• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

MSI RTX 2080 Ti pink artefacts on screen, after RMA for the same issue !

Associate
Joined
20 Oct 2011
Posts
674
I returned a MSI Duke 2080 Ti back to MSI Poland for an RMA, due to the card showing pink artefacts on screen, card was returned back to me fairly quickly around 8 day turn around, which was pretty fast, but the card came back , it worked for around 2 hours then crashed, now exactly the same issue......tested the card in my two PC's , my GTX 1080ti works fine in both machines.....so it is not an issue with my system

Is the pink artefacts a fairly common RTX 2080 ti issue ?

I have contacted MSI to request a collection of the card, Should I request they exchange the card for another of the same cards but perhaps a different batch ? the card returned to me was my original card.
 
Can you upload a screenshot (not photo) of the pink artifacts?

Tried reinstalling graphics driver I imagine, but how about clean install of Windows? Searching for this issue pulled up some results where pink artifacts were driver related. Worth looking into on the off-chance that the GPU is fine so you don't need to return it again.

Also, manufacturers often test cards vertically on test benches, not horizontally as most cases have them. Could try laying your case on its side, to rule out a sag issue of some sort. I would re-insert the card after you lay the case on its side as well.
 
Hi Danny, thank you for your reply, the card was delivered yesterday to my office and was tested vertically, the drivers are the latest revision.

I didn't take any screenshots or photos, but I can do but might be Monday before I get a chance.

The card worked perfectly for around two hours, I left the card running heaven benchmark, the temps where good no issues,the screen froze, rebooted then it was back to the original issue the card was sent for RMA, which I'm guessing would suggest it is not a Windows or Driver issue, as the card ran fine for around 2 hours.

I can tell that the card received back was the same one sent and the RMA paperwork points towards that too... I am getting the feeling that perhaps only a small component has been replaced on the graphics card, but the real problem my be a little deeper.
 
Last edited:
I had a look on Google to see if I could see a similar issue with a photo, the link below looks similar

https://bit.ly/2I2itt5

One other thing I did note when reading about other people's similar issues, if the card is running at 800 by 600 no artifacts are displayed, but once the resolution goes above 800 x 600 then the issues begin.

Still waiting on a reply from MSI with regards to returning, I hope they are not going to expect me to pay a further charge to send the card to them, I am assuming that they will collect the card this time at their cost.
 
Last edited:
Could you potentially see if your psu has more than 1 12v rail to feed the card? I would be doing a fresh windows install and trying different rails or a 1000w psu with a single large rail in this circumstance as 2080ti memory errors are usually a weird space invaders shape not pink like that.
 
Could you potentially see if your psu has more than 1 12v rail to feed the card? I would be doing a fresh windows install and trying different rails or a 1000w psu with a single large rail in this circumstance as 2080ti memory errors are usually a weird space invaders shape not pink like that.

Ty for your reply, the PSU is a Corsair RM 1000 (can't remember if is an x or an i) and runs my gtx 1080ti perfectly with no issues or signs of artefacts.

The card worked fine for 2 hours after return from RMA.
 
Last edited:
Ty for your reply, the PSU is a Corsair RM 1000 (can't remember if is an x or an i) and runs my gtx 1080ti perfectly with no issues or signs of artefacts.

The card worked fine for 2 hours after return from RMA.

Ah yeah single rail 83A yeah it won't be the PSU. I'd say you must reinstall windows 10 from scratch at this point to ensure you don't get hit with charges when the card comes back next time.
 
Ah yeah single rail 83A yeah it won't be the PSU. I'd say you must reinstall windows 10 from scratch at this point to ensure you don't get hit with charges when the card comes back next time.

But why would it work fine for 2 hours and then my gtx 1080ti work fine afterwards, all on the same install of Windows? Windows seems to be working fine.

I've tried the graphics card driver uninstaller..
 
But why would it work fine for 2 hours and then my gtx 1080ti work fine afterwards, all on the same install of Windows? Windows seems to be working fine.

I've tried the graphics card driver uninstaller..

You are right that it is probably vrm cooling or something on the card, testing a fresh install is the only other way you can progress the investigation without going straight back to RMA unless you have a way to monitor vrm temps under load. I'm assuming vrm temps after 2 hours just tipped it unstable. Can you get further than 2 hours with fans at 100% to prove a cooling issue?
 
I love when people jump straight to reinstall windows, makes no sense in this case imo.

If another card works perfectly then it isn't Windows.

It sounds to me like MSI tested the card, probably for about 10 mins, found nothing and so sent it back to you 'fixed', even though they likely did nothing.

RMA again and point out to them the fact that it's the second time you are returning the card for the same fault. Make them aware that you don't think they fixed (did anything) last time and that the problems arise after extended period (not 5 mins on a test bench).

Hopefully, you will get a replacement next time.
 
You are right that it is probably vrm cooling or something on the card, testing a fresh install is the only other way you can progress the investigation without going straight back to RMA unless you have a way to monitor vrm temps under load. I'm assuming vrm temps after 2 hours just tipped it unstable. Can you get further than 2 hours with fans at 100% to prove a cooling issue?

I might get chance to reinstall windows on a spare 120gb ssd drive, the 2080ti Duke I believe is based on the founders edition card with just bigger coolers and an extra fan.

I wonder what they fixed on the card to enable the card to work for a short while? As I presume the fault is deeper in the hardware as opposed to perhaps replacing something fairly low level.
 
Last edited:
I love when people jump straight to reinstall windows, makes no sense in this case imo.

