Is my hardware faulty or am i delusional here?

Lmememan212 on 26/08/2019 - 17:42

Maybe i might be able to get some help in these forums.

Round about december last year i bought a PC off of a deal in the compiuting section of this site. It was a fairly good deal at the time, directly from a popular group that frequently posts their new deals on this site and whatnot. Reviews are perfect n all so it aint a case of legitemacy here.

Specs are:

  • OS: Windows 10 pro 64-bit
  • CPU: AMD Ryzen Threadripper 1920x
  • RAM: 16gb DDR4
  • Motherboard: ROG STRIX X399-E GAMING
  • GPU: GALAX NVIDIA RTX 2080
  • SSD: Allied SSD 240Gb
  • HDD: Western Digital WDC 1tb

Seems great on paper but i received faults with it on day 1. More specifically no matter game i play i receive the error 'DXGI-DEVICE-ERROR-REMOVED' going on about the GPU being physically removed from the system while playing. This can happen on high demanding games or low demanding games with the same error. I mostly play APEX LEGENDs or at least try to with my mates, these crashes are stupid sporadic, no indication that they will occur. However i cant got at least 1-2 hrs without crashing once. These crashes 90% give this error but on rare occasions the whole PC will freeze forcing me to restart. Spoke with tech support for APEX on their forums back and fourth for weeks, only concusion they came to was to send it back the them.

Should also mention this PC has a tendecny to freeze up while watching a livestream or video on google chrome, these are rare but do occur and brick the whole PC.

I seriously have tried everything under the sun to fix it, monitoring temps, reg edits, clean uninstall of GPU drivers, updating drivers from windows to motherboard, lowering settings for the game, different PSU, different GPU. I told the provider of these problems and sent the computer back. Around a month they said they could not find any errors and sent it back. Sure enough the errors were still there so i sent it back again. This time they had it for around 2 motnhs or maybe 3, they could not find any errors so they sent it back. Once i received it i made sure to not touch or alter anything on the computer and to just play. On the 7th round of palying APEX legends the entire PC froze once again. This brings it to now where im getting these errors at least once every two games, its unbearable.

Its been this was since i received it, and from their words i can't get a refund as they cant find a problem with it.

Am i really dumb or something, there's gotta be a thing im missing here. Dont know where else to turn here, this thing was a lot of f***ing money btw.

Any help in any way would be much appreciated. Cheers.

Computing

Comments

  • scrimshaw on 26/08/2019 - 17:54
    +15

    Welcome to the Ryzen RAM troubleshooting club, on your left, is a huge group of people with RAM related issues and on your right, a group of people with motherboard BIOS problems. Please take a ticket from a counter and stay seated, our helpful tech support assistants will be with you soon.

    On a serious note: have you:

    • Updated the motherboard's BIOS to v1002
    • Checked the QVL list on your motherboard to see if the RAM is supported, or spoke to someone from ASUS on their forums
    • performed any sort of memory testing using memtest?
    • Tried increasing the DRAM voltage beyond the reference 1.2v spec?
    • Used the 1usmus DRAM calculator to find safe timings for your RAM?

    (If you are confused about what the final point was, you should watch Hardware Unboxed video on how to fine tune memory for Ryzen. See)

    • mememan212 on 26/08/2019 - 20:37

      the first 3 of your points yes I have tried.

      Will have to look into the last two. Cheers

      • netsurfer on 28/08/2019 - 10:59

        Try this:

        How To Fix DXGI Error Device Removed Error - Solve …

        Hopefully, the registry fix would do the job… Though I suspect your already tried. I would also try to downclock the GPU and VRAM on the graphics card.

        Otherwise, RAM is often an issue.

        Start with the most conservative RAM settings. If your motherboard BIOS supports command rate of 2T, use that. If your motherboard BIOS separates gear down and command rate, also set gear down mode to on/enabled. If you have 4 DIMMs or you mix two different sets of RAM modules, 2T is inevitable.

        If you MUST use the DRAM calculator, make sure you know how to export your current RAM settings properly (using Thaiphoon Burner, you need to change the export settings). Don't use overclock or aggressive settings in DRAM calculator. While the Hardware Unboxed video is good and useful, it is more suited for people with quality RAM. If you have dodgy RAM modules, don't overclock.

