My GTX690 has been my worst Computer experience.

Discussion in 'Videocards - NVIDIA GeForce' started by Denial, Jul 22, 2012.

  1. dune2

    dune2 Guest

    Messages:
    3,427
    Likes Received:
    19
    GPU:
    RTX 4090
    +1

    @OP: Good luck getting this sorted.
     
  2. Denial

    Denial Ancient Guru

    Messages:
    14,206
    Likes Received:
    4,118
    GPU:
    EVGA RTX 3080
    As someone already said I crosshipped the cards so they definitely two different cards. Unless I'm going crazy that is =d

    It's definitely a possibility it's the RAM. I already purchased 16GB of DDR3 2133 it should be here tomorrow.

    Yeah I honestly haven't tried screwing with the voltage. I have tried only running on one stick, and a cycled through a few of the sticks and I still had problems. The Maximus seems to be running the voltages for my memory at around 1.5 in auto if I recall correctly. When I get home from work I'll try screwing with the voltages to see what happens plus cycling the sticks again.

    I should be getting my new ram tomorrow. I figure I might as well upgrade everything now since I already did the first half.

    Well it's not like I can't afford to spend it. I was planning on waiting to Haswell to upgrade everything, but I just couldn't figure out what the hell was going on here and it annoyed me. So I did it now. And I'll probably be doing it again when Haswell drops =d

    You know that is actually one thing that's been a constant. My ASUS card in my PCI-E 16x slot. I didn't even think of that until now. ****, now I'm annoyed at myself.

    Thanks for trying the test setup. The weird part is though, that I can run Heaven for hours under my new motherboard no problems. I can even watch DOTA 2 via their in game HLTV system for a while without issue. Soon as I start interacting though, for longer period of times, that's when I start getting issues. At least that's what I've noticed.

    Like last night I played DOTA 2 for 2-3 hours. I fired up DayZ and I couldn't even get past the load screen before all my displays flickered and the game went invisible. It's really strange. And when I restarted after that, my computer defaulted to like the 800x600 resolution or whatever. I had to reinstall my graphics driver in order to get it detected again.

    I still think this is related to my system though, I don't think it's a problem between the Formula 5 and Kepler or anything.

    Yeah that's no problem. Thanks for doing that. I'm pretty sure it's a problem either with the card, my memory or now after your post my soundcard being in. Either way it's strange so hopefully i'll be able to provide some information for them and if its a configuration problem hopefully we can get it fixed

    I started the RMA process for a 3rd card from EVGA but again I'm getting it crosshipped. So I'll have this card for at least another 30-45 days before I send it back.

    When I get home I'll update with info about removing the soundcard and playing with the voltages on the memory. Tomorrow I'll be able to rule out the memory entirely.

    Anyway, thanks again guys, lots of help <3
     
  3. AlmondMan

    AlmondMan Maha Guru

    Messages:
    1,036
    Likes Received:
    345
    GPU:
    7900 XT Reference
    Terrible story man :( I only know the pain too well.

    If it's any help, I can say that my Asus brand 680 works beautifully.
     
  4. dune2

    dune2 Guest

    Messages:
    3,427
    Likes Received:
    19
    GPU:
    RTX 4090
    The seeming randomness in the timing when the errors occur gave me two other ideas you might want to test:

    1) Could this be a temp issue? Not necessarily at the card but maybe on the mobo/memory/PSU? How are your mobo temps when under load? How do you cool your rig? I know the CPU/vidcard is watercooled, but do you have airflow across the mobo? Even with watercooling the 690 will dump more heat into your case than the 480.

    2) PSU - I know you replaced your mobo, but this was effectively a like-for-like swap (1000w, single rail). The newer PSU seemed to help somewhat, but the issues came back. So maybe the new PSU is a little bit more effective than the old one but still not powerful enough to run everything you have in your rig? Not sure if you have done this already, but try stripping the rig to its bare minimum. Unplug all drives except the system drive, take out all other cards, drop any overclocks etc. See if the problem still persists.
     

  5. Neo Cyrus

    Neo Cyrus Ancient Guru

    Messages:
    10,780
    Likes Received:
    1,393
    GPU:
    黃仁勳 stole my 4090
    Sorry to hear you've had such a hard time and I'm happy to hear you're getting some help now.

    If it makes you feel any better I've got a boring story I've written before and I've probably forgotten and warped the details. :wanker:

    [boringstory] I've gotten multiple dead cards (or cards which died extremely rapidly) in a row from MSI. It wasn't just 2, it was literally something like 3 or 4 (yes it was so long ago I forgot). The 4th or 5th finally worked and remained working for years.

    Each time I had to wait 2-4 weeks (actually it may have not been faster than 3 weeks for any) for MSI to ship me a "new" one (lol they weren't new, they were obviously used and beaten up). So I was essentially stuck without a video card for months. I took the cards into their RMA facility in person, the place is a joke. As you can guess to this day I haven't bought another MSI product and I don't know if I ever will again.

