I built myself a new computer earlier this year. When I first built it, I was playing a lot of Dark Souls and watching streaming videos, and I encountered no issues. A few weeks after I built it, I tried playing Farcry 3: Blood Dragon and it happened for the first time: A short, highpitched sound comes over my headphones, the screen freezes on whatever image it is currently displaying, and accepts no mouse+keyboard input. This freeze reliably happens in many different games, such as Dark Souls 2+GeDoSaTo+GEM (~30 minutes), Star Citizen (~5 minutes), Witcher 2 (~10 minutes), or Borderlands 2 (~8 minutes). It also happens outside of gaming, though with a much lower frequency.
I immediately suspected my graphics card. I borrowed an Asus GTX 660 ti DirectCUII from a coworker, went home during lunch, played ~30 minutes of the Witcher 2, and then returned the borrowed card. That seemed like a clear indictment of my graphics card.
I logged my temperatures, and my GPU never went above 62. I tried reverting to earlier drivers, but to no avail. I used Asus GPU Tweak to underclock my GPU, which did increase the time I could play before it freezes. I did some research and saw many reports of the Asus GTX 780 ti DirectCU II OC having problems with OCP/OVP. I RMA'd my card to Asus (which still hasn't returned), and I ordered a new one from Amazon. Then the new one had the same problem. I immediately returned it to Amazon, and they shipped me a new one. That one had the same issue. I read online about using an OCP/OVP disabler from KingpinCooling to fix the problem. I tried it, but it did not fix my problem. But it started seeming improbable that I'd get 3 bad cards in a row.
A couple days ago, I thought that it might be the OCP/OVP on the PSU that's cutting it out. I was using two 6+2 connectors to power the graphics card. I used two 2x6 pin to 8 pin converters, thus now using four 6 pin connectors to power my GPU. This seemed to fix the problem entirely. I was able to play Star Citizen's AC: Vaduul Swarm through wave 18 (when I quit, no freeze), whereas before I never made it past wave 4. I could play Dark Souls 2 for hours with no crashes. Then I was looking at my PSU and wondered what would happen if I switched its fan mode from "Normal" to "Hybrid". After playing Dark Souls 2 for ~10-15 minutes, it had the same freeze that I experienced before. Returning the fan mode to "Normal" fixed it.
It seemed everything was going well, and that I'd finally addressed the problem. I took yesterday off from work to celebrate, and spent the whole day gaming (mostly playing Dark Souls 2 and Star Citizen) with no problems. After a long day of almost nonstop gaming, I started up Borderlands 2. After playing for ~5 minutes, the freeze happened. I kept doing a hard reboot and trying different things, running the Disable.exe, changing the seating of the graphics card, replacing the power cables.
Relevant links:
My computer specs: http://pcpartpicker.com/p/9LJQK8 (though, I only have one graphics card in it at the moment)
Via Speccy: http://speccy.piriform.com/results/5lf0Zsk1i2nT8vNnsbZ8gtS
My 3DMark: http://www.3dmark.com/3dm/3572415
My GPU-Z log leading up to a freeze: https://docs.google.com/file/d/0B272VczsphYrZmFSNzRQaTZleVk/edit
I immediately suspected my graphics card. I borrowed an Asus GTX 660 ti DirectCUII from a coworker, went home during lunch, played ~30 minutes of the Witcher 2, and then returned the borrowed card. That seemed like a clear indictment of my graphics card.
I logged my temperatures, and my GPU never went above 62. I tried reverting to earlier drivers, but to no avail. I used Asus GPU Tweak to underclock my GPU, which did increase the time I could play before it freezes. I did some research and saw many reports of the Asus GTX 780 ti DirectCU II OC having problems with OCP/OVP. I RMA'd my card to Asus (which still hasn't returned), and I ordered a new one from Amazon. Then the new one had the same problem. I immediately returned it to Amazon, and they shipped me a new one. That one had the same issue. I read online about using an OCP/OVP disabler from KingpinCooling to fix the problem. I tried it, but it did not fix my problem. But it started seeming improbable that I'd get 3 bad cards in a row.
A couple days ago, I thought that it might be the OCP/OVP on the PSU that's cutting it out. I was using two 6+2 connectors to power the graphics card. I used two 2x6 pin to 8 pin converters, thus now using four 6 pin connectors to power my GPU. This seemed to fix the problem entirely. I was able to play Star Citizen's AC: Vaduul Swarm through wave 18 (when I quit, no freeze), whereas before I never made it past wave 4. I could play Dark Souls 2 for hours with no crashes. Then I was looking at my PSU and wondered what would happen if I switched its fan mode from "Normal" to "Hybrid". After playing Dark Souls 2 for ~10-15 minutes, it had the same freeze that I experienced before. Returning the fan mode to "Normal" fixed it.
It seemed everything was going well, and that I'd finally addressed the problem. I took yesterday off from work to celebrate, and spent the whole day gaming (mostly playing Dark Souls 2 and Star Citizen) with no problems. After a long day of almost nonstop gaming, I started up Borderlands 2. After playing for ~5 minutes, the freeze happened. I kept doing a hard reboot and trying different things, running the Disable.exe, changing the seating of the graphics card, replacing the power cables.
Relevant links:
My computer specs: http://pcpartpicker.com/p/9LJQK8 (though, I only have one graphics card in it at the moment)
Via Speccy: http://speccy.piriform.com/results/5lf0Zsk1i2nT8vNnsbZ8gtS
My 3DMark: http://www.3dmark.com/3dm/3572415
My GPU-Z log leading up to a freeze: https://docs.google.com/file/d/0B272VczsphYrZmFSNzRQaTZleVk/edit