PC shutsdown while gaming and restarts after a few mins

Toggle sidebar Toggle sidebar
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630
Hello everyone,
I have been struggling with this build, off late my computer wouldn't power on at all, I tried all the steps of discharging the cabinet etc, then finally, I am not sure what worked, I took it out of the cabinet, removed the CMOS battery for a few mins, just had the CPU and RAM connected and it booted up.
I put it back together, and I sometimes have this problem at boot, I dont get video signal from my GPU, I have open it, take out the card put it back, and it comes on.
Boot to my desktop, and fire up some games, trying to Far Cry Blood Dragon, btw, I finished FC3 on this very build, I start gaming after about 20-30 mins (maybe more) the system shuts down and will restart after a few mins with the GPU fan running full speed! I then downloaded the FURMARK software, ran the burn in test at the default resolution, and I got a BSOD one time, other times after a few mins the system would shut down and restart, once the video signal went missing, while the CPU was powered on. While running FURMARK, the GPU would touch 90c, before you guys get excited, please note, this is India, the ambient temps here are high, and another point to note is that the room is air conditioned.

My system specs
I5 2500K
Asus P8Z68-V
Gskill ripjaws 2x4GB 1600
MSI 7970
256GB SSD
2x 1TB HDD
Corsair 750TX – this was RMA’d last year


Edit - Have attached the GPU sensor log, please ignore the 8/25 and 8/26 as they are with the onbaord GPU. Please take a look at the values and let me know if you guys see anything out of the blue.GPU log
 
Sort by date Sort by votes
vbright

vbright

Honorable
Mar 12, 2012
20
0
10,510
Hello jasku,
Sorry to hear you are having problems with your rig. I have a few questions to ask before I can attempt to help you.
Question 1: What are your CPU and GPU temps when playing video games?
Question 2: does this problem arise ONLY when playing video games?
 
Upvote 0 Downvote
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630


Hi, thanks for responding

1) The GPU while gaming touches 85C and the CPU around 60C, while running the benchmark though GPU went from idle of 58C to 90C in a minute.

2) Well, post my PC revival, yes, prior to that, I used to get random display driver has stopped responding and have been recovered, do also recall the odd reboot while watching media.

 
Upvote 0 Downvote
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630



Well, I am not sure if 58C idle is bad, because the fan speed is at 20%, if jacked to 50%, its comes down to 50C, but its like a jet plan in my room. I could take apart the card, wondering if it would void the warranty, as it is still under warranty.
 
Upvote 0 Downvote
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630


I wonder why you ask that, when I took out the mobo from the cabinet a couple stand off came unstuck with the screws, I dint think they would have anything to do with the odd behavior. I just put em back, a couple of them dont have screws.
 
Upvote 0 Downvote
vbright

vbright

Honorable
Mar 12, 2012
20
0
10,510




GPU idling @ 58C is a bit on the high side but airflow and ventilation can play a part in this and adding more case fans may help but make sure your GPU fan is actually working, I am sure it is but it is always a good idea to inspect your GPU just as Pconsoles has suggested. As for the 90C temps are concerned, is this a stress test you are running using a benchmark program?
 
Upvote 0 Downvote
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630
That is correct, it is called Furmark, just running the standard 'burn-in' test. I want to make sure that there is no hardware problem here. Do you guys think the computer restarting after a few mins, with the GPU fan blazing, points towards overheating issue?
 
Upvote 0 Downvote
Ponyface

Ponyface

Honorable
Jan 23, 2013
217
0
10,760
im sorry i misread "gpu" as "cpu"! i would not try change the thermal paste on the gpu. Perhaps try older version drivers for your card? sudden shutdown without any error message, bsod, freezing ,etc. means it could well be heat causing a problem.
 
Upvote 0 Downvote
vbright

vbright

Honorable
Mar 12, 2012
20
0
10,510


It is normal to see GPU temps at or around 90C during a stress test such as Furmark. These programs are meant to push your GPU to its limits and most video games will not come close to that, hopefully. I would suggest testing your memory first. I assume you are running Windows 7.

To test your memory in Windows, first open the Start Menu and type in mdsched.exe, then press enter. This will open a pop-up window asking how you'd like to test the memory. The first option will restart your machine and check the memory right then and there, the second option will check the next time you choose to reboot. After selecting your desired option, your computer will load a screen that shows the progress of the test as well as the number of passes on the memory (this test can take some time). Keep an eye out on the memory diagnostic tool for errors. No errors means your RAM is more than likely not the root of the cause. This will mean further troubleshooting.
 
Upvote 0 Downvote
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630


I apologize for not mentioning this in the OP, but I did run the memtest and it returned no errors, again from what I have been reading over the net, this may not be fool proof, did return no errors for whatever its worth!

 
Upvote 0 Downvote
Mahisse

Mahisse

Distinguished
Nov 26, 2012
941
6
19,315
First you do not mention anything about overclocking so I guess we all can assume that you have not overclocked your card? If you have, take it back to stock clock.

