System crashes and reboots while gaming on 7950

[Orija]

Until a week or so back all games had been running just fine on my system. I had been playing some heavy stuff like BF3, Crysis 2 and Dishonored without any problems. It was after I took a break off gaming for a month or so that things got hairy yesterday. When playing games like Blacklist and Saints Row 4, the system just reboots, no BSODs and no glitching. Even games that used to run fine, like Shogun 2, are crashing now.

I checked the GPU temps before and after the crash and they were around 40C. I haven't made any changes hardware-wise.

Here are my system specs:
Sapphire 7950 3GB
Asus P8-Z77M Pro
Corsair 650TX V2
i5 2500k

Drivers: 13.4

Games run fine if I lower that resolution and play them windowed.

 

[Mushkins]

Sounds like a power issue. Last time I had that same thing happen, it was because I updated my video card and my PSU just couldn't handle the new load over extended periods of time. I'd get like an hour of gaming in and it would just reboot itself.

650w is plenty for that system, the PSU itself could be failing. Do you have another you can swap out to test?

 

[Orija]

Nope. Is there any way to test it?

 

[Jaydip]

Test a different set of pcie cables

 

[Orija]

Before I get knee-deep into cables and screws, could there be any other issue that's causing this?

 

[Jaydip]

It seems your system just can't handle the load so I believe it's a high chance that psu is the culprit

 

[Orija]

Alright, I'll test with a different set of cables. Thing is, the pc had been working fine for about four months. I stopped playing games for a month and now I get this.

It's kind a weird that this has actually happened to me earlier too. Some months back, games started crashing the system when I started gaming after a break of month or so. The issue resolved itself in a couple of days so I didn't give it much heed then.

 

[Orija]

The cables seem to be fine. Anything else I should be trying?

 

[Jaydip]

Is your 7950 a 6+6 or 8+6 pin model?

 

[Orija]

I think it's 6+6. Not sure though. It's this one: http://www.newegg.com/Product/Produc...82E16814202026

 

[Jaydip]

Yeah it's 6+6, so you tried a different set of pcie connectors and had the same result?

 

[Orija]

Yes.

 

[Jaydip]

Well looks like the PSU is dying, system reboot typically happens when PSU is the culprit.If it was VC you would get bsod,driver crashes, garbled texts and so on.

 

[Orija]

Is there a way to test the stability of the PSU using a software? I'd rather be sure about the PSU being responsible before going through the hassle of an RMA.

 

[Jaydip]

You can download AIDA64 and check for the 12V rail voltage.You cannot check noise and ripple without an oscilloscope.

 

[Orija]

Okay, here's the thing, the issue seems to have resolved itself. This has happened for the second time when running video games after a month long break cause the system to crash, for a few days. Any opinions on why this keeps happening?

 

[CuriousMike]

This is a case where having some spare hardware would be really useful.

Do you have a friend that has a beefy system? Try running your videocard in his system for awhile and verify it's OK.

Or, beg/borrow a new P/S for a few days and beat on it and verify it's OK.

 

[Lorne]

You dont have a location in your avatar info so not sure if this would be relavent, But have you stopped using AC or heating in you home.
Has the temp in your area cooled recently.

 

[Rikard]

While my bet is on the PSU, check the CPU temps. I had that behaviour on a work pc a while ago; the CPU cooler was full of dust...

 

[tweakboy]

Put video card to stock clocks, and CPU, stock everything. Then let us know what happens. gl

 

[Orija]

The CPU temps reached 66C before the system crashed while running TW Rome 2. Are these temps too high for a i5 2500k?

PS: In case the CPU or RAM were the culprit, would I be seeing a BSOD before the reboot?

 

[Rikard]

The CPU temps reached 66C before the system crashed while running TW Rome 2. Are these temps too high for a i5 2500k?

PS: In case the CPU or RAM were the culprit, would I be seeing a BSOD before the reboot?

No that is not too hot, but certainly hotter than I have with my i5 2500. I think the specified max temp is 73 C, and that is typically with a fair margin of safety. Depending on what the failure is, you could just get a system reset without any warning.

 

[Orija]

I disabled "Restart on System Failure", which hasn't stopped the system from rebooting. Running OCCT crashes the system whenever it uses the GPU for stress testing.

I tried running the same tests using the integrated Intel gpu, no crashes this time. Running stress tests on the CPU and Memory don't lead to any crashes either. Which has led me to inferences:

a)the gpu isn't able to properly take power from the psu
b)the psu is unable to properly supply power to the gpu

How do I find out whether it's the gpu or the psu?

 

[Rikard]

I disabled "Restart on System Failure", which hasn't stopped the system from rebooting. Running OCCT crashes the system whenever it uses the GPU for stress testing.

I tried running the same tests using the integrated Intel gpu, no crashes this time. Running stress tests on the CPU and Memory don't lead to any crashes either. Which has led me to inferences:

a)the gpu isn't able to properly take power from the psu
b)the psu is unable to properly supply power to the gpu

How do I find out whether it's the gpu or the psu?

The easiest is to try a different GPU with a similar power draw in the same box. Ask a friend if you do not have a spare. Or bring your GPU to your friend at let him try it in his PC.

 

[Orija]

Would a 9800GT do?