whea_uncorrectable_error in gaming or GPU test

[Renan]

My PC (2 years of use):

Windows 10 x64

AMD FX-6300
Gigabyte ga-78lmt-s2
2x4gb de RAM
Sapphire r9 280 3gb
EVGA 430W
HDD 1Tb

Since last week my PC has been crashing every time I try to play.
If I play CS: GO or LOL, I can finish a match most of the times, but if I enter on a more demanding game like NBA 2k17 the PC crash as soon as start the 3D part. (don't crash on menu).

When the crash happens, it does not even show BSOD.
The screen goes to a single color, always with the color that was most prevalent in the screen just before crashing.
And the PC restart.

With WhoCrashed.exe I can see that the .dmp was created in the Minidump folder.
And it shows that the error is WHEA_UNCORRECTABLE_ERROR.

With AIDA64 I tested the stability and temperatures in Full LOAD.
Everything is normal.
Run Prime95 for 1 hour, no errors.

But when I'm going to test my GPU with FurMark, it crashes in less than 30 seconds.

I already cleaned the PC and reconnected all the cables.
I installed video driver from Novemver/2016, when everything worked fine.
All drivers are up to date, BIOS included.
Reset BIOS to default
I never overclocking anything.
I even underclocked my video card and disable Turbo Boost of my FX6300 (not work)

Here's the Dump File if anyone wants to help: https://onedrive.live.com/?authkey=%21AIHZLEm2i_kzAnI&c...

Thanks in advanced.

P.S. Sorry my english, is 75% google translator

 

[jboydgolfer]

same error on different site reports Votage issues on the GPU.

post OP mentioned flashing Bios, Not that i recommend that step unless as a last resort(unless your capable , and knowledgable ofc), but i figured id mention it.

i see you posted Stock bios, ...maybe a different cause then...sounds like instability related tho

what are your temps during? before the issue?

 

[eidairaman1]

Pull the Card apart and replace the thermal compound on the GPU, and any thermal interface material such as pads on the VRMs and Memory. Try again. If still acting up see if the card is still in RMA Warranty (free replacement, no charge).

If Not- Grab a 470 and Call it a day.

 

[Renan]

same error on different site reports Votage issues on the GPU.

post OP mentioned flashing Bios, Not that i recommend that step unless as a last resort(unless your capable , and knowledgable ofc), but i figured id mention it.

i see you posted Stock bios, ...maybe a different cause then...sounds like instability related tho

what are your temps during? before the issue?

It's the same as it was before.
With AIDA64 stress test GPU stabilizes in 161 ºF and CPU stabilizes in 149 ºF.

When it crashed in the NBA 2k17, the GPU was at 64 °C (147 ºF) and the CPU at 58 °C (136 ºF) on MSI afterburner.
And with Furmark it's even colder, because it doesn't even give time to start to heat up and already crash.

 

[eidairaman1]

It's the same as it was before.
With AIDA64 stress test GPU stabilizes in 161 ºF and CPU stabilizes in 149 ºF.

When it crashed in the NBA 2k17, the GPU was at 64 °C (147 ºF) and the CPU at 58 °C (136 ºF) on MSI afterburner.
And with Furmark it's even colder, because it doesn't even give time to start to heat up and already crash.

What Brand and Model of Case are you using? The CPU seems too hot too, I'm using an 8 Core that is overclocked to 5.0 GHz and it stays lower in temps than that 6300.

What revision of motherboard are you using?

It looks like there is no heatsinks on the VRM stack- not good.

Are you using the factory heatsink on the CPU?

FYI furmark is a terrible test to use on ANY GPU- quit using it.

 

[Renan]

What Brand and Model of Case are you using? The CPU seems too hot too, I'm using an 8 Core that is overclocked to 5.0 GHz and it stays lower in temps than that 6300.

what revision of board are you using

Are you using the factory heatsink on the CPU?

FYI furmark is a terrible test to use on ANY GPU- quit using it.

oops, it's not 161 ºF, it's 151 ºF

I have a cougar mx300.
And cooler Xigmatek Loki II

The MB is rev 1.0 (with F6 BIOS).

I do not think the temperature is too high. I live in Brazil, here it is 30 ºC (86 ºF) even at night.

And like I said, with 30 seconds testing at Furmark, it's not enough time to heat up.

BTW, which 3D test do you recommend?

 

[eidairaman1]

oops, it's not 161 ºF, it's 151 ºF

I have a cougar mx300.
And cooler Xigmatek Loki II

I do not think the temperature is too high. I live in Brazil, here it is 30 ºC (86 ºF) even at night.

And like I said, with 30 seconds testing at Furmark, it's not enough time to heat up.

58C on a 6300 is way too hot. plus no cooling on the VRMs is causing system instability.

http://www.tomshardware.com/reviews/radeon-r9-280x-third-party-round-up,3655-4.html

Unigen Valley/Heaven

what do you think @cdawall @fullinfusion?

