Do YOU have Problems with YOUR RTX card?

[Slade]

EVGA is NOT pre-testing them. My replacement was a brand new unit fresh from the factory. Unless they are "testing" them before packaging at the plant, they are just outright replacing them with available stock. I still had to do the MCU firmware flash update which has been available since the start of November if that is any clue to the state of the replacement cards.

 

[Legendary Gamer]

Can always be a driver or game related issue that affects the 20 series.

It's not game related because the general screen flickering is happening on the desktop and more noticeably while watching videos on things like Amazon prime and Netflix. I have seen the 2080Ti perform flawlessly in the same games and watching video and general operation, so It's not the drivers and it's the card. People that won the silicon lottery have cards that work. People that didn't are stuck with dead/dying cards and even persistent issues like I am.

Card #1: My first card had CTD issues, instability, screen tearing, flickering, massive artifacting all of it started happening withing a couple hours of installation and use.

Card #2: The second card initially seemed ok, then it started the flickering, gradually getting worse, screen tearing appeared on day 3-4 of use, atrifacting has been minor overall but I have noticed it here and there. It's begun to stutter and chop in games, so my frame rates are all over the damn place on this card. I would imagine that it is leading up to becoming unstable and CTD or worse.

The above are no driver related issues I have ever encountered. It's not isolated to simply gaming, general use will cause the issues to appear. Doesn't matter if I'm in a word processor, watching youtube or playing games, the issues persist. If I drop in any other video card, the issues cease immediately. There are big problems in this release, across every product in the 2000 series. Maybe the 2080, 2070 and eventual 2060 cards don't fail as often as the 2080Ti's do... However, this generation of cards is failing at a massively higher rate than anything I have seen in a really long time.

My first PC was an AMD 386DX-40, I have been building systems a REALLY long time. I have seen electronics smoke, light on fire and even a motherboard that turned into a pool of goo around one of the early generation Pentium processors. I have seen graphics cards fail in different ways before too. For all of the graphics cards that I own or have owned, I have never encountered a failure rate like this one and persistent issues like these. I have pretty much owned one or two of every major release Nvidia has made over the years and more than a couple ATI/AMD boards, in addition to all the oddball shit I have bought as well like 3DFX Voodoo1, Voodoo2, Voodoo3, Matrox, S3, etc.

From what I have experienced, first hand, as well as read here and elsewhere... Nvidia released some alpha/beta grade horseshit into the wild.

 

[Legendary Gamer]

EVGA is NOT pre-testing them. My replacement was a brand new unit fresh from the factory. Unless they are "testing" them before packaging at the plant, they are just outright replacing them with available stock. I still had to do the MCU firmware flash update which has been available since the start of November if that is any clue to the state of the replacement cards.

That's good to know, because they keep telling me they are "testing and burning them in" which I took to be bullshit since both of my cards have had the same issues and they were apparent in relatively short order.

 

[TheRagnarok]

I wish it was an unstable CPU causing my problem, but I popped in the Asus 2080ti and continued gaming like nothing has changed. After work today, I saw I had a EVGA reply email to try clean driver install, reset my bios, install approved drivers, flash to a new bios etc. I could not even get the evga card to get past the post screen, let alone windows. I think when the memory starts going, it REALLY goes. Also in the short span I had it in there to test, the card warmed up significantly. The asus card was in their since last night and on all night and day and wasn't close the the warmth of the dead card when compared to the few minutes it ran. I suspect once its dying, its a runaway effect.

So my issue returned and it was my PSU! I ran a PSU calc tool and my recommended PSU wattage was 740w. I heard the PSU fan ramp up a few times then, bam! I got another hard shutdown.
Set the cpu back to stock and it was fixed again..... bought a Seasonic 1200w Platinum and oced cpu again, rock solid.

 

[brentsg]

So my issue returned and it was my PSU! I ran a PSU calc tool and my recommended PSU wattage was 740w. I heard the PSU fan ramp up a few times then, bam! I got another hard shutdown.
Set the cpu back to stock and it was fixed again..... bought a Seasonic 1200w Platinum and oced cpu again, rock solid.

I'm not closely tracking each individual saga here, lost track.. But this is the benefit of keeping an open mind. These cards are definitely more problematic than most, but the solution is sometimes not as straightforward as it seems.

Glad you got it sorted.

 

[cybereality]

Yeah, totally. I had many problems with the 2080 Ti and I think the PSU was one of them, along with my memory speed.

