Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Nvidia RTX 2080 Ti graphics cards are dying on a lot of users.

PsYcHoGBRPsYcHoGBR Member UncommonPosts: 469
«1

Comments

  • GorweGorwe Member EpicPosts: 6,098
    That the cards are dying is nothing new(even sub 1 year), the real question is in which %.
  • OzmodanOzmodan Member EpicPosts: 9,505
    Rumor is they have a major heat problem.
  • PsYcHoGBRPsYcHoGBR Member UncommonPosts: 469
    edited October 2018
    Just been on to Nvidia forums and there's a lot of posts about faulty cards.

    https://forums.geforce.com/default/board/227/geforce-rtx-20-series/
  • RidelynnRidelynn Member EpicPosts: 6,826
    No real indication of the breadth of the problem. Could just be a few very vocal users, or it could be a systemic problem.

    I do hear that nVidia is at least offering advance replacement on RMA (they ship you a replacement unit before they receive the malfunctioning hardware). I would think, for the price of the cards, they would include concierge home delivery and installation, but there I'm getting a bit biased.
  • DarkestOverlordDarkestOverlord Member UncommonPosts: 545
    This is why you never buy the first batch of anything ;)
    OzmodanTorvalGaeluianMrMelGibsonsausagemix




  • QuizzicalQuizzical Member LegendaryPosts: 20,996
    The two critical questions aren't knowable:

    1)  What fraction of the cards are dying?
    2)  How were the dead cards treated before they died?

    Some fraction of every SKU produced in large quantities will fail.  If it's in line with other cards, then it's just the price of buying a consumer video card.  It's only if it's abnormally high that there's a real story there.

    On (2), one important factor is whether the card was overclocked.  In the case of the founders' edition cards, the answer is that they have a considerable factory overclock, and that can't be good for failure rates.  But there's still a question of whether they were further overclocked beyond that.  There's also the question of whether the power supply and case airflow were built to be appropriate for a high end card.  But I doubt that would be substantially different from what is typically done with other high end cards.
    Phry
  • RenoakuRenoaku Member EpicPosts: 3,068
    This is why you never buy the first batch of anything ;)
    This is why you refund within 30 days if there are problems.

    Run my 1080ti 6 months straight no power off no fail.
    PsYcHoGBR
  • Asm0deusAsm0deus Member EpicPosts: 2,946
    The article seems to be implying that a far far greater number or % of cards are dying than what is usual in this short amount of time which really could be indicative of an issue in design or quality control.

    Brenics ~ Just to point out I do believe Chris Roberts is going down as the man who cheated backers and took down crowdfunding for gaming.

    case: Coolermaster HAF932
    PSU: Antec EA 750watt
    RAM: 4x2g G-SKILL DDR3-1600mhz 9-9-9-24
    Mb:Gigabyte GA-P55-UD4P
    CPU: i5-750 @4ghz
    GPU: gtx msi N760 TF 2GD5/OC
    cooling: Noctua NH-D14
    storage: seagate 600 240GB SSD, samsung evo 860 500gb SSD, 500GB x7200rpm HDD


  • RidelynnRidelynn Member EpicPosts: 6,826
    Asm0deus said:
    The article seems to be implying that a far far greater number or % of cards are dying than what is usual in this short amount of time which really could be indicative of an issue in design or quality control.
    Yeah, but a lot of articles claim or imply a lot of things just to get you to click.
    Quizzicalgervaise1laseritPhry
  • Asm0deusAsm0deus Member EpicPosts: 2,946
    edited October 2018
    Ridelynn said:
    Asm0deus said:
    The article seems to be implying that a far far greater number or % of cards are dying than what is usual in this short amount of time which really could be indicative of an issue in design or quality control.
    Yeah, but a lot of articles claim or imply a lot of things just to get you to click.
    Clickbait is usually in the title itself not so much in the article.  Also they seem to be saying that lots of complaints are from people that didn't OC their cards. 

    Dunno might be worth checking out the nvidia forums.

