General question about NVENC, video capture and remote desktops


I have a strong impression of the answer, but wanted to get a direct answer from those who know better.

Are kepler enabled cards able to receive drawing instructions and output encoded video rather than directly out to a monitor? That is, would one be able to use the graphics acceleration of a kepler card, as well as the h264 encoding, and send that over the network through x11 forwarding, VNC, xrpa, or something similar. I think this is the case but I’ve gotten feedback that x11, xpra and VNC can’t do hardware accelerated graphics.

If so, would all kepler cards do this, or only certain ones? Do the graphics have to bounce back to the CPU between render and encode? Are some graphics forwarding applications (like X11, xpra and VNC) capable of this and others not? Any information or direction would be appreciated. Any hardware (other than graphics card) limitations or requirements I need to be aware of?



By the way, I’m seeing some bizarre behavior from this forum. The list-of-posts page says this post has one comment, made by basimaro4 about an hour after I posted, but then there’s no comment when I click on the post! basimaro4’s profile page has contradictory information about whether or not they’ve made a reply. weird.

That normally is due to deleted SPAM. The forum software doesn’t show banned posts but still counts them.