        Run Memtest, it's a must for AMD. I had two identical pair of RAM modules, one set works fine, one set is faulty BUT that set only failed memory tests like crazy on AMD setup. When putting the RAM modules on an intel system, memtest passes.

        Start with the slowest settings, run memtest and gradually make your way up.

  • Bid Sniper on 26/08/2019 - 17:56
    +1

    Not your job to diagnose, if the system is faulty then return it.

    If it was me, id do the nuclear option and reformat. Mobo bios up to date?

    • mememan212 on 26/08/2019 - 20:38
      +2

      I did drop the nuke previously. I have attempted to return it but out of the two times the provider hasn't found a problem with it.

  • HighAndDry on 26/08/2019 - 18:03
    +1

    Could be any number of issues, but I'm mostly worried about:

    I seriously have tried everything under the sun to fix it, monitoring temps, reg edits, clean uninstall of GPU drivers, updating drivers from windows to motherboard, lowering settings for the game, different PSU, different GPU.

    Uh - are you saying you've physically swapped out a different PSU or GPU?

    But honestly it could just be a problem with the PCI-E slot on the motherboard that's making your system think the graphics card is being randomly physically disconnected.

    Edit Reddit user says this: https://www.reddit.com/r/BattlefieldV/comments/a080d2/dxgi_e…

    I found the solution myself:

    IF YOU HAVE DXGI DEVICE HUNG \ REMOVED DIRECTX11 CRASH TO DESKTOP THIS IS THE SOLUTION:

    DO NOT BELIEVE WHAT EA EMPLOYERS WILL TELL YOU, IT IS NOT YOUR HARDWARE. Here it is 24 pages about this problem:

    As you can see BAD PROGRAMMED SOFTWARE DOES THAT.

    Several leaks are present both in Tomb Raider, Fifa 2019 and Battlefield V.

    No idea how correct that is - do you play any of these games?

    FULL SOLUTION:

    Go To Windows Search 2) Search Regedit 3) Right click and select run as administrator 4) Open HKEY_LOCAL_MACHINE 5) Open SYSTEM 6) Open CurrentControlSet 7) Open Control 8) Click GraphicsDrivers 9) Right Click And Select New In Windows Registry 10) Select DWORD (32-bit) Value 11) Type TdrLevel (Write it exactly T and L should be capital) 12) Double click on TdrLevel 13) Make sure value data is 0 14) Now restart your pc Done! Problem solved, dxgi error device removed FIXED Alternatively you can just delay the detection (10-15 sec. is enough, 2 sec. (default) is not enough for these bad software). The following link explains how. This is the explaination of what it does:

    Put the system BACK INTO ITS ORIGINAL CONFIGURATION, and contact the seller about sending it back for repairs with a description of the issue.

    • Baysew on 26/08/2019 - 18:11

      I told the provider of these problems and sent the computer back. Around a month they said they could not find any errors and sent it back. Sure enough the errors were still there so i sent it back again.

      and from their words i can't get a refund as they cant find a problem with it.

      • HighAndDry on 26/08/2019 - 18:16
        +4

        Honestly there's nothing else for it - take a video of it dying (maybe putting a dashcam over his shoulder to record in loop?), and send it back again.

        Product's faulty from what I can see - seller should deal with it, OP's problem is just that they need to show that it is faulty and how it's faulty.

        Maybe ask them to swap out a motherboard - again, if it's a physical issue, I'm putting my money on a PCI-E slot problem.

        • mememan212 on 26/08/2019 - 20:40

          I have given a description on how to replicate the issue. Alongside screenshots of the same error with different games. They haven't been able to replicate (somehow ffs).

          • skid on 26/08/2019 - 21:34
            +1

            @mememan212: I think it is a PCI-E slot problem too. Your motherboard has 4 slots for the GPU. Have you tried moving the GPU to another slot?

            Have you tried swapping the power cables from the PSU to the GPU?

            Lastly, I'd kick up a massive stink until TechFast replaces the whole PC. Isn't that option the reason why people buy whole systems from one shop?

