Some time ago there was some discussion surrounding versions of Nvidia drivers that had something to do with erroring out after XX number of hours of continuous usage.

I can't remember what version(s) it was.

Also I'm thinking I need to set my systems to do a daily reboot.

My newly reimaged GTX 590 machine ran great for over two days and then suddenly I'm back to TDR BSODs.

I then have to manually abort whatever WUs are in the middle of processing and lose the credits.

I was thinking it had something to do with the 'Nvidia XXhr flu' thing maybe.

Anybody remember this discussion from a while back?

Operator
____________

I started a thread on it back in August, the drivers your using don't have that bug though (I have an uptime of 685 hours).

http://www.gpugrid.org/forum_thread.php?id=3443

If you google 36 hour TDR bug you can find lots of info on it including a long thread in the gForce forums.

I have driver 326.80 on my GTX770 and that PC had a continuous up-time of 13 days when I needed it to boot for updating mouse drivers.
And no problems with this driver on any WU also.
____________
Greetings from TJ

I usually reboot once a month, I check for Microsoft updates, run Ccleaner and defragment. So far everything has been working perfectly (knocking on wood), I was really banging my head on the wall over that one until I found the thread in the gForce forums.