PC Freezing Issue (NON-BSOD)

[Shrub Nuts]

Hey all,

 

I recently have (within the last two days) found that my desktop PC is consistently crashing.  The crash is unlike anything I have ever seen before and I am leaning towards thinking this could be some kind of imminent GPU death sign.  Essentially my PC will almost at random drop everything on both of my monitors and simply display one color on one and one on the other. The colors themselves seem like random colors based on an average of the most displayed color on the screen but often one is black and one is white. It is incredibly disheartening because Minecraft of all things is what seemed to make this problem become evident.  I had been playing Cyberpunk 2077 on the previous days after release and despite how it runs on my age old 1080 TI it never caused this. I just yesterday and today played some Minecraft for the first time in a long time and now it seems to be a consistent occurrence regardless of what I am doing. I ran some stress tests from 3Dmark and it ran for 10 minutes without stalling, I quit out of it by hitting escape and while it was loading results it happened again. I am now just leaving my desktop up to see if it does it while idling with just my background standard stuff running and no interaction at all. Attached is an image of the monitors in their frozen state.

 

Specs are listed here:

Gigabyte X570 Aorus Elite

AMD Ryzen 3800X

Zotac Nvidia GeForce GTX 1080 TI Amp Edition

2 Sticks of G.Skill TridentZ Neo 16GB 3600 Mhz (18-22-22-42) Memory

Sound BlasterX AE-5 Sound Card

(A Bunch of drives but I don't think they are an issue unless you think so I can share those as well)

 

 

Please any help as to what you think is going on would be greatly appreciated. These are lowly times to buy a GPU which blows but if I need to I will scrounge for some second hand stuff.

 

[CommanderAlex]

What are your GPU temperatures like?? Have you recently updated your drivers at all? If so, roll back to the previous version to rule out the drivers are not causing you issues. 

[WhitetailAni]

I had this issue too.

Turned out it was my card overheating (my cooler was loose after repasting) so you could try repasting and remounting the cooler.

Once my idle temps weren't 71C it stopped.

[Shrub Nuts]

17 hours ago, CommanderAlex said:

What are your GPU temperatures like?? Have you recently updated your drivers at all? If so, roll back to the previous version to rule out the drivers are not causing you issues. 

So my temps were fine and I did actually just update to the latest a few days ago I will try rolling back cause I would like to save the card though I did something else too.

I have an old 980 TI that I popped in there and replicated the same thing.  I launched Minecraft (since this was the game that caused it) and ran the game fine for until the 980 TI got hot. HWInfo read out that the 980 TI was peaking at 73 degrees. That old thing runs really hot, but it ran the whole time. I closed Minecraft and waited for the dreaded monitors freezing to occur. Waited for 20 minutes and nothing. Shut it down and put the 1080 TI back in. Launched the game and did the same thing ran till my 1080 TI leveled out at about 68 degrees.  I don't think 68 is too hot, but it didn't climb any higher. It stopped there and I noticed the game was running like shit. Like 25 FPS bad. ( I have shaders and stuff sure but the game ran at 80 FPS just fine earlier) I shut the game down and within 15 seconds, both monitors hard froze two different colors.  I did just repaste recently, but I don't think its temp related.

 

I ran FurMark for 30 minutes; the GPU failed to climb past 44 degrees Celsius. I was blown away, FurMark was running like hell and the GPU was barely heating up at 100% load. That didn't make sense. I took out my heat gun and blow some warm air around the GPU to see if the sensors are bad, but it climbed up and reached that magic 65 degree plateau.  Also to my surprise when the card warmed up FurMark went from 4 FPS to 30. I have no idea what the hell that was.

 

I don't know if this information helps, but I can repaste and see what happens and put the cooler on differently. That isn't a problem, though I did get a deal on a used 2080 SUPER so now I have a backup if I can fix this up. (Also I know the 2080 SUPER is basically a 1080 TI but it was dirt cheap so its fine on me if it holds me till 40 series cards drop)

[Shrub Nuts]

17 hours ago, ragnarok0273 said:

I had this issue too.

Turned out it was my card overheating (my cooler was loose after repasting) so you could try repasting and remounting the cooler.

Once my idle temps weren't 71C it stopped.

I have more information in the reply above, I am going to try this too though just to see. Worth a shot at least.

[Guest]

If you are running the XMP profile for the memory, try turning that off and see if the issue reoccurs. 

[Shrub Nuts]

1 minute ago, ShrimpBrime said:

If you are running the XMP profile for the memory, try turning that off and see if the issue reoccurs. 

I can try that too, though I should mention I ran AIDA64 for 45 minutes stressing CPU and RAM and let the computer sit with my standard background apps. No crashes at all. I can still try this though and see what happens.

[Guest]

1 minute ago, Shrub Nuts said:

I can try that too, though I should mention I ran AIDA64 for 45 minutes stressing CPU and RAM and let the computer sit with my standard background apps. No crashes at all. I can still try this though and see what happens.

The frequency is too high with this particular memory cpu and board combo. 

You could try 3200mhz instead and see if that brings better stability.

[Shrub Nuts]

Alright everyone. I tried several things, but my problem was still the same. Quit Minecraft, 15 seconds later dead screens. My girlfriend is letting me borrow her 1070 and sure enough I try to recreate the issue with no problems at all. Also this card has the loudest coil whine sound I have ever heard.  I think I just lost the long lasting lottery and my card actually died within two years. It just doesn't work the same.

6 hours ago, ShrimpBrime said:

The frequency is too high with this particular memory cpu and board combo. 

You could try 3200mhz instead and see if that brings better stability.

Also I had a question for you beside the crash if you don't mind.  The RAM I have is CL 18 and 3600 do you think I could squeak out 16 CL if I underclock the speed of the memory? 

[Guest]

2 hours ago, Shrub Nuts said:

 

Also I had a question for you beside the crash if you don't mind.  The RAM I have is CL 18 and 3600 do you think I could squeak out 16 CL if I underclock the speed of the memory? 

It's possible yes. You have to do the testing to find out what it likes.