- Apr 11, 2013
- 15
- 0
- 10,510
Hi Everyone,
Your help with this would be really appreciated. I know it's a long post, but I'm trying to give as much detail as possible. There's a tldr at the bottom.
I recently RMA'd a faulty HD7870 (the factory confirmed that it was faulty). I was sent a different model of HD7870 in return, and I noticed it ran quite hot (a fair amount hotter than the other GPU, which I highly doubt overheated).
BUILD:
OS: Windows 8.1
CPU: Intel i7 3770k
RAM: 2 x Crucial Sport 4GB, Ballistix 240-pin DIMM, DDR3 PC3-12800 Memory Module
GPU: Radeon HD 7870 (VTX 3D Black Edition)
Sound Card: Asus Xonar D2X
PSU: Corsair Gaming Series 2013 Edition GS 600W ATX/EPS 80 PLUS Bronze
Motherboard: Gigabyte Socket-1155 Z77-DS3H Motherboard
THE SYMPTOMS:
1. I was playing Tomb Raider (the new-ish one) and my computer crashed. I noticed a slight reduction in fps that the GPU seemed to do when it was throttling due to overheating, and chose to continue playing to see if the GPU would sort itself out. It had been fine with about 5-6 hours of gameplay on Tomb Raider before this, and probably about 4 hours of gameplay on the horror game 'Anna'. Perhaps this was a mistake: the screen froze with vertical brown stripes on a white background, and the sound was stuttering with this screen. The computer was unresponsive, so I had to do a hard restart.
2. When I restarted my PC, I got the initial motherboard screen followed by the Windows logo, but where I would normally be asked to log in a screen didn't show. After several restarts,the Wingows 8 start up problems dialogue came up, and I system restored to before the problem. This fixed nothing; I had the same issues. After a successful start-up I noticed a few odd things; Avast needed reinstalling (the hard restart may have caused a problem) and I couldn't find my crash logs (Bluescreenview helped me diagnose a problem before, but for some reason it can't find any minidumps anymore, and I don't know why but there is no 'Minidump' folder in C:/Windows - any help in simply finding what caused the crash would be really useful).
3. It seems that the computer thinks the GPU still gives out a signal. The GPU definitely gets power (the fan whirs), and my motherboard doesn't let internal graphics kick in when the GPU is plugged in. I can only get a screen when my GPU is unplugged.
EXTRA INFORMATION;
-As soon as I received the card I tested it in Furmark a bit. It ran quite hot, and seemed to stabilise around 85 degrees Celcius. Furmark is demanding software, so I presumed the GPU wouldn't normally be under this load. Also, I noticed that the GPU activated throttling above about 65 degrees; my fps would drop from about 44 to 33 between 60 and 75 degrees.
-I have not OC'd the card.
-My case has a lot of air flow: a decent Zalmann CPU cooler, along with 5 fans from good brands. I even taped another 80mm fan (again from a good brand) to the side of the case to blow directly onto the side of the GPU when I noticed it ran hot. A sound card sat about two inches below the GPU fan and may have blocked some airflow.
MY TAKE
Well, it seems likely that the GPU was fried, but shouldn't my computer should shut down before permament heat damage was done? I had been gaming with the GPU for at least 8 hours before this occurred, with at least two longer gaming sessions that I can think of before this which were absolutely fine. If overheating is what happened I'd be very unimpressed that a new GPU in a case with decent airflow would have such terrible cooling (though I have mentioned the sound card perhaps reducing airflow around the GPU). The GPU definitely isn't completely dead: when it is plugged in the fan still runs, and the card still gets hot despite giving me no picture.
TLDR
¬¬My GPU might have fried. Does anyone have some advice? It would be especially useful if someone could tell me how to find the crash dump that might have been created when my computer crashed, or recommend some way of fixing or diagnosing the card. Also, would I be able to get away with a second RMA? Or was this really stupid of me and not covered by a warranty?
Thank you so much for any answers, these GPU problems are driving me mad.
I should add that I fairly thouroughly tested my RAM and CPU because of the previous faulty GPU: Ram- memtest, tried one stick at a time. CPU - some CPU testing software I can't remember the name of.
