• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

A tale of an unsuccessful EVGA Graphics Card RMA

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Tech Tweaker

Member
Joined
Dec 14, 2010
So, this story begins back in December of 2017. My 980 Ti burned out an Inductor/Choke, and the video out suddenly went to a black screen and there was a smell of the mysterious electronics smoke in the room.

Had to wait a week or two to RMA, because EVGA's Customer Support staff and their entire RMA department were on vacation for Christmas/New Years, not that I'm begrudging them for having time off, but it was an inconvenience being down a video card and having to use an old 660 Ti for a week and a half.

Replacement Card #1 (1080)... Got a new card a week or so after I sent my failed card back to the manufacturer. This card worked for about a month, then started displaying strange issues such as random frame rate dips and video lag. I put up with that for a few weeks, up until it randomly started having the video crash to a black screen while I was just sitting at the desktop checking my email and/or browsing the internet (with the message: "video driver has crashed and recovered"). The driver started crashing three to five times a day, then it was just ridiculous. Failed Firestrike Stress test miserably with a whopping 94.5% on the regular test, and less than 91% on the Ultra Stress Test IIRC. I don't know if stress tests in Firestrike are all that definitive of actual stability, but I know it was crashing at the desktop while not even gaming.

Replacement Card #2 (1080)... Sent first replacement card back, waited another week. New card came, took out of box and installed in my main PC. This card was crashing with the same symptoms as the first card in less than 12 hours of installing it. Tested card in another system in case my PC was the problem, same symptoms in another system random black screens (even more frequently than the first card) and failing Firestrike stress tests. Reinstalled OS on main PC in case OS was a problem (which was an inconvenience to do with having to reinstall so many programs (20+)), and in the end that didn't even help. Sent second card back, for another RMA.

Replacement Card #3 (1080)... So, after reinstalling my OS to make sure that wasn't a problem one would think I could get a working card. Let's see how this one goes, after waiting another week. Installed card in two of my backup systems, fails Firestrke Stress Tests in both. Merely attempted to run a benchmark in NH/NHML, failed almost immediately and went to a black screen with message "display driver failed and has recovered" when the display came back up to the desktop.

Am I just being unreasonable, or have I gone crazy? What do I have to do to get a working card?

EVGA seems to be unable to supply an actual working card that operates within specifications without randomly crashing. I would have probably accepted this if the third replacement card actually worked, and likely wouldn't have made this Post, but it doesn't. Three defective cards in a row is a little ridiculous, but believe it or not I'm not making this up. I'm still trying to get a working graphics card, and the warranty for my original card is nearly up at this point.

My previous RMA several years ago was flawless and was sent a working product in return for a failed one, this RMA though is less than successful with three flawed and/or defective cards in a row sent to me during the duration of trying to complete ONE RMA and get one working card.
 
Last edited:

UltraTaco

Member
Joined
Oct 28, 2017
I read the whole thing, sir and wow, at first I thought perhaps a power supply, but you mentioned it was same in another system. It's hard to comprehend what kind of luck it may be to either have 4 bad cards from a bad batch or two computers both having problems :shrug:

I would just give up at that point and go back to mechanical type writer. Really sorry to hear such problem
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
Firestrike stress test? NH/NHML? What are those? Can you play a game? Try different drivers? Try the card(s) in a secondary slot?

Did you track SN's?
 
OP
Tech Tweaker

Tech Tweaker

Member
Joined
Dec 14, 2010
Firestrike stress test? NH/NHML? What are those? Can you play a game? Try different drivers? Try the card(s) in a secondary slot?

Did you track SN's?

Firestrike is a GPU benchmarking software made by Futuremark Inc.

NiceHash Miner/Nice Hash Miner Legacy.

Tried four different driver versions, three different computers, and four different power supplies. I'm more thorough than you assume.

I've also tried all PCI Express slots on every board in every PC I tried them in capable of support video out, no difference.

Every card I got had a different SN.
 