            • mememan212 on 26/08/2019 - 21:49

              @skid: Yes i have tried swapping the power cables. Havent tried moving it to a different slot but its worth a try. As i said before they arent seeming to budge unless they can repeat the problem themself. I have no idea what they're testing to do that that.

    • mememan212 on 26/08/2019 - 20:40

      Tried those exact solutions from those exact links haha. to no avail though.

  • pjetson on 26/08/2019 - 18:07

    You need to find some way that the seller can see the problem. Have you tried running freeware burnin software? If you can find something that reliably shows up the problem, you'll have a much better chance of a resolution.

  • theHMASfriendship on 26/08/2019 - 18:11
    +2

    Nothing is real everything is an illusion

    • HighAndDry on 26/08/2019 - 18:21
      +3

      There is no PC…

    • [Deactivated] on 26/08/2019 - 20:45

      Agree: reality is merely an illusion, albeit a very persistent one.

  • garetz on 26/08/2019 - 18:20

    DXGI_ERROR_DEVICE_REMOVED:

    The graphics device has been physically removed, turned off, or a driver upgrade has occurred. This happens occasionally and is normal; your app or game should recreate device resources as described in this topic.

    Maybe you had a driver update ?

  • mememan212 on 26/08/2019 - 23:41

    Update:

    Playing with my mates, after a few games the whole PC just froze while playing. This used to be a rare occurance but now i get it every night.

  • mememan212 on 26/08/2019 - 23:55
    +3

    Some pics for anyone interested:

    These are only a handful of the many delights i have experienced with this amazing PC! Nothing wrong with it by the way! RMA team didnt find any problems! Totally didnt have my PC freeze up yet again 20 mins ago ruinging my game with my mates!

    Another bonus was that the copy of bfv bundled with this game didnt work, didnt really care tbh. Also they kept my power cable from when i sent it there last, now im using this old one from a decade ago, great stuff all round

    • Zachary on 27/08/2019 - 10:31

      The last one looks like your card might be borked….or your screen is borked….or both….

      • smashman42 on 27/08/2019 - 13:14
        +1

        They've already replaced the GPU and get the same sort of BS - I reckon the slot itself is borked, meaning need new motherboard.

        OP - You've got PCI lanes coming out your arse with a threadripper setup, so try another full 3.0x16 slot on the off chance just the one slot is borked - check the mobo manual for which slots to try, some might be limited to x4. Use both GPUs you've been testing with to be sure.

        It ain't a real fix but it might be a work around that stops you setting it on fire.

        • Zachary on 27/08/2019 - 13:52

          Oh yeah forgot about the motherboard being a possible dud too…..woops my bad…

        • mememan212 on 27/08/2019 - 15:07

          Even if that was a fix, wouldn't that mean the the MOBO i got is broken?

          • smashman42 on 27/08/2019 - 18:24

            @mememan212: Ya, if it only plays up in one slot it proves that the motherboard is borked.

            Whether the seller replaces it like they should or keeps screwing you around I have no idea, but worst case if it works in a different slot at least you'd be able to use it even if they refuse.

          • Seedy seed on 27/08/2019 - 22:55

            @mememan212: Could be a bent cpu pin also 😂 i got a asus mobo with a bent pin once nd couldnt get it to run half my ram.

  • veej on 27/08/2019 - 08:19

    I've finanlly got my new Ryzen build up and running. I ran into some interesting issues but have now been pretty stable.

    One, I used a case with a riser and I found that it (the riser) was a couple of generations old and I had to manually change the PCI-e slot to Gen 2.
    Benchmarking was crashing around 10FPS and fixing it pushed it past 50FPS.

    Two, the motherboard didn't detect what RAM I was using and picked something way slower than I had in it.

    I would look at the motherboard settings for the graphics slot.

  • jonkvh on 27/08/2019 - 11:19

    What's the power supply? Also, try plugging it into another power outlet.

    • mememan212 on 27/08/2019 - 15:08
      +1

      PSU is 750W Thermaltake, i've tried also with a 1050 PSU and multiple power outlets to no difference

  • Diji on 27/08/2019 - 11:55

    This must be the "Techfast flava" I've read about.

