PC reboots under stress

cheeseandpickles · Feb 19, 2014

Hi

A couple of months ago my PC started to reboot under stress (usually of a game,) upon starting up again it would give me an overheating message. I eventually realised I should check the cpu temps and they were massively high, so I bought a new heatsink (as opposed to stock) and more thermal paste, which gave me temps of about 50 Celsius idle. I thought this had fixed the problem, but I'm still getting reboots, now without the overheating message.

My assumption is that the power supply is either on the way out, or it's not giving my system enough power, however I'm probably wrong. Another thing to note, my power supply was second hand, and occasionally on starting up it will be louder than usual and not allow boot, just staying in its initial state. This is remedied with restarting it again.

System specs:
CPU: i7 920
GPU: HD 6950
PSU: OCZgxs1010
Motherboard: Asus P6T SE

I appreciate any help so thanks.
Olly

nukemaster · Feb 19, 2014

I do not think 50c idle is that good for that cpu. Even the stock cooler should have managed that. It is load temperatures you want to see.

Please run a short game test with something like HWmon running. Do not run long enough to crash, you just want an idea of how hot it is getting(and without logging the crash would loose this data).

http://www.cpuid.com/softwares/hwmonitor.html

And yes, an old second hand power supply could be an issue. Not ways, this is why you have to do some testing.

thx1138v2 · Feb 19, 2014

Overheating can leave some damage while not completely destoying a chip. Over power causes higher resistance in the circuit which causes it to draw more power which causes more heating and it feeds on itself until you get shutdown. When it cools down it may still have higher resistance if permanent damage was done. So you may need to really hike up your cooling to overcome that new level of resistance.

A bad power supply can cause these problems and since everything in the box runs off of that device you are in danger of toasting everything in the box. Replace it now.

cheeseandpickles · Feb 19, 2014

I agree the idle temp isn't that good, but it's significantly better than what it was. Having just ran a few minutes of Far Cry 3 HWmon is giving my mins of 45,45,48,43 for the different cores, from just before running the game, to max 62,61,61,57. I'll upload a screenshot of it in a sec.

As for the the second answer, this was my primary concern, I do feel that the previous overheating has had long term effects.

Edit: http://i.imgur.com/NqwUj4q.png

nukemaster · Feb 19, 2014

Modern cpus should shut down before actual damage will occur[and before this they tend to clock down quite a bit to try to cool off first]. many users have run for months with temperatures in the 90's and even over 100 without actually damaging a cpu(I am sure the life was slightly shortened).

Do you have access to a friends power supply to test with. Failing that, Can you test some of your parts on that system(video card/memory ect) on another system.

Your temps look fine for that cpu and video card. If HWmon is getting accurate voltages(it is not always accurate) from your system the 12 volt looks low. What does the bios show for your 12 volt line?

cheeseandpickles · Feb 19, 2014

OK, unfortunately I don't really have the means of testing with a different power supply or for testing other components. Well bios showed about 12.4 for the 12. I'm inclined to think it is a power supply issue, just with the strange boot difficulties and the fact there's never a conclusive error logged for a reboot.

Search

PC reboots under stress

cheeseandpickles

Reputable

nukemaster

Titan

thx1138v2

Splendid

cheeseandpickles

Reputable

nukemaster

Titan

cheeseandpickles

Reputable

TRENDING THREADS

Latest posts

Moderators online

Share this page