I think my video card is about to kick the bucket.

[TickleMyElmo]

Hello all. So lately my MSI Gaming X Trio RTX 2080 has been acting strange. The last few days it will be working fine and then the fans all the sudden decelerate to 40% from my normal fan curve of 100% that I set for 50c and up. Then the screen will go black and nothing. I have to hard reset my pc and then all works again for a few hours. I have been keeping eyes on temps with afterburner and they never go above 62c or so. Could it be all the crazy windows update bugs lately? I just did an uninstall with DDU and downloaded the newest drivers. I will let you guys know if it happens again but I'm almost certain it will given the time.
Anyone have a clue or idea? Thanks.

 

[Dr. McCoy]

From what you described it does sound like a driver crashing, I've had that happen before in the same way with my stuff causing a reboot.
Chances are that's the issue and updating the driver was a good move. Update bugs due to Windows is another possible thing, can't rule that out.

In all likelyhood the card itself is fine.
Still possible and I'd just keep an eye out for this to increase in severity/frequency and if it does, the card could well be saying bye-bye.

 

[EarthDog]

I wouldnt count the card out yet either..

If the driver reinstall doesnt work, try it on a different known stable system. If it works there, then we know it is something on your rig.

Does this happen in one game? All games? Etc...?

 

[TickleMyElmo]

All games after so long. Sometimes 30 min., others 4 hrs. So far I am good since the new driver installation. We will see.

 

[freeagent]

Could be the 5ghz at 1.28v.

 

[TickleMyElmo]

I guess. I wouldn't think CPU degredation would start already. But I guess anythings possible.

 

[TickleMyElmo]

Well it was fine all night last night after the driver update but today it is doing it again every 30 min. Guess I'll have to return it again.

 

[TickleMyElmo]

So I took it to Micro Center to return it and they ran furmark for 15 min. without any problems. Said it sounded like a power delivery problem. So I upped the cpu voltage from 1.29 to 1.30. Now I wait and see what happens. If it acts up again I'll have to bring in my whole pc. This heavy ******* is going to be a pain to drag around. While I was there I bought 2 more sticks of G.Skill trident Z 3200 ram, the royal ones since I planned on using them in my next build anyhow. $140, good deal, I say. So at the moment I have 32 gigs of ram. I doubt it will make any difference in gaming but it's good to have nonetheless.

 

[EarthDog]

Furmark, lol.

Furmark throttles the card and doesnt run it close to the clock speeds. Sure it uses similar power, but not clocks.

So, they say a power delivery issue, and you then RAISED the voltage on your system? You are aware that uses more power, right? Not that I believe them anyway...but that was a counterproductive move to their suggestion, even if it isnt by much.

Save your overclock profile and go back to stock and see if you are stable.

Edit: so, to be clear, you are having system issues and nkw have added more RAM? The point is to try and isolate the issue not add more variables such as dropping in mismatched ram sticks........you are shooting yourself in the foot by your testing methods.

 

[Zerileous]

Agreed. On one hand it was nice of them to test it at all. And testing it in another system is better than not. But I don't think 15 minutes of Furmark tells you anything useful other than that the card boots in a different system.

 

[trents]

As has been said, return everything to stock frequencies and voltages and see if the problem disappears.

 

[TickleMyElmo]

Everything has been fine since upping the voltage. I also lowered the frequency to 4.9 from 5.0. Yea, I laughed when they said Furmark as well. I told them to run Heaven but they wouldn't.

 

[wingman99]

Also they only ran the testing for 15 minutes and you had the problem every 30 minutes.

 

[tekjunkie28]

If it happens again take off the OC. I quit doing OCs mostly. I get tired of dealing with the crashes and the search for stability. I will OC older stuff but never newer stuff. Electro-migration is too much of a problem with smaller nm products.

 

[Zerileous]

I mean everyone is welcome here, but you do realize what forum you're on right?

 

[EarthDog]

And, I'm not worried about electromigration... Dialing in an overclock today is easier than ever, really.

 

[tekjunkie28]

Lol yea I know. I’ve done my fair share of OC but not much anymore. Especially now that I have an AMD. If I had an Intel I probably would do some OC. Haswell really ruined overclocking for me. I hated how dumb adaptive voltage was and I’m not a fan of blasting 1.3v through a processor for 24/7. My 4670k would spike voltage by 0.1v under heavy load. The last thing I need is to fry my cpu. I have to have dang near 100% reliability now days.

 

[TickleMyElmo]

Alright. I'm bout sick of this pc. I'd take a hammer to it if I had the cash to replace it. I took out the new ram, and set everything to stock. Temps are hitting 90+, not good. I have no idea what else to do.

 

[EarthDog]

Alright. I'm bout sick of this pc. I'd take a hammer to it if I had the cash to replace it. I took out the new ram, and set everything to stock. Temps are hitting 90+, not good. I have no idea what else to do.

I'm confused...

Did the problem this thread is about (the black screen/fans go 100%) get resolved and now it's a temp issue?

Basic troubleshooting man... is the pump running? What action gets it to 90C? What voltage is it sending to the CPU at stock? What have you done so far to resolve the issue? More information would be helpful, please....

 

[TickleMyElmo]

The black screen was fixed but then came the 90+ temps. According to corsair link the pump was working fine, so were the fans. the voltage never went over 1.28. All I did was reset the pc a few times and it was fixed. (for now).