hate these ads?, log in or register to hide them
Page 1 of 2 12 LastLast
Results 1 to 20 of 22

Thread: PC dying?

  1. #1
    Ancient Ways's Avatar
    Join Date
    July 2, 2011
    Posts
    108

    PC dying?

    Hello all, just trying to get some feedback before I spend much more money. I have a decent gaming rig for several years now (Windows 10 Pro, i5-3570k, 16 gb ram and now a gtx 970 (got one 2nd hand from a friend over a year ago) and had no problems until recently. I had an SSD running the OS and a 2 TB hybrid drive for games and other programs. Just before Christmas the system started slowing down and in particular anything on my secondary drive started taking a long time to do anything. I figured the drive might be dying and used an older drive that I cleaned and formatted to replace it and everything seemed fine when I left town to visit family. After the new year I was back in town and the rig was fine. Work started a few days after that and then the system died. Came home and brought the system out of sleep via spacebar and nothing happened. Computer had power, but neither monitor changed from black, even after cycling power on them. Fans were running, but no activity on hdd light. I left the system overnight (it was already quite late and I had work in the morning) and there was no change so I risked a reset. At that point my SSD drive with the OS was not detected by bios at all although it was detecting the secondary drive. Swapping cables did nothing, the SSD remains dark

    So I figured that was that, the SSD died on me. I bought a new one and had it shipped to me. I was able to borrow a laptop and get a usb windows install. I was able to install windows to both the new SSD and later the secondary HDD (to see if the new SSD was bad), but new problems sprang up all over. I was able to update via windows update, but while quick scan worked, full scan on windows defender always hangs and stops at some point. Installing new gpu drivers via GeForce Experience and direct download fail almost everytime (I this point I have formatted and reinstalled windows multiple times) and even if get them in, the two games I installed (Prey and World of Warships) both crash as does a benchmark for FFXIV that I downloaded. I had several bluescreens of death, the one I caught said Clock Watchdog Timeout for the error. I'm fairly stumped at this point. MemTest 86 came up fine as did the intell CPU test program. However Event viewer is showing multiple instances of

    Reported by component: Processor Core
    Error Source: Corrected Machine Check
    Error Type: Internal parity error
    Processor APIC I 0

    And the system has crashed/restarted several times. Given the tests above, I was thinking the GPU might be failing, but honestly I have no idea at this point. Anyone think of something that will cover those bases or is it time to just do a complete rebuild from the ground up.
    Last edited by Ancient Ways; January 17 2018 at 04:06:31 PM.

  2. #2
    Donor
    Join Date
    April 9, 2011
    Posts
    1,359
    When you get loads of different, seemingly unconnected problems, it's usually the PSU but can occasionally also be the motherboard. I'd change the PSU and see how it goes. You can at least reuse that if you decide to go for a rebuild.

  3. #3
    Daneel Trevize's Avatar
    Join Date
    April 10, 2011
    Location
    T L A
    Posts
    12,378
    Apparently Clock Watchdog Timeout can be caused by a bad/corrupted driver. The full fresh OS install on the new drive, combined with a GPU driver from before issues appeared, should be able to eliminate that.
    If you've not yet installed an AV, try going without one for a testing period too, as another listed cause of this memory corruption.

    Do you have the latest firmware for your mobo and drives (BIOS and SSD)?
    Quote Originally Posted by QuackBot View Post
    Idk about that, and i'm fucking stupid.

  4. #4
    Ancient Ways's Avatar
    Join Date
    July 2, 2011
    Posts
    108
    I'm currently updating my mobo firmware and will check my drive's once it finishes (it seems to be taking a long time, which worries me, over 10 minutes now). I'll see if I can borrow a PSU for tests.

  5. #5
    Daneel Trevize's Avatar
    Join Date
    April 10, 2011
    Location
    T L A
    Posts
    12,378
    It can be slow, it varies per manufac, just don't bloody interrupt it until you're certain that's the best option. Like, give it over an hour if it's not obviously hard-locked.
    Quote Originally Posted by QuackBot View Post
    Idk about that, and i'm fucking stupid.

  6. #6
    Ancient Ways's Avatar
    Join Date
    July 2, 2011
    Posts
    108
    Oh, I'm not touching a damn thing; it is still moving an icon on the updater and task manager shows it using different amounts of cpu, so just waiting for the very small file to finish.

  7. #7
    Ancient Ways's Avatar
    Join Date
    July 2, 2011
    Posts
    108
    Well its been well over 3 hours, and the firmware update has not completed. METool is bouncing around .5 and 1% cpu use, but console window host and Intel Firmware Update Utility are both solidly 0% and I have not seen them move yet. I'll give it a touch longer, but it seems the actual firmware updated is not running.