Was getting frequent BSODs while playing games (usually within a few minutes). Replaced the PSU and the BSODs pretty much went away, but I was still getting crash to desktop in games.

After lots of troubleshooting, and finding that each individual component seemed to be working, I dialed back the OC on my system and that fixed it.

Ultimately it ended up being the RAM speed OC that was problematic. I had it on 4133, the XMP profile for my sticks, which was no problem with GTX 1080, but adding the 2080 Ti caused this to an issue.

I reset CMOS, which set the sticks to 2133. Ran this for about a week, no problems at all. Then I put the XMP profile back on, but with a slight voltage bump on the RAM and that was it. Been running solid for like 2 months with no problem.

Really, explore every option. Just because something was working before (in my case, the memory speed) doesn't mean it will continue to work if you make hardware changes to the machine. Test everything.

 

[Legendary Gamer]

Well, card number 3 shows up today. Should be interesting.

 

[Furious_Styles]

Well, card number 3 shows up today. Should be interesting.

3rd time's a charm!

 

[Slade]

So my issue returned and it was my PSU! I ran a PSU calc tool and my recommended PSU wattage was 740w. I heard the PSU fan ramp up a few times then, bam! I got another hard shutdown.
Set the cpu back to stock and it was fixed again..... bought a Seasonic 1200w Platinum and oced cpu again, rock solid.

I'm running a 1200W EVGA PSU so its not for a lack of power that my card died. Asus card is back in my other system churning away experiments while I game on the EVGA replacement one. The replacement fired up into windows like the system never missed a beat. Even precision X1 shows a correct 2080 ti XC where with the asus card just says 2080ti. I noticed when the evga card was dying, precision x1 could NOT detect it was a 2080 ti xc, it read it like it was a generic card. Something to note when your card starts dying if it "forgets" its identity.

 

[VF31Tomcatters]

Maybe I'm just lucky but my video card has been problem free.

 

[Slade]

There are some lucky ones out there. It's not a 100% failure rate lol. I'm hovering 50%, my first 2080ti is still in excellent working order.

 

[Snowdensjacket]

I've got my new 2070 replacement since the first one went space Invader. Finally got it all running and I've realized something.

I'm coming from a crap 470 from AMD I bought two years ago to upgrade my last ATI or whatever card.

My last card ran loud. I mean the Duke 2070. It was loud. I didn't get to use it for long but it was LOUD. The new card? It's quiter than my 120mm case fans I don't even notice it.

I was never able to run any tests or get any real data, but anecdotally I think my space invader 2070 was running the fans at 100% or close to it all the time. I think perhaps it had thermal issues.

I'll be running a bunch of benchmarks over the weekend fingers crossed this thing won't fail.

 

[Legendary Gamer]

I just installed the new card, my system failed to post immediately after installing it. I restarted to no video and then I restarted again and I'm at my desktop. I'm attempting to remain positive and I'm chalking up the initial oddities to "Space Magic". No odd flickering or graphical issues yet, I am loathe to say the card is stable until I actually do some sort of load testing or game play. There are no BIOS updates for this card, I checked. It's likely because it's a release after the initial wave of FE cards (I was hoping for a magical BIOS update for the card after reading Slade's post). It's almost 2 AM here, I need sleep but I will try some light gaming on this for a bit and see what the hell shakes loose.

Update: I just played about an hour and a half of Battletech at max settings at 4K. I had one initial game crash that I'm on the fence about, that put me on edge. After that, a solid 1.5 hours of FAWLESS gameplay. No glitches, no artifacts, no flickering, solid and fluid FPS, no stuttering. Sweet Mother of unmerciful Nvidia... It took three cards to get here. I'm still going to remain critical as all hell about the card. However, maybe, just maybe, I finally got a good one.

The other two both manifested tangible issues almost immediately: Screen Flickering, stuttering, glitches/artifacts, etc.

My experience has been pretty good, and that's about all I'm willing to grant this card in terms of compliments until I can verify it's been stable for at least a month.

 

[Legendary Gamer]

This old man has a hot date tonight, lol. So, I won't be doing heavy load testing on card. I did notice what I would call an artifact a couple hours ago tho. Damn near anything that looks out of the ordinary sets me on edge with this series of Nv cards. I will update you guys later after I've had some alone time with... the 2080Ti, the girl, is gonna come first tho!

 

[FrgMstr]

This old man has a hot date tonight, lol. So, I won't be doing heavy load testing on card. I did notice what I would call an artifact a couple hours ago tho. Damn near anything that looks out of the ordinary sets me on edge with this series of Nv cards. I will update you guys later after I've had some alone time with... the 2080Ti, the girl, is gonna come first tho!

Let me know when it goes Space Invaders on you, I will buy it.

 

[jamse101]

Picked up Palit 2080 Ti dual from post office at Thursday 13th, gave it 50+ hours testing during week and now this Friday it started giving me Space Invaders.

 

[Slade]

Replacement in since tuesday. No problems yet. Benchmarks are fine, micron memory still. Cross fingers is stays this way.

 

[brentsg]

Kids’ new Evga 2070 XC Ultra has Samsung VRAM for what it’s worth.

Also I’m a bad dad so I didn’t buy a DisplayPort cable. Thought I could use the DisplayPort to DVI cable that came with my 2080 Ti but that’s single link and their old monitor is dual link.

 

[Shadowed]

Anyone that signed that NDA is not going to be reporting about defective cards or a possible issue. At best you may get a passing article, however if you check the Nvidia forums then you see many have issues.

Do we have a list of who signed the NDA? Or maybe it is easier to find a list of who DIDN'T sign the NDA. All I know is Kyle did not sign it.

 

[Legendary Gamer]

Do we have a list of who signed the NDA? Or maybe it is easier to find a list of who DIDN'T sign the NDA. All I know is Kyle did not sign it.

I'm guessing Kyle has more than an educated guess, with his contacts in the industry. However, it's like Fight Club, you don't talk about Fight Club. He can't burn his sources, by outing the information.

I have a feeling Nvidia is sorting the issues with these cards because my third has been seemingly devoid of issues. I see a graphical anomoly here and there but, then again, I saw those on my 1080Ti... I think all the damn problems I had with the first two cards made me utterly paranoid and critical of anything that appeared to be a graphical flaw. Even to the point of getting me to think my toaster oven of a 1080Ti FE was possibly on it's way out.

I'm still waiting for this card to shit the bed, however it's not giving me any reasons, yet, to start doubting it. I did notice late game stuttering in Unity powered Battletech, however, that's just a flaw in their engine optimization. It also happened on every other system I've run it on. The stuttering I experienced in all my games on the other two cards was the cards themselves.

Before I forget, Merry Christmas everyone! You guys have provided me with a much needed outlet for my inner nerd .

 

[Jayham]

I've tested, stress-tested, overclockead and gamed with over 100 hours so far in slightly over 30 days that I've owned my card no issues whatsoever thankfully. EVGA RTX 2080 TI XC Black

 

[brentsg]

Second 2080 Ti and first 2070 XC Ultra still going strong.

 

[Slade]

40 days is when my EVGA 2080ti XC went. I was 100+ hours on gaming with that card. The Asus 2080ti card I have is nearing 60 days, but has only 50 hours of gaming on it.

 

[Legendary Gamer]

40 days is when my EVGA 2080ti XC went. I was 100+ hours on gaming with that card. The Asus 2080ti card I have is nearing 60 days, but has only 50 hours of gaming on it.

So, what do you think? If these cards pass the 100 hour mark, under load with no issues, are we in the clear?

It's somewhat annoying to me... After the first two cards, I still find myself critically scanning my screen for issues. While it somewhat detracts from game play, I suppose I will remain paranoid until I have placed enough strain on my card to believe it's not going to fail.

I live on my PCs, they are the gateway to my media and offline storage and gaming, research and writing books. It's why I dump the cash into the high end video card, I want the best possible visual experience. I don't need more than 60 FPS, however, I want stable 60 FPS! I was kind of underwhelmed by Intel's 9 series 14nm++++++++++++++++++++++(lol) architecture, lots of people had proclaimed you would hit 5.2Ghz on the 9600 easy and I never got it over 5 (I think that was pretty much the consensus in the article here as well). After that let down (all the while knowing the 2080Ti was shitting itself) I still bought into the 2080Ti because, if it works, I only need one card.

3rd EVGA Black, Cheapo Edition, 2080Ti still functioning properly...

 

[Slade]

I was paranoid about glitches too so when I saw some "line tearing" while watching videos, the oh shit not again thought popped into my head. Then I doubled checked my driver version and found I had installed 417.xx drivers which have been problematic for high refresh/low refresh multi monitor setups. Sure enough video glitches are part of the 417.22 set, while 417.35 has out right screen flickering. A reversion to 416.81 works for me, but this is also the version of driver where my first 2080ti failure happened. EVGA did suggest something to "some driver versions have been known to cause problems" line in their diagnostic e-mails, but I loathe to think that any particular driver version can KILL a card outright unless it was doing something with power/thermal targets outside of spec.