    Just had a quick look @ https://forums.geforce.com/default/board/227/geforce-rtx-20-series/1/  and there seems to be quite a few complaints about 2080ti issues and cards simply dying.



    Brenics ~ Just to point out I do believe Chris Roberts is going down as the man who cheated backers and took down crowdfunding for gaming.

    case: Coolermaster HAF932
    PSU: Antec EA 750watt
    RAM: 4x2g G-SKILL DDR3-1600mhz 9-9-9-24
    Mb:Gigabyte GA-P55-UD4P
    CPU: i5-750 @4ghz
    GPU: gtx msi N760 TF 2GD5/OC
    cooling: Noctua NH-D14
    storage: seagate 600 240GB SSD, samsung evo 860 500gb SSD, 500GB x7200rpm HDD


  • OzmodanOzmodan Member EpicPosts: 9,505
    edited October 2018
    Still think it resides in the fact that they did not cool the card properly.  Most of the complaints I have seen deal with heat issues.

    I just hope the OEMs are paying attention to this.
  • krulerkruler Member UncommonPosts: 589
    Spent a fortune on mine and still got massive wife aggroe, but I went for the ASUS card plus water cooling hence the fortune spent and its purring nicely at load well below average stats ive seen for the card heat wise...…. 


    The returns policy is good so hopefully no problems, and at the price of the card I expected 6 months of back rubs and happy endings but we cant have everything.
    Phry

  • CleffyCleffy Member RarePosts: 6,052
    Defect rate is ridiculously low for most hardware. Something like 1 in 1000 or lower. Considering how many of these cards have sold. Anything more than a handful having a defective card is a problem compared to the norm.
    Asm0deus
  • 13lake13lake Member UncommonPosts: 708
    edited November 2018
    Similar to Vega, 2080 Ti have some monster power and voltage spikes which don't play nice with multi-rail PSUs (Lot's of vega 64 users especially OC-ed could not run the cards on multi-rail PSUs of any wattage), and either trip OCP protection and/or don't play nice with voltage regulation on some PSUs. 

    On reddit a few people mentioned that the way that some PSUs react to those spikes damages the GDDR6 memory or its controllers, which is why some people keep having RMAd and different cards die (same PSU that causes no issue on any other card).

    Another redditor swapped PSUs with a friend and had no issues with a problematic card.
    https://www.reddit.com/r/nvidia/comments/9su4da/nvidia_2080_ti_failures_vs_power_supply/

    He thinks it's voltage regulation on the 12v rail, but bullzoid from actually hardcore overclocking thinks it's too agressive OCP protection on some PSUs, hopefully he does indepth testing on this on his youtube channel.
  • PsYcHoGBRPsYcHoGBR Member UncommonPosts: 469
    edited November 2018
    Looks like it might be DDR6 memory overheating. Got this from Tom's Hardware. From the picture it looks like they got the label wrong from the German Tom's Hardware.

    https://www.tomshardware.com/news/rtx-2080-ti-gpu-defects-launch,37995.html



    OzmodanAsm0deusQuizzicalPhry
  • OzmodanOzmodan Member EpicPosts: 9,505
    PsYcHoGBR said:
    Looks like it might be DDR6 memory overheating. Got this from Tom's Hardware. From the picture it looks like they got the label wrong from the German Tom's Hardware.

    https://www.tomshardware.com/news/rtx-2080-ti-gpu-defects-launch,37995.html

    From what I heard, it has been happening when people are running high resolutions in some games.
  • Asm0deusAsm0deus Member EpicPosts: 2,946
    Been reading the https://forums.geforce.com/default/board/227/  forums and BOY are the nvidia fanboys working hard to deny there's any problems.

    It's pretty clear by now there's an issue and they missed it, most likely they did very poor quality control in their rush to make the big bucks and keep up with demand.