Last edited:

UltraTaco

Member
Joined
Oct 28, 2017
Yes sir, I watched your other posts(nothing creepy), so I figured you're quite seasoned in troubleshooting:) this is crazy though, I don't know how this may be even possible. I believe you though.
 

ihrsetrdr

Señor Senior Member
Joined
May 17, 2005
Location
Close to the San Andreas Fault
I don't think that EVGA is necessarily bound to send a 'new' replacement, but perhaps a card that had been returned and repaired/tested. Maybe cryptomined to death by previous owner.
 
OP
Tech Tweaker

Tech Tweaker

Member
Joined
Dec 14, 2010
I don't think that EVGA is necessarily bound to send a 'new' replacement, but perhaps a card that had been returned and repaired/tested. Maybe cryptomined to death by previous owner.

I don't expect it to be new, I just expect a replacement to actually work.

Also, says in their TOS for RMA that most replacement units are Refurb'd or Repaired units. Apparently not tested that well though, if my experience is anything to go by.
 

wingman99

Member
Joined
Dec 10, 2003
I don't expect it to be new, I just expect a replacement to actually work.

Also, says in their TOS for RMA that most replacement units are Refurb'd or Repaired units. Apparently not tested that well though, if my experience is anything to go by.

Is it a possible monitor problem?
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
NiceHash Miner/Nice Hash Miner Legacy.

Tried four different driver versions, three different computers, and four different power supplies. I'm more thorough than you assume.

I've also tried all PCI Express slots on every board in every PC I tried them in capable of support video out, no difference.

Every card I got had a different SN.
we can only go by what was posted.. ;)
 

Bill Dimwit

Disabled
Joined
Mar 21, 2018
I don't think that EVGA is necessarily bound to send a 'new' replacement, but perhaps a card that had been returned and repaired/tested. Maybe cryptomined to death by previous owner.

Yes EVGA sends second hand cards back for replacements. That is not a big deal, the card you send them is likely not new when you RMA it.
But they need a better system of testing the cards before sending them to you. I revived many junk EVGA cards back from RMA just like the OP.
They all worked under lighter loads, but under a full load or maxing out the VRam and the cards would crash. I had similar problems with their B stock.
 
Last edited:
OP
Tech Tweaker

Tech Tweaker

Member
Joined
Dec 14, 2010
Is it a possible monitor problem?

Unlikely, I've tested three different monitors (2x Dell U2414H, 1x Samsung S22A300B) with this card. Tried DisplayPort, HDMI, and DVI cables for video output. Unless my monitors are somehow incompatible with the GTX 1000 series/GTX 10 series.

These same monitors work fine with my GTX 970, GTX 660Ti, GTX 580, and GTX 470. And worked with my 980Ti, before it suffered from an electrical component failure.
 
Last edited:

jeffyt

New Member
Joined
Jul 12, 2019
Wow.. scary similar experience here with 2080ti from EVGA. Also remember getting a DOA 1080ti when stepping up from 1070... apparently didn't learn my lesson then and had to go through 5 2080ti's hahaha
Reading your post made me relapse, lmfao. Worst part was "Am I just being unreasonable, or have I gone crazy? What do I have to do to get a working card?" such a horrible state of mind to be in... hope you got everything working in the end.

If you don't mind... did you end up replacing any other hardware? Did you ever determine if it was your hardware causing any issues or was it just bad re-certified cards from EVGA?
 

jeffyt

New Member
Joined
Jul 12, 2019
5 dead 2080ti's replaced!:clap: you nearly backrubbed evga!
lmfao, with all the hassle/headache/paranoia they put me through to end up being nothing other than bad cards, not sure if I was actually the one standing behind...
but to be fair, the first one (brand new) had one bad/dead fan on arrival and second one (brand new) lasted 6+ months but had a rare flicker that evga couldn't find/detect anything wrong with. only 3 dead 2080ti, all "re-certified" :escape:
 

Janus67

Benching Team Leader
Joined
May 29, 2005
lmfao, with all the hassle/headache/paranoia they put me through to end up being nothing other than bad cards, not sure if I was actually the one standing behind...
but to be fair, the first one (brand new) had one bad/dead fan on arrival and second one (brand new) lasted 6+ months but had a rare flicker that evga couldn't find/detect anything wrong with. only 3 dead 2080ti, all "re-certified" :escape:
What was wrong with the other 3? Seems like an awful lot.

