Hi all - first post here but a lurker for many years. Unfortunately ive got a problem I simply cant fix, so here I am with a first post!
On Sunday night I thought Id just have a quick game of GTA 5. Fired it up as normal, played for about 20 mins then WHAM - total crash, no BSOD, back to BIOS and it rebooted. "Thats odd", I think, as it reboots. So I try another game - XCom. Same thing happens again, this time pretty much as soon as it starts rendering the 3d stuff at the menu. Now im concerned. So I try Dying Light as well. That lasts about 20 seconds then dies the same way.
The next day I think, lets give it another go. XCom. Same happens again. This only seems to happen when the GPU is put under load, so I do some normal browsing, YouTube, etc etc. All is fine - just GPU. So that makes me think, well it must be overheating or something. So I run three GPU-Z logs with Dying Light every time it crashes. These are the readings :
CRASH 1 - GPU Clock 1164.4, GPU MemClock 1752.8, GPU Temp 70c, Load 98%, Power Consump 129.9w, TDP 85.9%, CPU Temp 51c
CRASH 2 - GPU Clock 1240.3, GPU MemClock 1752.8, GPU Temp 72c, Load 99%, Power Consump 145.2w, TDP 96.0%, CPU Temp 60c
CRASH 3 - GPU Clock 848.0, GPU MemClock 1502.3, GPU Temp 46c, Load 0%, Power Consump 39.7w, TDP 26.3%, CPU Temp 54c
OCCT tests kill it within 10 seconds, generally. GPU temp off load right now : 33c
These dont seem like overheating temps to me, and ive had mammoth gaming sessions with a modded Skyrim install and had no heat shutdowns of any kind in the past. Ive used MSI Afterburner for years and always had a very slight OC on the card, only about 30MHz or so on the core clock and again had no issues. The PC case itself is an Alienware X51 R2, which many have put GTX970's in with the 330W PSU you get on those machines. But even then I think, maybe its a power thing. So I test the PSU (its an external 330W PSU like a laptop PSU as is standard with these older Alienware PC's). Voltage and wattage absolutely fine both on the kettle lead and the power brick itself. All Windows can say is that its an Event ID 41 - the usual for an unforseen crash due to power loss?
The next day it then decides after its latest crash that its going to get stuck in the W10 Infinite Repair Loop issue. So I then have to completely reinstall Windows from USB as it refuses to boot to Windows, System Restore doesnt work. Fresh install onto formatted SSD.
I then go through the following list of actions over the next couple of days :
System Config :
Alienware X51 R2 w/ A12 BIOS
i7 i7-4770 @ 3.4GHz
16GB Hyper Fury X RAM DDR3 1600MHz
256GB Sandisk SSD
Palit GTX970 4GB (Blower)
330W Dell External PSU
Cheers all!
Letto
On Sunday night I thought Id just have a quick game of GTA 5. Fired it up as normal, played for about 20 mins then WHAM - total crash, no BSOD, back to BIOS and it rebooted. "Thats odd", I think, as it reboots. So I try another game - XCom. Same thing happens again, this time pretty much as soon as it starts rendering the 3d stuff at the menu. Now im concerned. So I try Dying Light as well. That lasts about 20 seconds then dies the same way.
The next day I think, lets give it another go. XCom. Same happens again. This only seems to happen when the GPU is put under load, so I do some normal browsing, YouTube, etc etc. All is fine - just GPU. So that makes me think, well it must be overheating or something. So I run three GPU-Z logs with Dying Light every time it crashes. These are the readings :
CRASH 1 - GPU Clock 1164.4, GPU MemClock 1752.8, GPU Temp 70c, Load 98%, Power Consump 129.9w, TDP 85.9%, CPU Temp 51c
CRASH 2 - GPU Clock 1240.3, GPU MemClock 1752.8, GPU Temp 72c, Load 99%, Power Consump 145.2w, TDP 96.0%, CPU Temp 60c
CRASH 3 - GPU Clock 848.0, GPU MemClock 1502.3, GPU Temp 46c, Load 0%, Power Consump 39.7w, TDP 26.3%, CPU Temp 54c
OCCT tests kill it within 10 seconds, generally. GPU temp off load right now : 33c
These dont seem like overheating temps to me, and ive had mammoth gaming sessions with a modded Skyrim install and had no heat shutdowns of any kind in the past. Ive used MSI Afterburner for years and always had a very slight OC on the card, only about 30MHz or so on the core clock and again had no issues. The PC case itself is an Alienware X51 R2, which many have put GTX970's in with the 330W PSU you get on those machines. But even then I think, maybe its a power thing. So I test the PSU (its an external 330W PSU like a laptop PSU as is standard with these older Alienware PC's). Voltage and wattage absolutely fine both on the kettle lead and the power brick itself. All Windows can say is that its an Event ID 41 - the usual for an unforseen crash due to power loss?
The next day it then decides after its latest crash that its going to get stuck in the W10 Infinite Repair Loop issue. So I then have to completely reinstall Windows from USB as it refuses to boot to Windows, System Restore doesnt work. Fresh install onto formatted SSD.
I then go through the following list of actions over the next couple of days :
- reinstall the drivers / rollback to a previous set of drivers
- Completely reinstall all drivers recommended by manufacturer
- Enable / disable PCI3 Gen3 Support in BIOS
- Switch my RAM out to an older working pair that I know works / reseating and switch current RAM
- Remove all traces of dust by airdusting
- Reseat GPU
- Use DVI/VGA instead of HDMI
- Remove all other devices, USB and internal optical drive so its just SATA SSD and GPU left
- Update the BIOS (A06 to A12)
System Config :
Alienware X51 R2 w/ A12 BIOS
i7 i7-4770 @ 3.4GHz
16GB Hyper Fury X RAM DDR3 1600MHz
256GB Sandisk SSD
Palit GTX970 4GB (Blower)
330W Dell External PSU
Cheers all!
Letto
Last edited: