1. OCAU Merchandise is available! Check out our 20th Anniversary Mugs, Classic Logo Shirts and much more! Discussion in this thread.
    Dismiss Notice

New 3700x build crashing w/optimized defaults in OCCT CPU test

Discussion in 'Troubleshooting Help' started by randomman, Jul 9, 2021.

  1. OP
    OP
    randomman

    randomman Member

    Joined:
    Oct 21, 2007
    Messages:
    5,279
    Motherboard back with the store for RMA, they said Gigabyte could choose to do repair, replace or credit; at their cost, not MSRP, WTF is this shit? Will wait and see what happens. If they give me another board and aren't helpful in any way I'll just flat out tell them I'm leaving them a bad review and never spending my money their again. I was less than a month outside of their return window when I first reported their issue and I spent over $1k.
     
  2. OP
    OP
    randomman

    randomman Member

    Joined:
    Oct 21, 2007
    Messages:
    5,279
    Fucckkkkkkkk. Locked up with new MB, then wouldn't post. Forgot new board had diagnostic LEDs so didn't check them... managed to post after a few power cycles. Updated the UEFI to latest. Then it crashed with IRQL not less or equal BSOD. Rebooted immediately so couldn't check diagnostic LEDs... will just have to wait for it to come back.

    EDIT: IRQL BSOD again, wouldn't post after, diagnostics LEDs stuck on VGA, then power cycle a few times and it was always diagnostic LED RAM....... Pretty much exact same behavior as before so it likely is the RAM (or possibly VGA). Reseated the RAM and back up and running. I upped the DRAM voltage to 1.375 (1.35 stock). I did that on the previous MB so I'm not holding my breath that it'll fix it. It might be worthwhile reaching out to G.Skill and asking them wtf to do.
     
    Last edited: Sep 9, 2021
  3. Elmf

    Elmf Member

    Joined:
    Jan 6, 2007
    Messages:
    2,770
    Location:
    Melbourne
    Borrow some RAM from an OCAUer, chuck it in, hope for the best.

    Same with a GPU. Or buy a $20 one and just use it for a week playing Tetris and see if it crashes
     
    t8y likes this.
  4. t8y

    t8y Member

    Joined:
    May 5, 2002
    Messages:
    4,731
    Location:
    adelate
    i work in an RMA dept for a national supplier - thats GB standard these days. they attempt repair as first priority (due to covid/shortages this makes sense). have noticed some stuff can go back/forth a few times before they finally decide to replace/credit instead, though.
    v frustrating

    since you don't have spare parts;
    tried testing w/ just 1 stick of ram at a time at least? set everything stock. try XMP off and set a low mem multi (so it rules everything memory related out).

    looking more like CPU issue beyond that
     
    Last edited: Sep 15, 2021
  5. Elmf

    Elmf Member

    Joined:
    Jan 6, 2007
    Messages:
    2,770
    Location:
    Melbourne
    I've had similiar sounding issues with a GPU, so not sure, literally could be anything unfortunately.
     
  6. OP
    OP
    randomman

    randomman Member

    Joined:
    Oct 21, 2007
    Messages:
    5,279
    I'm not in Oz so that's not an option and (un?)fortunately hasn't crashed since so it's back to being quite infrequent making single slot RAM/GPU testing quite difficult. I'll continue to wait and see if it increases in frequency like it did on the old MB.
     
    Elmf likes this.
  7. OP
    OP
    randomman

    randomman Member

    Joined:
    Oct 21, 2007
    Messages:
    5,279
    Okay so I replaced my Win11 host with Proxmox (NVMe with Win11 is now passed through to a VM guest, along with GPU - it's sick) and that just went down and the MB diagnostics light for RAM was lit up. Just pulled one stick out and will see if it still happens, if so, switch with other stick and test again.
     
  8. OP
    OP
    randomman

    randomman Member

    Joined:
    Oct 21, 2007
    Messages:
    5,279
    Crashed on the single stick, MB diag LEDs yellow pointing to RAM. Switched it out, will see if this one also crashes. Hopefully not, that would make it simple and maybe G.Skill will let me hold onto it while they replace the dud.
     