You live in India, where you have high temperatures and I pressume also have very high humidity as well. Temperatures can be dealt with by additional cooling but high humidity is much more critical for you PSU, video card and all other components for that matter.

When you are not getting a signal from you video card, is the card's fans still spinning?

You say that you've played FC3 without any issues before. Did you do any driver upgrades since then? If so try to roll back.

In all honesty I believe your video card is faulty or overheating. Have you tried taking it out and using the integrated video card and see if your system runs fine then?

One thing you have to remember when dealing with cards that reaches high temperature is that the temperature that is being monitored is the temp for the GPU but the temperatures on the VRAM is just as crtitical. If you have a card running around 85 celsius, chances are that you VRAM is moving towards critical temperatures as well and tha can definitely cause errors on the video card. Even though it sounds like a jet plane in your room when the fans are spinning 100%, try having it on from the very beginning to see if that will result in fewer crashes.

If you can get a fan that can get some fresh air directly on the GPU this would probably also help.

Also if your card is still under warranty, you could consider to just RMA the card.



 
Upvote 0 Downvote
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630


Thank you for your response.

Your assumption is correct, there is no overclocking involved here.

There used to be the odd restart or game would quit to the desktop with the driver has recovered error. I would tend to agree with your, but I want to make sure it is the GPU, please help me validate prognosis. I want to be sure, because the RMA process in India is nothing like the west.

I have the GPU log, will post it once I get home, you will get an idea of the temperatures and voltage till crash.
 
Upvote 0 Downvote
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630
Hello all, I have attached the GPU log in the OP and have since taken out my 7970 and using the onboard video, seems to be working fine, not sure how to stress test this setup, ran the same Furmark, it seems to run fine for as much time, which my 7970 did not make it more than 60 seconds.
 
Upvote 0 Downvote
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630


Well, I played FC:BD for an hour with the manual fan speed set to 60%, the load temps did not cross 72C and it ran fine, I am guessing it was the temps. But again, is it normal for the card to jump from 50C idle to 74C during load, shouldn't the jump be more gradual, or is it cause the card take more juice and hence the temp also immediately spikes?

 
Upvote 0 Downvote
vbright

vbright

Honorable
Mar 12, 2012
20
0
10,510


I would say it is normal depending on card type and card fans, My card idles at 34C and when I start up BF4, the temps climb rather quickly to around 70-75C. You should be ok.
 
Upvote 0 Downvote
Mahisse

Mahisse

Distinguished
Nov 26, 2012
941
6
19,315
That's very normal behaviour so don't worry about that at least. If you can keep your card under 80 celsius that would be ideal for you I think.
So you are just running 60% fan speed all the time now? The reason why I ask is because your GPU log also indicates that your card ran 60% fan speed when running full load on the card but it managed to get to 93 celsies anyway.
 
Upvote 0 Downvote
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630


Good observation, prior to using my onboard GPU, I had the video settings to max, but when I was trying out the onboard GPU, it reset the video settings on FC;BD, when I went back to my 7970, I only upped the resolution to full HD, and did not turn on all the eye candy, maybe that is why the difference in temps.

Having said that, do you guys think I should RMA the card?
 
Upvote 0 Downvote
Mahisse

Mahisse

Distinguished
Nov 26, 2012
941
6
19,315
Any Antia Aliasing related settings will effect your GPU performance drastically. You don't really need more than x2 AA in newer games if you are already running 1080p in my opinion.
If you had AA maxed out in your previous settings I wouldn't be surprised if that was causing the high temperatures due to the extra performance needed.

Also the shadow setting (AO, HBAO and HDAO) has a big impact on the performance. For AMD card users they should use HDAO. HBAO is for Nvidia users.

It's hard to say if you should RMA the card or not at this point because Far Cry 3 does become unstable with maxed out settings on systems that cannot handle it. Also you want to make sure that your driver installation is not corrupted so if you haven't done it before, you need to completely uninstall the old drivers before installing the new one. You can use Driver Sweeper for that.

If you experience similar issues with any other games I would RMA the card right away though.
 
Upvote 0 Downvote
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630


Thank you for responding.

I was playing BF3 and after about 30mins it shutdown, and restarted after a few mins, this is getting really annoying, this the temps were also under control.

I am thinking it could be mobo? I keep getting weird messages while booting, from settings have changed, to chasis intrude, and new CPU installed.

I did not mention these earlier, but faced new CPU installed today, have no idea whats going, any help in this matter would be greatly appreciated.

Here is the log, please look at 8/29 around 01:15am
 
Upvote 0 Downvote
J

jasku

Distinguished
Nov 18, 2006
52
0
18,630


I havent updated to the latest version, because the release notes talk about improving performance of some of the controllers, I didnt think it would impact the stability and hence did not update.

I am also worried about the RMA, if it is 2 year, I may be out of luck, anyways let me update the BIOS and report back.
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom