bambiboom said:
oatmealpacket,
Interesting problem. Checking, it appears that the Asus Essentio CG5290 has either a 500W or 600W PSU and while the original GTX 260 requires over 200W, the 660ti uses less(!) >150W. So unless you have some other extreme draw, my guess is that your problem is not from running out of electrons.
Some guesses>
1> Conflict with previous driver not completely uninstalled > Uninstall GTX 660, enable and run on integrated graphics, reinstall GTX 660 and check "Clean install", disable integrated graphics. It is worthwhile to try running on integrated graphics anyway because if the system misbehaves as before, then the problem is not with the 660. Also, going through the procedure to get the clean installation will also ensure here's no conflict with integrated graphics.
2> Reinstall GTX 260 and see if it runs properly on the 320.00 beta driver
3> Use of beta driver > find previous version
4> As the problem occurs after some time of use, and the system is shutting itself off, is there any chance you're overheating the CPU or other component? > Consider installing CPUID HW monitor and check temperatures- this should also display fan speeds.
5> The i7 920 is a 2.5GHz > Is yours overclocked? If so, reset to standard clock speed and test.
6> Try games on lowest settings and if crashes stop, try each setting level until crashes occur.
Just some ideas- wish I could say something more definite.
Cheers,
BambiBoom
Hi,
Thanks for your prompt reply! I should have mentioned that before posting the thread I tried a clean driver install; I should also mention that at this point I've tested with three different drivers. I've used the drivers that the card came with (very old), the latest stable release listed on Nvidia's site and the latest beta. All three produced the same issue - the drivers the card came with froze with red squares when I booted a game for the first time after installing the card (Crysis, the freeze was as the game was displaying copyright information before even getting to the main menu), the stable drivers lasted a bit longer but eventually crashed, the latest beta drivers lasted the longest but also eventually crashed.
The last install was the 320.00 beta drivers; I uninstalled with add/remove programs, rebooted in safe mode, used Driver Fusion to remove all Nvidia display drivers, rebooted to normal mode and installed the beta drivers. Is it possible that earlier drivers would solve the issue?
I've installed CPUID HW Monitor. Without running it at load, the temperatures include:
Temperature 0 41°C (105°F) [0x29] (SYSTIN)
Temperature 1 47°C (115°F) [0x5D] (CPUTIN)
Temperature 2 25°C (77°F) [0x32] (AUXTIN)
Temperature 0 30°C (86°F) [0x1E] (TMPIN0)
SYSTIN - Current: 42c, Min: 41c, Max: 42c
CPUTIN - Current: 47c, Min: 46c, Max: 97c
AUXTIN - Current: 25c, Min: 25c, Max: 50c
Core 0 - Current: 61c, Min: 60c, Max: 64c
Core 1 - Current: 62c, Min: 60c, Max: 63c
Core 2 - Current: 62c, Min: 60c, Max: 63c
Core 3 - Current: 61c, Min: 60c, Max: 63c
GPU - Current: 30c, Min: 34c, Max: 37c
I can get more information from the logs HWMonitor produces if that would help at all.
I haven't intentionally overclocked the CPU at any point; again, I've just run it stock. The machine came with a program called EPU-6 engine which was running in "High Performance" mode; I've turned that down to "Medium Power Saving" mode but haven't tested with it like this yet. Would this make any difference?
I'll need to test with progressively increasing settings later, but does any of this information provide further insight that might lead to an answer? Would it be wise to RMA the card and try using a new one?