.
Thanks, I sent a request
Well, I tried to figure out what the claim concerning the alleged essential
OpenGL support means.
In the realease notes for the current drivers for Tesla cards one can find that there
is a "Support for OpenGL 3.2"
On the other hand I found articles saying, for instance, that the Mercury Playback
Engine for CS5 is CUDA-based (and I am aware that CUDA and OpenGL are quite differnt things)
Is it really that naive to believe that Tesla cards should be supported ???