 

jeffyt

New Member
Joined
Jul 12, 2019
What was wrong with the other 3? Seems like an awful lot.
Hahah are you sure you wanna dive in to that mess? I'll try to keep it as short as possible.
First RMA, re-certified card from EVGA died after 2 weeks, artifacting.
Second one died after 8 days, artifacting once again.
Before getting the 3rd, I replaced the PSU.
Third one died within 2-3 days, artifacting once again.

Since then, I have ran multiple stress tests over 5 different GPUs. Two most recent have been two different Gigabyte 2080ti.
The most recent is a Gigabyte Windforce that has lasted since May 30. No other hardware/software/OS changes have been done.
All three EVGA re-certified cards died with mediocre usage (maybe 2-3 hours of Division 2 daily at most)
Followed up both Gigabyte 2080ti cards with extensive testing. First one 11 days and second one 30 days. 6/7 hour daily Heaven stress tests with 100% gpu usage.

Here are the full details: https://forums.evga.com/Want-honest-opinion-about-my-four-2080ti-RMA-m2953621.aspx
 
OP
Tech Tweaker

Tech Tweaker

Member
Joined
Dec 14, 2010
Wow.. scary similar experience here with 2080ti from EVGA. Also remember getting a DOA 1080ti when stepping up from 1070... apparently didn't learn my lesson then and had to go through 5 2080ti's hahaha
Reading your post made me relapse, lmfao. Worst part was "Am I just being unreasonable, or have I gone crazy? What do I have to do to get a working card?" such a horrible state of mind to be in... hope you got everything working in the end.

If you don't mind... did you end up replacing any other hardware? Did you ever determine if it was your hardware causing any issues or was it just bad re-certified cards from EVGA?

I tried three different ram kits, four power supplies (even monitored the rails on all four with a digital multi-meter and rails on all PSU's were stable), three different computers, and three different monitors. All the hardware I used displayed the same issues, only if I used these defective GTX 1080 GPU's.

Eventually I gave up and repaired the third GTX 1080 myself (by that point the original warranty was up anyway, and I had a theory as to what the problem was), it did stop crashing after I repaired the graphics card.

Either these 1000 series cards were more susceptible to this type of failure than previous series cards, or they were all somehow damaged by the same type of usage/strain/stress. I had never seen this many failed cards in my possession from any previous series (GTX 600, 700, 900, etc.)

I've had failed AMD GPU's too (they were worse, no video output) HD 4870, HD 7970, R9 390. AMD seems to be more prone to the Black Screen or no video output issue, at least in my experience.
 
Last edited:

jeffyt

New Member
Joined
Jul 12, 2019
I tried three different ram kits, four power supplies (even monitored the rails on all four with a digital multi-meter and rails on all PSU's were stable), three different computers, and three different monitors. All the hardware I used displayed the same issues, only if I used these defective GTX 1080 GPU's.

Eventually I gave up and repaired the third GTX 1080 myself (by that point the original warranty was up anyway, and I had a theory as to what the problem was), it did stop crashing after I repaired the graphics card.

Either these 1000 series cards were more susceptible to this type failure than previous series cards, or they were all somehow damaged by the same type of usage/strain/stress. I had never seen this many failed cards in my possession from any previous series (GTX 600, 700, 900, etc.)

I've had failed AMD GPU's too (they were worse, no video output) HD 4870, HD 7970, R9 390. AMD seems to be more prone to the Black Screen or no video output issue, at least in my experience.
Thank you for the update.

Very crazy... I've never experienced so many failed cards with any other series as well. Not sure if they don't have a proper way to stress test these 'repaired' cards but they need to just send them back to nvidia to inspect/repair at that point. Recycling them through RMA is a horrible way to do it imo. By the way, if you don't mind, what was the fix for your card?