 

[jboydgolfer]

I agree, furmark is a pretty terrible load to put on a GPU thats throwing what could be thermal errors ,its way more than stress while playing regular games. Until you isolate the issue I'd recommend not running that

Imagine if your cars engine was possibly damaged, and to figure out the problem you held down the throttle 100% for extended periods of time. It's equivalent to furMark

 

[Renan]

58C on a 6300 is way too hot. plus no cooling on the VRMs is causing system instability.

http://www.tomshardware.com/reviews/radeon-r9-280x-third-party-round-up,3655-4.html

Unigen Valley/Heaven

FX6300 should be able to handle 70 ºC.

My never surpassed 68ºC in full load.

My GPU never surpassed 72 ºC.

 

[eidairaman1]

FX6300 should able to handle 70 ºC.

My never surpassed 68ºC in full load.

My GPU never surpassed 72 ºC.

at somepoint it did but you cant see it.

Trust me that CPU is running too hot. Plus the VRM set on that motherboard has no place to dissipate the heat.

Drop the CPU clock speed back by lowering the multiplier first by 2.

we need a screenshot of your bios Voltages too.

Clean out all of the fans in the case to include the CPU heatsink and fan and the GPU heatsink and fan too

 

[Renan]

So, I lowered the CPU clock to 2.4ghz.

And I ran a test on the Unigine Valley in 1080p, High Quality, without AA.
GPU temp stabilized on 66º C.
Crash in about 8 minutes.

we need a screenshot of your bios Voltages too.

I can't do that today
But the only one that is a little off is the 12v, it is v12.240, I think is normal.

Clean out all of the fans in the case to include the CPU heatsink and fan and the GPU heatsink and fan too

Already did that. I changed the CPU thermal as well.

 

[jboydgolfer]

So, I lowered the CPU clock to 2.4ghz.

And I ran a test on the Unigine Valley in 1080p, High Quality, without AA.
GPU temp stabilized on 66º C.
Crash in about 8 minutes.



I can't do that today
But the only one that is a little off is the 12v, it is v12.240, I think is normal.



Already did that. I changed the CPU thermal as well.

Just a quick question it seems like you're battling two Fronts here , the GPU and the CPU. Why don't we try removing 50% of the problem and then tackling one @a time , through process of elimination .

Do you have a second or spare GPU lying around? If so try installing that and then testing the integrity of the CPU. Once that testing is done and you have verified that the CPU is not the problem then you can add the video card back in and begin testing of it.

Even if you don't have a spare video card does your CPU have a integrated GPU? All you need is video output while you test the integrity of the CPU.

It just seems like it's a bit of a ping-pong ball game back-and-forth here.

It seems that instead of looking for the solution to the problem,we need to figure out where the problem is first, atleast that is how I troubleshoot personally.

 

[Renan]

Just a quick question it seems like you're battling two Fronts here , the GPU and the CPU. Why don't we try removing 50% of the problem and then tackling one @a time , through process of elimination .

Do you have a second or spare GPU lying around? If so try installing that and then testing the integrity of the CPU. Once that testing is done and you have verified that the CPU is not the problem then you can add the video card back in and begin testing of it.

Even if you don't have a spare video card does your CPU have a integrated GPU? All you need is video output while you test the integrity of the CPU.

It just seems like it's a bit of a ping-pong ball game back-and-forth here.

It seems that instead of looking for the solution to the problem,we need to figure out where the problem is first, atleast that is how I troubleshoot personally.

Do you think the problem is not the PSU?
Like, if I run the test on the onboard video and work fine, can't it be the PSU that can't provide enough power for the whole system when it's in full load?

Yes, I have onboard video.

With which program do I test the integrity of the GPU? Because I run Prime95 for one hour (with GPU plugged) and did not crash.

 

[cdawall]

58C on a 6300 is way too hot. plus no cooling on the VRMs is causing system instability.

http://www.tomshardware.com/reviews/radeon-r9-280x-third-party-round-up,3655-4.html

Unigen Valley/Heaven

what do you think @cdawall @fullinfusion?

Everything in the rig is super cheap. Psu or board imo, but then again the 280 is known for bad ram.

 

[Vya Domus]

That error is typically linked to an unstable overclock , check the CPU voltages in BIOS if they are set on auto or something else , and monitor the voltages using HWmonitor they should jump from 0.9 to 1.425 at stock . If for some reason your motherboard fails to set correct voltages automatically try setting it manually to something between 1.35-1.38 with turbo core off and see what it does.

As far as temps , note that what most programs report as CPU temp is actually the socket temperature , package or core temp are the actual temperatures of the CPU this one should be lower than CPU temp.

That being said 68C isn't a dangerously high temperature for the socket with a stock cooler but you don't use one , it does seem a bit strange the temps are that high , mine reaches 55C on the core and 65C on the socket max under load but it has a decent overclock. 65 on the core and 70 on the socket is what AMD considers safe for nonstop use. I would rule out CPU temperature for being the problem if temps really where the issue you would first experience throttling and then a complete shutdown assuming you haven't disabled thermal protection in the BIOS. I do agree that motherboard isn't the best to use for an FX CPU due to older chipset but it should handle a 6300 just fine at stock.