If another card works perfectly then it isn't Windows.

It sounds to me like MSI tested the card, probably for about 10 mins, found nothing and so sent it back to you 'fixed', even though they likely did nothing.

RMA again and point out to them the fact that it's the second time you are returning the card for the same fault. Make them aware that you don't think they fixed (did anything) last time and that the problems arise after extended period (not 5 mins on a test bench).

Hopefully, you will get a replacement next time.

Maybe the one jumping without doing any research is you. Nobody is saying the graphics card isn't the likely culprit, and in fact I believe it is. But troubleshooting as much as possible, since you have more time with your card than the manufacturer, can sometimes go a long way.


recently I have been getting an occasional start-up issue. On a cold start the screen would display pink artifacts in a pattern form. Resetting the computer seems to fix the issue. Graphic card or drivers?

update: updated flash for IE manually(probably nothing to do with my issue) but also updated to the new driver and so far symptoms have vanished as far as the strange graphic bug goes.
Windows logs was error free for a change...Impressive!

update final : no more issues with latest driver, solved.

https://forums.geforce.com/default/topic/879626/pink-artifacts/

So in that case - driver. Now imagine there's still some left-over driver conflict somewhere in current Windows install. What's wrong with being thorough, instead of playing "my guess is better than yours"?

Even if it doesn't solve anything (which is a long shot) some clues might appear while troubleshooting that will help them track down the specific issue.
 
Maybe the one jumping without doing any research is you. Nobody is saying the graphics card isn't the likely culprit, and in fact I believe it is. But troubleshooting as much as possible, since you have more time with your card than the manufacturer, can sometimes go a long way.

Really? You should read the OP. He tested the card in two machines, both showing the same symptoms. Are you saying both Windows installs are at fault? This testing methodology is sound, especially since he also tested another GPU in both machines without issue. Everything points to a faulty card and wasting time and effort to reinstall Windows is pointless.
 
Maybe the one jumping without doing any research is you. Nobody is saying the graphics card isn't the likely culprit, and in fact I believe it is. But troubleshooting as much as possible, since you have more time with your card than the manufacturer, can sometimes go a long way.




So in that case - driver. Now imagine there's still some left-over driver conflict somewhere in current Windows install. What's wrong with being thorough, instead of playing "my guess is better than yours"?

Even if it doesn't solve anything (which is a long shot) some clues might appear while troubleshooting that will help them track down the specific issue.

On Monday I shall use a spare SSD to install a totally fresh version of Windows 10 Pro and see if it makes any difference, if nothing else I can at least have peace of mind that I tried that and to be honest, Windows 10 tends to install pretty quickly, I shall download the latest version from Microsoft.
 
Really? You should read the OP. He tested the card in two machines, both showing the same symptoms. Are you saying both Windows installs are at fault? This testing methodology is sound, especially since he also tested another GPU in both machines without issue. Everything points to a faulty card and wasting time and effort to reinstall Windows is pointless.

Yes, really.

Justwondering doesn't mention if the second machine had a fresh Windows install when testing the 2080Ti.

His previous 1080Ti is fine, and the testing with the 1080Ti proves the drivers on both machines are fine... for the 1080Ti.


if nothing else I can at least have peace of mind that I tried that

Exactly. And guess what? If you return the card and MSI says they find no fault and that they are returning it and to reinstall Windows, you can now tell them you already tried that. It's as much about anticipating such fob-offs and cornering them, as about trying on the off-chance for a fix.
 
Yes, really.

Justwondering doesn't mention if the second machine had a fresh Windows install when testing the 2080Ti.

Agreed it is best to try everything, I know I mentioned it previously, it is the part where the card worked just fine for 2 hours running heaven benchmark on a loop, the drivers appeared to of installed correctly, but I will still try the fresh install on Monday.

Thank you all for your help and advice.

His previous 1080Ti is fine, and the testing with the 1080Ti proves the drivers on both machines are fine... for the 1080Ti.




Exactly. And guess what? If you return the card and MSI says they find no fault and that they are returning it and to reinstall Windows, you can now tell them you already tried that. It's as much about anticipating such fob-offs and cornering them, as about trying on the off-chance for a fix.
 
Yeah, and if you can also lay the case on its side then you can tell them you've tested in same orientation like they did, and it still had the error.
 
Yeah, and if you can also lay the case on its side then you can tell them you've tested in same orientation like they did, and it still had the error.

Hi, my case is on its side in my office, I really hope the fresh install does the trick, I would rather the issue be software related, otherwise I'm not going to be left with much peace of mind with the card in the future, if it does go back via RMA and they don't swap the card for the same type.
 
otherwise I'm not going to be left with much peace of mind with the card in the future, if it does go back via RMA and they don't swap the card for the same type.

Yup. The other issue is - what are your options for RMA'ing through the retailer you bought it from? Because that's where you have rights* like one attempt at repair and if that fails then refund or replacement must be issued if customer doesn't agree to second repair. The manufacturer has no obligation in that sense.

* Depending on the country.
 
Yup. The other issue is - what are your options for RMA'ing through the retailer you bought it from? Because that's where you have rights* like one attempt at repair and if that fails then refund or replacement must be issued if customer doesn't agree to second repair. The manufacturer has no obligation in that sense.

* Depending on the country.

I bought it from a MSI reseller that sells returned cards, they have no more stock, so it's either RMA or a refund... I'd rather RMA to be honest due to the cards price.
 
Back
Top Bottom