    • mememan212 on 27/08/2019 - 15:08

      Im not sure what you're refferring to. care to share?

      • Diji on 27/08/2019 - 15:16
        +1

        Just something I read posted by their marketing rep, this topic brought that to mind for whatever reason.

  • AncientWisdom on 27/08/2019 - 15:39

    I feel for you OP, that's not a good experience. I would just put my foot down for either a fix or a full refund from seller.

  • jace88 on 27/08/2019 - 20:01
    +1

    Have you tried running benchmarks eg 3dmark, heaven, etc… to see if they consistently produce the error? That might help make it repeatable.

  • Ghosteye on 27/08/2019 - 21:11
    +1

    I had similar problems awhile back on a ryzen build. Did a fresh install of windows (hard reset). I even downloaded the most 'stable' windoes which was build 1703 I think. YMMV

    Do not let the pc connect to the internet to auto-download drivers.

    Downloaded all the drivers from manufacturer's website onto a usb stick from another pc and manually installed them. (mainly just the motherboard and gpu) just make sure you download the right ones from the manufacturer's website.

    This should fix your problem.

    edit1: You can also try to install the older drivers instead of the latest one. I think the older ones tend to be more stable and bug free. Maybe one or 2 builds before the latest one. Usually when you head over to overclock.net you can find the model and make of your mobo and people who own it. There will usually be complaints about the current update (if any) in those forums. Maybe you can trace some of your problems there and see if anyone else who owns your exact board has faced this issue before.

    • mememan212 on 27/08/2019 - 22:12

      thanks for the suggestion but i have tried that lol. also on the odler driver part yes i have tried multiple. mobo is the same

      • Ghosteye on 27/08/2019 - 22:17

        did you OC the ram at all? maybe try 1 stick at a time in different slots? The games crashing can probably narrowed down to either your gpu or ram I think.

        You could always try to seat in on the other pcie slots and see how that goes.

        • mememan212 on 27/08/2019 - 22:20
          +1

          Nah didnt oc anything - ill have to put this stuff down on the list to try as from last contact they've said they did "everything in their power", which consisted of running 3d benchamrk and playing apex for like 1-2 hrs over the course of 2-3 months.

          But ye i'll have to put that down, writing on this thread while my mates finish the game that i crashed from lmao.

          Cheers

          • Ghosteye on 27/08/2019 - 22:26

            @mememan212: yeah try those. Normally if its just games it means some part of the system can't handle being stressed out.

            You should try to look out for 'oddities' in day to day tasks too. eg mouse skipping/ disappearing, mouse not loading on boot, website randomly not loading, sometimes windows taking longer than normal to boot etc.

            if you can find some of these and document them there might be a precise solution you can find online.

  • nushydude on 28/08/2019 - 08:19

    Didn't read all the comments so might have been mentioned before.

    I've seen this happen when the core overclock of my GTX 1080 wasn't stable. Specifically on F1 2018 DX11 mode. Lowering the overclock fixed it.

    I know you aren't overclocking, but this probably means that the GPU cannot run at the rated clock speed. Something you can try to verify this is underclocking the GPU core by 100MHz or more and see if the issue goes away. You can use MSI Afterburner to do this. If the issue goes away, that means it's a hardware fault (doesn't work at the advertised specs) and you need to get it replaced.

    If it doesn't at first, I would lower the clocks a but more and see. It could be the GPU memory clock speed but unlikely.

    I seriously doubt it's CPU or platform related. Has to be a dud GPU.

    • mememan212 on 28/08/2019 - 17:44

      Tried some of that earlier in the year. Thought it was a dud GPU at first too, sent it to them and over 2 months they havent found a problem with the 2080. Not to mention i've goten this error with another GPU (my old 760).

  • the Unforgiven on 10/07/2021 - 20:36

    I wonder if the voltage is fluctuating coming out of the wall/plug, ie. not 230v. maybe too high or low

    • spackbace on 12/07/2021 - 14:33

      Post is 2yrs old, and OP hasn't been around since 2019…

      • the Unforgiven on 13/07/2021 - 17:33

        Why don't they weed out these old posts so us old farts with bad eyes don't get caught out then?

Login or Join to leave a comment
Login Join