I would focus on the GPU and PSU , they might be failing.

 

[Renan]

Everything in the rig is super cheap. Psu or board imo, but then again the 280 is known for bad ram.

Here in Brazil the minimum wage converted into your currency (U$) would be U$260 per month.

A very cheap CPU here like AMD FX-6300 is U$140...

It is very difficult to have a good PC, with good quality without being too rich or stealing a bank.

That error is typically linked to an unstable overclock , check the CPU voltages in BIOS if they are set on auto or something else , and monitor the voltages using HWmonitor they should jump from 0.9 to 1.425 at stock . If for some reason your motherboard fails to set correct voltages automatically try setting it manually to something between 1.35-1.38 with turbo core off and see what it does.

As far as temps , note that what most programs report as CPU temp is actually the socket temperature , package or core temp are the actual temperatures of the CPU this one should be lower than CPU temp.

That being said 68C isn't a dangerously high temperature for the socket with a stock cooler but you don't use one , it does seem a bit strange the temps are that high , mine reaches 55C on the core and 65C on the socket max under load but it has a decent overclock. 65 on the core and 70 on the socket is what AMD considers safe for nonstop use. I would rule out CPU temperature for being the problem if temps really where the issue you would first experience throttling and then a complete shutdown assuming you haven't disabled thermal protection in the BIOS. I do agree that motherboard isn't the best to use for an FX CPU due to older chipset but it should handle a 6300 just fine at stock.

I would focus on the GPU and PSU , they might be failing.

I think I can get a PSU to test with my cousin.

But it will not hold the whole system, it's a 350W PSU.

Can I put it to power my MB, CPU and HDD, leaving my current PSU only for my r9 280?

If it works out, it will show that my PSU is not powering the PC under load, right?

 

[cdawall]

The PSU you have is a common rail using it to just power the 280 is not a good plan at all. Those need a load on all rails to provide even power.

 

[alucasa]

That specific error has something to do with voltage regulation.

I don't recommend BIOS flashing but this can only be fixed by BIOS flashing if there is a fix available.

 

[fullinfusion]

I've took a look at the thread and my opinion it's a weak PSU. Sounds like the exact issue I had a few years ago. Turned out I replaced the OSU and never had a single issue after that. Good luck.

 

[cdawall]

That specific error has something to do with voltage regulation.

I don't recommend BIOS flashing but this can only be fixed by BIOS flashing if there is a fix available.

If the voltage to the video card is not a stable ATX spec 12v and it bounces around it will act exactly like an unstable clock. The voltage regulation on the cards can only correct so much.

 

[Renan]

I've took a look at the thread and my opinion it's a weak PSU. Sounds like the exact issue I had a few years ago. Turned out I replaced the OSU and never had a single issue after that. Good luck.

"I think I can get a PSU to test with my cousin.

But it will not hold the whole system, it's a 350W PSU.

Can I put it to power my MB, CPU and HDD, leaving my current PSU only for my r9 280?

If it works out, it will show that my PSU is not powering the PC under load, right?"

Do you think I should try to do that?

 

[Vya Domus]

The thing is a PSU might show a problem only when you put a certain level of load on it , if you power only your video card it might work just fine but when you power the whole system it may not, unfortunately you need to use a different PSU for the whole system to be 100% sure. Before that did you check the CPU voltages ? Like I said that error is caused typically by unstable voltages , we should clarify this before you start testing components , I am insisting on this because I've seen motherboards dropping the voltage to a small value automatically if certain settings are are changed in the BIOS.

 

[Renan]

The thing is a PSU might show a problem only when you put a certain level of load on it , if you power only your video card it might work just fine but when you power the whole system it may not, unfortunately you need to use a different PSU for the whole system to be 100% sure. Before that did you check the CPU voltages ? Like I said that error is caused typically by unstable voltages , we should clarify this before you start testing components , I am insisting on this because I've seen motherboards dropping the voltage to a small value automatically if certain settings are are changed in the BIOS.

in the BIOS:
+1.5v = 1.5v
+3.3v = 3.335v
+12v = 12.240v

But with HWMonitor:

 

[Vya Domus]

Can't really determine anything yet , they do seem bit low tho, keep HWmonitor on and run the benchmark tool in CPU-Z and post another pic of this.

 

[cdawall]

Using HWmonitor for voltage monitoring is not a route you want to take.

"I think I can get a PSU to test with my cousin.

But it will not hold the whole system, it's a 350W PSU.

Can I put it to power my MB, CPU and HDD, leaving my current PSU only for my r9 280?

If it works out, it will show that my PSU is not powering the PC under load, right?"

Do you think I should try to do that?

NO

The PSU you have is a common rail using it to just power the 280 is not a good plan at all. Those need a load on all rails to provide even power.