• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

GPU overheat causing system crash during load on custom water loop?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Bageland2000

Member
Joined
Jul 20, 2005
Location
Chicago, IL
I've had an issue with any games causing system crashes for a really long time now. I've thought that it was a symptom of me failing to dial-in my overclock, but I'm realizing that I think it's a heat related issue. I've set my CPU/GPU to stock and the issues persist. My loop is a 2x120 rad; Swiftech CPU block; EK 290 full cover block w/ back plate, and an XSPC dual bay res/pump. When running benches like AIDA64 w/ GPU stress test, the back plate stays fairly cool. When running games, my system will typically crash after about 10-30 minutes. When I touch the back plate, it's almost too hot to keep my hand on.

My CPU and GPU temps from AIDA64 and GPU-Z are consistently mid to high 50s, so I didn't think there could possibly be a heat issue. But I can't think of what else could be causing the crashes. Is there a chance my VRMs are overheating? I got the GPU block and back plate and installed it myself, and I used heat pads and thermal compound on the VRMs. I got the card as a miner on this forum, but I don't remember it crashing like this until I put on the water block. It was over a year ago though, so my memory is fuzzy.

Any thoughts?
 
your gpu is crashing, not a full os crash, so
let's start with the free stuff first.
lets go to guru3d.com and download display driver uninstaller and download a fresh gpu driver.
extract display driver uninstaller.
take all clocks off the rig.
reboot the system.
open display driver uninstaller, it will want to reboot to safe mode, yes to that dialog box.
when it self opens remove your driver, all of it, select remove driver and restart.
it will reboot and you can install a fresh driver.
 
your gpu is crashing, not a full os crash, so
let's start with the free stuff first.
lets go to guru3d.com and download display driver uninstaller and download a fresh gpu driver.
extract display driver uninstaller.
take all clocks off the rig.
reboot the system.
open display driver uninstaller, it will want to reboot to safe mode, yes to that dialog box.
when it self opens remove your driver, all of it, select remove driver and restart.
it will reboot and you can install a fresh driver.
I'll try that, but it seems like a full OS crash. I have to restart the system and usually recover the BIOS.
 
OK, about to try that now.

Since you mentioned it's a driver issue, I realized that every time this happens, I have to remove and reinsert my USB wireless adapter before I can see any SSIDs. Could this driver be causing the crash somehow?
 
OK, about to try that now.

Since you mentioned it's a driver issue, I realized that every time this happens, I have to remove and reinsert my USB wireless adapter before I can see any SSIDs. Could this driver be causing the crash somehow?
Probably not causing it but that problem is definitely a bi-product of the crash. Same thing happens on one of my rigs.
 
the lack of bsod, not self restarting and audio working says the os is still alive.
 
Yup, I would do as Caddi said and then post back. Give it a few days and go from there.

I will add though that it doesn't look like a efficient running loop with only 120.2 of rad for a presumed OC'd CPU and GPU. When you say your CPU and GPU temps are in the 50s, are those idle temps?
 
OK so I fully uninstalled and reinstalled the driver, tested for a few hours, then overclocked the CPU (GPU is still stock) and ran games for about an hour so far. So far, so good. Is it possible that running CRU to overclock my monitor is causing the problem?
 
OK so I fully uninstalled and reinstalled the driver, tested for a few hours, then overclocked the CPU (GPU is still stock) and ran games for about an hour so far. So far, so good. Is it possible that running CRU to overclock my monitor is causing the problem?

So I ran a really long time without an issue. Added GPU overclock, played some Witcher 3, still no problems. Was getting 60fps at med/high settings. Cranked up settings (ambient occlusion, etc.) and it crashed almost instantly. Black screen, looping audio (several times a second causing almost a buzzing sound) and no ability to exit to Windows. Reset, and had to clear CMOS before PC would POST (Same as before.)
 
GPU OC is not stable enough for the higher settings. Try a different driver or lower your OC.
 
R290 on water.

OC was 947->1036 core/1250->1360 RAM

OS:Windows 10 x64
Motherboard: ASRock z87 Extreme4
CPU: i5 4670K
RAM: 2x4GB Crucial Ballistix TT 1866
GPU/VIDEO: XFX R290 w/ EK WB
HDD: Corsair Force GS 240GB SSD
Display: Acer G257HU 1440p
Sound: Onboard
 
OC doesn't appear to be that radical. Was this a mining card?
I'd try a different driver.
 
I've had this same issue and it was heat related for me as well so I'm willing to bet my last dollar that it is a heat issue for you as well.

I tried the clean driver reinstalls, hell I tried RMA'ing parts and nothing seemed to work...that is until I regarded it as a heat issue.

One thing which helped me is I went into the bios and for my Asus MB there is a fan control. I raised the CPU and chassis upper temp limits to 65°C respectively and that really helped decrease the frequency of the black screen with looping audio, but it was still happening for me infrequently.

Come to find out my liquid cooler is malfunctioning and my CPU is at 40°C while I sit here typing this, the only application I'm running is the web browser. My water block is a clear see through one and a few days ago I noticed droplets of moisture forming inside of it looking like condensation. This signifies to me air is getting inside of the block.

If you treat this like a heat related issue I'm sure you will find a solution.
 
Back