    Anyway, 3-4 duds beats 2, what do I win?

    The only other time I had such bad luck was with Asus motherboards, exploding capacitors left and right. Though they didn't send me any duds as replacements and IIRC it took them 2 weeks to send each replacement, not 4. I haven't bought a single Asus product since then either. For the record I do consider Asus products now, with solid state capacitors I doubt there will be any more exploding.[/boringstory]

    Um... I'm sorry I don't have any real advice for you as most of it has already been covered. Good luck.
     
    Last edited: Jul 24, 2012
  6. Denial

    Denial Ancient Guru

    Messages:
    14,206
    Likes Received:
    4,118
    GPU:
    EVGA RTX 3080
    I'm always paranoid about my watercooling setup. At the moment I don't think I have enough rads to effectively cool my loop as good as it could be. I have a Corsair 800D with 3 top fans blowing out across a XSPC RX360. I also have a rear fan blowing out across RX120. Intake is through the bottom, there is a single 140mm fan that is raised slightly above the mesh to pull air in. I've been contemplating cutting out two additional fan holes for side intake on the panel or possibly even cutting out the bottom to add an additional RX240.

    Either way my CPU temps average about 50/60ish loaded at stock and my GPU temps are about 45-50ish loaded in a game, depending upon the ambient room temp which changes a lot I guess. This is obviously warmer than I wanted, considering I've been reading people getting 30c with 690's loaded, but most of those people also run dedicated loops with much more radiators. As for my motherboard and other components. I've noticed that my memory under my Maximus 5 is much cooler than my X58, X58 I manually put the memory at 1.6 in order for it to achieve the correct timings, I think the Maximus defaults to 1.5 auto, so that should explain that. Nothing else really seems to be hot to the touch, I had the ASUS PC Probe thingy running for a while when I first got the board, nothing looked out of the ordinary but I can check again tonight.

    As for the PSU and power, I currently have the following plugged in:

    3770K
    Asus Maximus 5 Formula
    2GB Memory (I've been testing with only one stick past couple days)
    GTX690
    450 Pump for the WC loop
    EssenceSTX Soundcard
    2x Vertex 4 256's

    My RAID 5 drives aren't even plugged in because I can't run the array on the Formula 5. So I pulled them out and will eventually get a external exclosure or a RAID card.

    Either way I don't really feel like there is an overpowering amount of stuff there for the PSU. The Essence eats a Molex, same with the pump. I don't even have my DVD drive plugged in because I simply never got around to doing it.

    The more and more I'm thinking about this the more and more I'm thinking it's the EssenceSTX for some reason.
     
  7. Wagnard

    Wagnard Ancient Guru

    Messages:
    2,746
    Likes Received:
    519
    GPU:
    MSI Geforce GTX 1080
    I would suggest to run the system as simple as possible. That implies even removing the water cooling and installing stock cpu cooler fans.
    Remove the special sound card. remove overclock.
    Run this system 100% stock.

    I have a GTX 690 too with a Maximus Gene V motherboard with 16gb corsair memory.
    And the card is really good. I hope you can benefit of it too.

    Also can we have a pic of the inside of your computer ? a clear one if possible just in case we see something. who knows...
     
    Last edited: Jul 24, 2012
  8. Bentez

    Bentez Guest

    Messages:
    1,309
    Likes Received:
    1
    GPU:
    Palit GR RTX 3080
    Standard volts for older triple channel X58 memory was 1.65v. If your Ivy Bridge setup is defaulting to 1.5v maybe bump it up to 1.65v.
     
  9. Denial

    Denial Ancient Guru

    Messages:
    14,206
    Likes Received:
    4,118
    GPU:
    EVGA RTX 3080
    http://imgur.com/a/3hEkf

    This is my old setup, it's the same card that was giving me problems, just under the X58 setup. It was right after I installed the WC block on it.

    When I get home tonight ill take some of my newer ASUS build, 4 more hours or so.


    Yeah this will be something I try tonight. My X58 I had it running at 1.6 I think or maybe it was .65, it was whatever Corsair recommended on the back of the box for it, I don't really remember atm.
     
    Last edited: Jul 24, 2012
  10. Denial

    Denial Ancient Guru

    Messages:
    14,206
    Likes Received:
    4,118
    GPU:
    EVGA RTX 3080
    So I'm home now and I tried a few things. First thing I did was remove the sound card and all but 2 sticks of ram. I then set my memory to 1.65v. I started DayZ launcher, the launcher is down or whatever so I manually entered an IP. Game launched and stuff, looked ok, server was fill. I ctrl alt deleted and immediately got a BSOD for memory management. Rebooted and set the voltage lower to 1.6. I haven't seen a BSOD yet but as soon as I tried DayZ i got the invisible game flicker thing again.