    Ozmodan

    Brenics ~ Just to point out I do believe Chris Roberts is going down as the man who cheated backers and took down crowdfunding for gaming.

    case: Coolermaster HAF932
    PSU: Antec EA 750watt
    RAM: 4x2g G-SKILL DDR3-1600mhz 9-9-9-24
    Mb:Gigabyte GA-P55-UD4P
    CPU: i5-750 @4ghz
    GPU: gtx msi N760 TF 2GD5/OC
    cooling: Noctua NH-D14
    storage: seagate 600 240GB SSD, samsung evo 860 500gb SSD, 500GB x7200rpm HDD


  • QuizzicalQuizzical Member LegendaryPosts: 20,996
    That looks like it's a milder case of what ailed the GeForce GTX 590, and to a lesser extent, the Radeon HD 5970.  The GPU puts out most of the heat, but other things can overheat, too.  While the GPU itself has to be the main focus of the cooling, other things can cause problems if you completely ignore them.

    The GTX 590 had the excuse that it was two GPUs that could crank out perhaps 450 W combined, as the nominal 365 W TDP was pretty much a lie.  The Radeon HD 5970 was also a dual GPU card.  The RTX 2080 Ti doesn't have any such excuse.  Is a $1200 price tag not enough to pay for a decent cooler or something?
    OzmodanPhry
  • QuizzicalQuizzical Member LegendaryPosts: 20,996
    Small sample size, but when a prominent tech media person says that his own RTX 2080 Ti died after there were various rumors that they were unreliable, I think we may be on to something:

    https://www.hardocp.com/article/2018/11/09/geforce_rtx_2080_ti_fails_after_gaming_for_2_hours/

    What I don't see is direct reason to believe that the RTX 2080 and RTX 2070 are also unreliable.  Maybe they are, or maybe not.  A thermal shot like the one Psycho linked above would be useful.
    Ridelynn
  • OzmodanOzmodan Member EpicPosts: 9,505
    It seems that Nvidia is making excuses for the faulty cards none of which make any sense.

    https://www.hardocp.com/article/2018/11/14/nvidia_on_cause_rtx_2080_series_card_failures/
    Ridelynn
  • RidelynnRidelynn Member EpicPosts: 6,826
    I read that - yeah it really doesn’t say anything.
  • QuizzicalQuizzical Member LegendaryPosts: 20,996
    It's interesting that the GeForce RTX 2080 Ti seems to have nearly vanished from stores.  It's not normal for that to happen to a part two months after launch, especially for such a low volume part that should be easy to keep in stock.  Perhaps Nvidia realized that they were defective and decided to stop shipping them until they can fix it.

    The RTX 2080 (non-Ti) and RTX 2070 don't seem to be affected.  They're available in ample stocks and aren't plagued by rumors of widespread failures.  The GTX 1080 Ti seems to be discontinued, as the ones that were available at reasonable prices are gone, and all that remains is some stragglers at stupid prices.  Presumably Nvidia wants people to move to the RTX 2080 for that performance level.  The GTX 1080 still seems to be around, though.
  • SedrynTyrosSedrynTyros Member EpicPosts: 2,924
    edited November 2018
    Well that sucks ... guess that's another reason to wait for the 2070.
  • QuizzicalQuizzical Member LegendaryPosts: 20,996
    Ozmodan said:
    It seems that Nvidia is making excuses for the faulty cards none of which make any sense.

    https://www.hardocp.com/article/2018/11/14/nvidia_on_cause_rtx_2080_series_card_failures/
    Let me translate that for you:

    "We were in such a rush to get cards to market that we didn't test them very well, so we didn't filter out enough of the defective ones before selling them."
    Ozmodan
  • OzmodanOzmodan Member EpicPosts: 9,505
    edited November 2018
    Well that sucks ... guess that's another reason to wait for the 2070.
    Why in the world would you waste money on a 2070?  You can get a 1080 much cheaper and the performance is about the same.  Or for a few bucks more get a 1080 ti which is much faster.

    Just to note, the 2070 does not have the power to do ray tracing in any effective way.


Sign In or Register to comment.