 

[brolioz]

I bought a EVGA 2080 Ti XC Ultra on Dec 7 2018 and it just died after 1 week of casual gaming. It first started with BSOD to eventually screen artifacting when booting into windows. The RMA process has been extremely slow and I won't be getting a new card for another 2 weeks by looks of it. Extremely upset as I have another $1300 charge on my credit card for collateral with me having to pay the interest and shipping. Will probably never buy anything high end from Nvidia again after this horrific experience.

 

[Slade]

Timing is rough for you due to holidays. I had my replacement within a week and I live in Canada across continent from EVGA. They received my RMA a week after that and promptly refunded me. As far as parts go, I've had mobos, PSU and video cards go over the past 15 years and I think all of the major companies I have dealt with (abit, Asus, EVGA, Corsair, Intel) have been really good with crossship RMA within 2-3 weeks avg for each of them.

 

[Domingo]

Probably unrelated to all of the issues people have been having, but there is a BIOS update for the launch EVGA 2080 cards:

https://forums.evga.com/EVGA-GeForce-RTX-2080-Ti-2080-XCXC-Ultra-BIOS-Update-m2858793.aspx

It's listed as being for the power limit setting, but I suppose there could be more there, too.

 

[FrgMstr]

Picked up Palit 2080 Ti dual from post office at Thursday 13th, gave it 50+ hours testing during week and now this Friday it started giving me Space Invaders.

You have pm.

 

[Slade]

Probably unrelated to all of the issues people have been having, but there is a BIOS update for the launch EVGA 2080 cards:

https://forums.evga.com/EVGA-GeForce-RTX-2080-Ti-2080-XCXC-Ultra-BIOS-Update-m2858793.aspx

It's listed as being for the power limit setting, but I suppose there could be more there, too.

I applied this BIOS update same day I got my first EVGA XC card. The replacement RMA card did not have it applied, so I did the update on that too the same day I installed it.

 

[Legendary Gamer]

I was paranoid about glitches too so when I saw some "line tearing" while watching videos, the oh shit not again thought popped into my head. Then I doubled checked my driver version and found I had installed 417.xx drivers which have been problematic for high refresh/low refresh multi monitor setups. Sure enough video glitches are part of the 417.22 set, while 417.35 has out right screen flickering. A reversion to 416.81 works for me, but this is also the version of driver where my first 2080ti failure happened. EVGA did suggest something to "some driver versions have been known to cause problems" line in their diagnostic e-mails, but I loathe to think that any particular driver version can KILL a card outright unless it was doing something with power/thermal targets outside of spec.

Well...Shit... Ran home for lunch, fired up Battletech to kill an hour and my frame rates took a nosedive for some reason... Then I got a single screen flicker and and a bit of artifacting. Was flawless last night... I have the latest drivers and there is no BIOS update for this card. The higher end models have one but I would be concerned I might brick the card applying them (even though they're essentially the same, this one just being lower binned). Have to come back and load it tonight. Dude... I am starting to get sick of this bullshit...

Update: I gamed on it all night with sensors on and it never went over 58c... Card performed flawlessly... I think this release will slowly drive me insane with it's issues.

 

[ripken204]

Well my ASUS card started having issues after 3 weeks.
Artificating and crashing in games.

ROG-STRIX-RTX2080TI-O11G-GAMING

Ended up RMAing back to Newegg and they are just going to give me a refund since they have none in stock.
I'm just going to wait a few months until this mess gets sorted out.

 

[Slade]

I personally think the design is flawed in some way. I think with cards dying that aren't reference design boards and cards dying with random temperatures attests to this hypothesis.

 

[Maddness]

I personally think the design is flawed in some way. I think with cards dying that aren't reference design boards and cards dying with random temperatures attests to this hypothesis.

Could very well be the case. Hopefully Kyle manages to get hold of a couple off the cards and can get to the bottom of this.