    Here is a pic of the newer stuff - I didn't bother routing some of the cables because I just wanted to see if it fixed the problem at the time.

    http://i.imgur.com/a6DCyh.jpg

    At this point I guess the only thing I can do is wait till tomorrow and see if the new memory fixes it, dunno.

    Nvidia hasn't contacted me yet btw
     

  11. Agent-A01

    Agent-A01 Ancient Guru

    Messages:
    11,630
    Likes Received:
    1,119
    GPU:
    4090 FE H20
    Im assuming you are using asus drivers and not the unified drivers? i would remove them, they are wonky and download the unified 1.61 drivers
    http://brainbit.wordpress.com/2010/07/19/asus-xonar-unified-drivers/
    install c-media control panel. or low dpc lantency one.

    When i was tightening my timings for my ram, i tried memtest+, it ran for 4 hours fine, with no problems. did like 5 or 6 passes iirc and had zero errors. funny enough load up a game and it BSODs instantly. I dont trust memtest+ as a valid stability test anymore. It bsoded in prime 95 in a minute.
     
  12. Denial

    Denial Ancient Guru

    Messages:
    14,206
    Likes Received:
    4,118
    GPU:
    EVGA RTX 3080
    Think you might be right about memtest.

    I just raised all the timings on my memory by a bunch, like up to 15 and running at 1.65 voltages. I haven't experienced an issue in the last hour since doing it.
     
  13. -Tj-

    -Tj- Ancient Guru

    Messages:
    18,097
    Likes Received:
    2,603
    GPU:
    3080TI iChill Black
    I saw the same thing with memtest, windows7 memory diagnostic tool was more accurate (try 8-10 rounds on default settings)

    Maybe try another thing, go to advanced ram timings and find tRTP - read to precharge delay, set it to 8 (usually its at 5-6 by default) and leave main timings at default.

    ^
    I saw this info by some Z68 mobo video review, it can be handy by higher ram OC or for stability.





    And a bit stupid question :D, did you enable pcie3.0 in mobo bios, gpu in 1st pciex16 slot?
     
  14. Denial

    Denial Ancient Guru

    Messages:
    14,206
    Likes Received:
    4,118
    GPU:
    EVGA RTX 3080
    The GPU is in the first slot and PCI-E 3.0 is on by default. I've tried fooling around with it off but it didn't really seem to help anything.

    That being said, as of right now I probably played 4 hours of DOTA tonight without a single issue since bumping up my memory timings. I'm only running on 2 gigs of ram atm though.

    This could be a coincidence though, because in the past I've tried stuff that I thought fixed it, and the next day it would go to hell again.

    And yeah I'll try screwing around with the timings a bit more. Now that it seems to be working for a bit I wanted to get some dota in.
     
  15. sykozis

    sykozis Ancient Guru

    Messages:
    22,492
    Likes Received:
    1,537
    GPU:
    Asus RX6700XT
    Memtest86+ is utter garbage....always has been. I've never actually seen it report errors, even from sticks that were known faulty.

    Windows 7 memory diagnostic tool is based on Memtest86 (developed by Chris Brady of BradyTech). Memtest86+ is developed by CanardPC using older source code from the original Memtest86.
     

  16. Agent-A01

    Agent-A01 Ancient Guru

    Messages:
    11,630
    Likes Received:
    1,119
    GPU:
    4090 FE H20
    Well i tried both of them, neither showed errors. Ill stick to prime/real wod tests. The win 7 mem test tool found nothing as well.
     
  17. BladeRunner

    BladeRunner Guest

    Messages:
    1,938
    Likes Received:
    1
    GPU:
    Asus 1080Ti STRIX
    I feel for the OP. My two new 670s werent a smooth sailing either. One of them was DOA. After two mins of play it would start artifacting and erroring out in standalone or SLI modes. The other card did not exhibit this when in standalone mode. I now have to RMA, and I hope to god its OK afterwards, but already, after leaving the RED team, its not been a great welcome so far.
     
  18. Hilbert Hagedoorn

    Hilbert Hagedoorn Don Vito Corleone Staff Member

    Messages:
    48,392
    Likes Received:
    18,564
    GPU:
    AMD | NVIDIA
    So how did it work out ?
     
  19. Wagnard

    Wagnard Ancient Guru

    Messages:
    2,746
    Likes Received:
    519
    GPU:
    MSI Geforce GTX 1080
    If you still have the problem, I suspect its the water coolong used on the gtx 690. I know that some part have to be cooled (voltage regulator for example) that some exotic cooler doesnt cool.

    I recommend to replace all original heat sink / fan on you evga signature and use it stock just to be sure. The could be a mem chip not cooled correctly. Anyway you ave nothing to loose but im almost sure its that!
     
  20. snowdweller

    snowdweller Guest

    Messages:
    492
    Likes Received:
    0
    GPU:
    Leadtek GTX 580 SLI
    Wow I just wanted to say I feel for you dude if it is the timings of RAM after buying and going through all you did. Man :bang:
     

Share This Page