  9. Elmf

    Elmf Member

    Joined:
    Jan 6, 2007
    Messages:
    2,770
    Location:
    Melbourne
    Good luck bro hopefully you found it! They do typically make you RMA full kits unfortunately
     
  10. OP
    OP
    randomman

    randomman Member

    Joined:
    Oct 21, 2007
    Messages:
    5,279
    Okay so after some very prolonged testing it appears as though one DIMM was faulty. G SKILL USA agreed to let me RMA only the one faulty DIMM, and even said I could return the working DIMM at a later date for the matching pair. Returned it to the vendor as they handle RMAs for 12 months, they said expect 4-8 week wait. No biggie, 16GB is still plenty.
     
    t8y, M1ck1, BlueRaven and 2 others like this.
  11. OP
    OP
    randomman

    randomman Member

    Joined:
    Oct 21, 2007
    Messages:
    5,279
    New stick arrived and not off to a good start. Instantly crashed as soon as Proxmox loaded and then the board wouldn't POST citing DRAM until I removed the new stick. I tested the working stick, no issues. Added the new stick again, no issues... for now. We'll see if this is resolved.
     
  12. albeeny

    albeeny Member

    Joined:
    Feb 25, 2002
    Messages:
    127
    Location:
    Syd
    I posted before on this thread about having the same issue.... Today, after 3 BSoDs in 2 hours... I cracked the sh!ts, ran out and purchased a new Ryzen 7 5800X.

    Plugged it in with no other hardware changes. Same BIOS setting (default + XMP-3200 + AMD-V)

    2hrs at full tilt, no crash yet.... I'll post back here in 24hrs

    UPDATE: 24hours have passed. The machine has been idling, running idle VMs with high MEM consumption. and it hasn't crashed, stalled or BSOD'd in that time. this is the first time (since it's build in 2019) that it has run for 24hrs without crashing. It appears that replacing the CPU solved the systems instability issues.
     
    Last edited: Jan 7, 2022
  13. t8y

    t8y Member

    Joined:
    May 5, 2002
    Messages:
    4,731
    Location:
    adelate
    missed this one with the seasonals, good news!
    first boot can be touch and go with big changes like memory or CPU. hopefully thats all it was, hiccup w/ initial training on the extra slot.
    nice upgrade :thumbup: gl & maybe a new thread with full specs things tried etc in the OP so everyones on the same page (literally) with where your fault is at
    think of it like opening a new tech support ticket, but crowdsourcing :) keeps mods happy too
     
    randomman likes this.
  14. albeeny

    albeeny Member

    Joined:
    Feb 25, 2002
    Messages:
    127
    Location:
    Syd
    Cheers t8y... :) I'm too shy to start my own thread... was just looking to empathise with OP... perhaps offer an alternative course of troubleshooting.

    Hey Randomman,
    out of interest... what's your CPU batch information? (https://www.reddit.com/r/Amd/comments/6scnlg/ryzen_reading_your_production_batch_number/)

    Mine is
    year = 2019
    week = 23rd
    ATMP (assembly, test, mark, and pack) = Suzhou, China
    Waffer manufacture = Texas
     
  15. OP
    OP
    randomman

    randomman Member

    Joined:
    Oct 21, 2007
    Messages:
    5,279
    So all has been stable so far. I'm running Windows 11 under a headless Proxmox and I've only assigned 24 (of 32GB) to the VM, so it's also possible that the memory isn't being stressed very often if ever. Nothing more to do but continue enjoying no crashes.

    I don't have a pic of the top unfortunately. I'm curious but also too lazy to check lol.
     
    t8y likes this.
  16. albeeny

    albeeny Member

    Joined:
    Feb 25, 2002
    Messages:
    127
    Location:
    Syd
    ...that's great news! Congrats.
     

Share This Page

Advertisement: