CMIIW, but I doubt the AMD patch will do anything on your NVidia card. To me this sounds like an hardware issue. The card was designed for games, which although have high usage, are nowhere near the utilization of vanitygen. Esp. not for prolonged periods of time. If you have a way to improve the cooling system of the machine/card Id give it a try. Another method - common among miners - is to underclock the card until you get it at <70° or even <65° with the existing cooling system.
Yes, the AMD patch was just something to try and I didn't expect it to correct the issue although it seems to have had no effect positive or negative. My temps aren't getting too high for the card and I don't believe I need to change anything with the cards cooling since when vanitygen has not stalled it runs at a solid 100Mkey/s. There's no indication this is a video card / artifact type error and everything points to the unresponsive vanitygen process which I assume is due to some sort of driver / OpenCL issue.
I tried running vanitygen by limiting the threads with the "-t" option, although it still stalled out after some time. It did take a lot longer to stall but it stopped eventually.
When the script catches it and restarts the process, it's back up and running at 100Mkey/s in less than 30 seconds from the process stalling.