Hello everyone, this will be my maiden post here so here it goes:
Note: I have searched and tried a few different things, but nothing has seemed to solve my problem completely.
 
Comp Specs:
OS: Win7 64 bit.
CPU: i7-3770k @ stock speeds
Mobo: ASUS Sabertooth Z77, Firmware should be near up to date, maybe 1 version behind.
Memory: 16 GB G.Skill @ 1866
GPU: 660 Ti SC (Best Buy Version) (Equivalent to a standard 660 Ti SC I believe).
PSU: Corsair GS700 (54 amps on 12v on 1 rail).
 
Problem: So I swapped out the card (didn't uninstall drivers while the 460 GTX was installed) and booted in no problem. Uninstalled everything then proceeded to install the 600 series drivers (clicking the option for a clean install). Things went fine so I proceeded to play DOTA 2. About 20-30 minutes in was experiencing a lot of stuttering. Graphically I was getting massive frame drops and my sound was "messed up". My comp ended up black screening and freezing with the sound in a crashed loop.

So I uninstall my drivers again, ran a registry cleaner and tried to delete any remaining files in the program files, and program data. No luck, so I did some digging and disabled the High Def Audio via device manager, I had the gpu running on bus 0 function 0, and a HD audio on bus 0 function 0. That seemed to have solved the audio issue, but I was still getting frame rate drops occasionally.
 
During this time I have a dual screen setup so I have precision running on the other window and my temps were running at around 72-74c and the CPU clock was running at around 1100-ish. So I said ok, the temps look a little bit high, even though I know GPU's run hotter than CPU's. I bumped the Fan Speed up to the 80% default max in precision and the temps went down. But a bit while later while playing another DOTA 2 game, I had my game crash and thrown back on the desktop saying my driver has crashed.

Anyone have recommendations for this? I want to try the fan curve thing but not sure what values and what temp range is safe and optimal. Also, should I try to bump the GPU clock offset downward to see if I get stability?
 
Thanks for reading guys.

I am experiencing similar issues, though only the driver crashes from the list of ones you're experiencing. Additionally, I am experiencing red, green and  black screens "of death" - my RMA request for my SLi setup has been approved so hopefully these issues will no longer pose an issue once I receive replacements.
 
Will keep you updated, although it sounds like you're going to have to do the same.

the Card does down clock starting at 70C and setting a manual fan profile is a great idea - I keep my card down to the mid 60's when its working hard
 
If you have updated to the latest Driver version, then you also need to uninstall Precision X DO NOT save any settings reboot then reinstall Precision X and reboot again - a pain but it will stop any issues in the future :)

Update: The card runs fine with Precision X and the auto fan curve, but anytime it is off the card occasionally will give me frame rate drops.
 
I'll probably sometime in the week run games and monitor all of the "vitals". Hopefully this doesn't turn into an RMA or a exchange at the store.

You need to clock it down more to like 1032 and -105 off the memory.

zeroseoul
Self Bump: Instead of making a new thread, here I am!
 

No problem with that in this case, since your still on the same problem. Hope your able to get it fixed. Post back, and tell us how it went.

I guess it is true that when i ran stress tests I never tested the GPU + other components. I never did check CPU utilization during OC Scanner or 3D Mark.
 
I'll try that as well, it makes sense I should draw as much power from the components to mimic a normal situation where I get my freezing.
 
A part of me wants to say that Dota 2 hardly puts a significant load on the CPU and OC Scanner may put in more, but I'll run the tests and see what the numbers churn out to.