I don’t think that’ll be necessary, but thanks for the offer… Also, those appear to be water cooled, and I’m not sure if they’re standalone or require a water-cooled configuration…either way, I’d rather not deal with it … Now, if you’re having issues with them…well, then maybe we’ll need to take a look.
I spoke with NVidia, and they’re currently not sure why performance is being reported as “jerky”, and didn’t really have any suggestions other than to report a bug. As for not being able to recognize the cards, it sounds like what I thought…If you configure your cards to be “compute driven”, which I guess happens when you use them in parallel (note: This is not the same as SLI), then there is no Primary Display Adapter on the system, and so Microsoft’s generic display driver is in effect…and as of right now, I don’t see (understand) how a program can create an accelerated device context when there is no primary adapter.
Lastly, do not confuse CUDA and its uses with OpenGL and its uses…they are not really the same thing, and have different usage and system requirements.
Please let me know if you run into problems with these cards asap.