Your help with this would be really appreciated. I know it's a long post, but I'm trying to give as much detail as possible. There's a tldr at the bottom.
I recently RMA'd a faulty HD7870 (the factory confirmed that it was faulty). I was sent a different model of HD7870 in return, and I noticed it ran quite hot (a fair amount hotter than the other GPU, which I highly doubt overheated).
BUILD:
OS: Windows 8.1
CPU: Intel i7 3770k
RAM: 2 x Crucial Sport 4GB, Ballistix 240-pin DIMM, DDR3 PC3-12800 Memory Module
GPU: Radeon HD 7870 (VTX 3D Black Edition)
Sound Card: Asus Xonar D2X
PSU: Corsair Gaming Series 2013 Edition GS 600W ATX/EPS 80 PLUS Bronze
Motherboard: Gigabyte Socket-1155 Z77-DS3H Motherboard
THE SYMPTOMS:
1. I was playing Tomb Raider (the new-ish one) and my computer crashed. I noticed a slight reduction in fps that the GPU seemed to do when it was throttling due to overheating, and chose to continue playing to see if the GPU would sort itself out. It had been fine with about 5-6 hours of gameplay on Tomb Raider before this, and probably about 4 hours of gameplay on the horror game 'Anna'. Perhaps this was a mistake: the screen froze with vertical brown stripes on a white background, and the sound was stuttering with this screen. The computer was unresponsive, so I had to do a hard restart.
2. When I restarted my PC, I got the initial motherboard screen followed by the Windows logo, but where I would normally be asked to log in a screen didn't show. After several restarts,the Wingows 8 start up problems dialogue came up, and I system restored to before the problem. This fixed nothing; I had the same issues. After a successful start-up I noticed a few odd things; Avast needed reinstalling (the hard restart may have caused a problem) and I couldn't find my crash logs (Bluescreenview helped me diagnose a problem before, but for some reason it can't find any minidumps anymore, and I don't know why but there is no 'Minidump' folder in C:/Windows - any help in simply finding what caused the crash would be really useful).
3. It seems that the computer thinks the GPU still gives out a signal. The GPU definitely gets power (the fan whirs), and my motherboard doesn't let internal graphics kick in when the GPU is plugged in. I can only get a screen when my GPU is unplugged.
EXTRA INFORMATION;
-As soon as I received the card I tested it in Furmark a bit. It ran quite hot, and seemed to stabilise around 85 degrees Celcius. Furmark is demanding software, so I presumed the GPU wouldn't normally be under this load. Also, I noticed that the GPU activated throttling above about 65 degrees; my fps would drop from about 44 to 33 between 60 and 75 degrees.
-I have not OC'd the card.
-My case has a lot of air flow: a decent Zalmann CPU cooler, along with 5 fans from good brands. I even taped another 80mm fan (again from a good brand) to the side of the case to blow directly onto the side of the GPU when I noticed it ran hot. A sound card sat about two inches below the GPU fan and may have blocked some airflow.
MY TAKE
Well, it seems likely that the GPU was fried, but shouldn't my computer should shut down before permament heat damage was done? I had been gaming with the GPU for at least 8 hours before this occurred, with at least two longer gaming sessions that I can think of before this which were absolutely fine. If overheating is what happened I'd be very unimpressed that a new GPU in a case with decent airflow would have such terrible cooling (though I have mentioned the sound card perhaps reducing airflow around the GPU). The GPU definitely isn't completely dead: when it is plugged in the fan still runs, and the card still gets hot despite giving me no picture.
TLDR
¬¬My GPU might have fried. Does anyone have some advice? It would be especially useful if someone could tell me how to find the crash dump that might have been created when my computer crashed, or recommend some way of fixing or diagnosing the card. Also, would I be able to get away with a second RMA? Or was this really stupid of me and not covered by a warranty?
Thank you so much for any answers, these GPU problems are driving me mad.
I should add that I fairly thouroughly tested my RAM and CPU because of the previous faulty GPU: Ram- memtest, tried one stick at a time. CPU - some CPU testing software I can't remember the name of.
Facebook
Twitter
Instagram