 

[Legendary Gamer]

Just submitted another Ticket with EVGA. Screen flickering is increasing in frequency, some mild artifacts here and there (not as bad as the first two but increasing in frequency and quantity). I have had my screen blank and go completely white with multi colored pixels (full screen). Screen flickering is noticeable in games, on the desktop (happens mostly at the bottom of the screen) and even more while watching video. Degraded performance in games is increasing. Initially it was hit or miss, now frame rates are taking a nosedive and performance in games is experiencing stutter and video rendering lag (3D model starts an animation and teleports to the end animation in the sequence and the game hangs noticeably while it's happening).

Looks like I will be moving onto card number 4 soon...

FYI: I've never gotten the space invaders issue. Just an entire host of other wierd stuff. EVGA seems incredibly aware of the other issues though. They keep issuing RMA's without me even asking for them, just based on my description of the issues.

 

[Slade]

Ever think something else is wrong with the system in general? I'd think at 3 cards down, something else may be killing the cards, but without a clear idea of failure rates I know its tough to point to anything but the 2080ti chip design.

417.xx has weird desktop video issues; but you stated 3d performance is acting up.

Maybe change up where the power is coming from if a modular PSU. I'd also monitor the rail voltages under load to see if something is dropping way too much out of spec.

 

[brentsg]

Ever think something else is wrong with the system in general? I'd think at 3 cards down, something else may be killing the cards, but without a clear idea of failure rates I know its tough to point to anything but the 2080ti chip design.

417.xx has weird desktop video issues; but you stated 3d performance is acting up.

Maybe change up where the power is coming from if a modular PSU. I'd also monitor the rail voltages under load to see if something is dropping way too much out of spec.

He doesn’t want to hear that. I tried to go there.

Even further evidence, he thinks his previous 1080Ti damaged some cards around it, and his 1080 Ti was also failing (but mysteriously didn’t RMA it).

 

[Legendary Gamer]

He doesn’t want to hear that. I tried to go there.

Even further evidence, he thinks his previous 1080Ti damaged some cards around it, and his 1080 Ti was also failing (but mysteriously didn’t RMA it).

Guys, I've swapped out everything else in my system. I've already commented on how the 2080Ti issues have been slowly driving me insane. My 1080Ti is not actually failing... After extensive testing the 1080Ti is fine. The issues I've encountered have made me hyper critical of damn near everything in my system.

The 2080Ti issues I'm facing are related to the 2080Ti. EVGA continues to support this fact each and every time I go through this process.

I also had (one bit the dust recently) 3 different systems to test on. Currently in the process of rebuilding the front room PC with a 8600, new MB, new RAM, new power supply, etc. If I see the same shit on that system as I have on my Ryzen and i7 4790 systems... It's the god damn 2080Ti.

FYI I have no other cards in my system. It's a clean build, I even posted pictures.

 

[Slade]

If you want to do a proper test, then you should run the replacement 2080 ti in another system altogether without running it in the one thag keeps dying. I have my asus 2080ti card running tensorcore experiments in a 2nd box albeit doing something other than gaming. Testing a glitching out and dying card in another box after its glitching and dying is kinda moot tbh.

 

[Legendary Gamer]

If you want to do a proper test, then you should run the replacement 2080 ti in another system altogether without running it in the one thag keeps dying. I have my asus 2080ti card running tensorcore experiments in a 2nd box albeit doing something other than gaming. Testing a glitching out and dying card in another box after its glitching and dying is kinda moot tbh.

Well, if it's manifesting the same issues on all three... of my systems. Then it's the video card. SO, by your logic I should run whatever replacement I get, exclusively, on the new system I'm building in my front room. Fine. Not sure how in the hell that's gonna matter if my lower binned video card is shit like the first three but, I am happy to oblige. It's essentially the same system, in my front room, as my den. 8600 vs 9600 cpu and only 16 GB of ram vs 32. Power Supply is supposed to show up today. We shall see.

This shit is utterly exhausting. None of the three cards I've had have the level of stability I would associate with a stable release of new tech. Dial things back to the 1080Ti era, the odds of receiving 3 bad cards in a row is insanely minor. I have to wait until the 7th for EVGA to respond anyway, they're on vacation according to their website. I have plenty of time to document the failures of the current card.

I am single, I have this inordinate amount of time to play games and such that allow me to study the output of the video card. I have never, in my life, seen this many issues associated with any video card I've ever used. For example, gamed on the current card flawlessly for almost 24 hours on Saturday. I was up til 6am. The issues manifested around 2am. Whatever the card is doing, it gives me a fucking headache, literally. Not because it's failing but because it's fucking with my head. The added strain on my brain from the glitches definitely is wearing my ass out.