  8. #8
    Daneel Trevize's Avatar
    Join Date
    April 10, 2011
    Location
    T L A
    Posts
    12,378
    I'd be scouring for logs/errors & contacting support, just a guess but it might well be detecting a problem that keeps causing some checksumming to fail and caution keeps the updater stalled/looping. Or it could be some dumb permissions/BIOS security setting thing, support should know.
    How long/intensively did you run that memtest for?
    Quote Originally Posted by QuackBot View Post
    Idk about that, and i'm fucking stupid.

  9. #9
    Ancient Ways's Avatar
    Join Date
    July 2, 2011
    Posts
    108
    BIOS updated, tried updating firmware again this morning, but its still not picking anything up. I ran the memtest (took at least 90 minutes, iirc) until it completed everything and gave me results, same with the Intel cpu test.

  10. #10
    Ancient Ways's Avatar
    Join Date
    July 2, 2011
    Posts
    108
    Progress of a sort: ran some Prime95 stress test which fail within minutes but yielded 2 bits of information. The first failure gave me:

    Taskmgr.exe - application error the memory could not be written

    and event log gave me

    Windows cannot access the file for one of the following reasons: there is a problem with the network connection, the disk the file is stored on, or the storage drivers installed on this computer, or the disk is missing. Windows closed the program PRIME95 appliation because of this reason.


    I'm leaning towards mobo/cpu issue now since i doubt both drives I've used since my SSD crashed completely are bad, but its hard to say.

  11. #11
    Daneel Trevize's Avatar
    Join Date
    April 10, 2011
    Location
    T L A
    Posts
    12,378
    Some mobos have extra SATA controller chips for the last few sockets, perhaps see if you can swap cables around between going through that or using the controller in the main chipset/cpu.
    The only other way I can think of to test chipset IO components would be transfering some large files (ISOs) via USB or Ethernet and checksumming the results.
    That the memtest works somewhat rules out a CPU issue.

    Have you tried running just one of your drives connected at once? Rereading the OP makes me wonder if the one that seemed to first die was perhaps not the culprit, or only thing at fault. I don't know if the other drive failing could cause some power issues on the 5/12V rail that would mess with all other drives on the same rail.
    Possibly getting a Linux Live ISO with badblocks or some other drive stress-tester included/able to be downloaded to ramdrive might be a way to test each drive individually.

    I'm assuming you have any data you care about backed up on a different machine/unconnected drive.

    Edit: googling hints another idea might be to ensure no overclocking/actually underclock your RAM/bump up you Vcore slightly.
    Or that you separately had a drive fail, and then the new Meltdown/Spectre patch is causing you issues.

    What was it that made you suspect your GPU? Just that the drivers weren't installing?
    Last edited by Daneel Trevize; January 18 2018 at 05:48:23 PM.
    Quote Originally Posted by QuackBot View Post
    Idk about that, and i'm fucking stupid.

  12. #12
    Ancient Ways's Avatar
    Join Date
    July 2, 2011
    Posts
    108
    I currently only have my hybrid HDD hooked up, running windows off that. Everything is backed up except for the undetectable ssd, which I'll deal with later. The GPU thought was because the biggest issue (sudden program halts and one bluescreen) only seemed to occur when in a game (only played a few minutes until they crashed) as well as the drivers refusing to install (all other drivers via windows update seem to install fine). I should get a PSU to swap in tomorrow, see if that clears anything up, but mobo and/or cpu seems more likely at this point.

    Edit:

    It has been recommended from another forum to try a ubuntuu live disk/usb and boot from that for a few hours to see if its hardware/software. I don't think its software since I've reformatted the drive several times and done new installs on top of that.
    Last edited by Ancient Ways; January 18 2018 at 06:33:22 PM.

  13. #13
    Daneel Trevize's Avatar
    Join Date
    April 10, 2011
    Location
    T L A
    Posts
    12,378
    But have you tried running off of only a different drive?
    Imagine your current one has some issue where it locks up the system when trying to read a specific set of sectors, where some DirectX driver is located. That'd be the same place the install and the game are going to try access, and explain why memtest works. Though it wouldn't immediately explain the Prime95 failure. Still, I'd try either using only a different drive, or purely a Live install. IIRC you can get many flavours with non-free 3D drivers included or able to be installed to a Live system running purely from ISO+RAM, which will let you test some 3D games (you can get Steam for Linux too), and/or maybe just a 3D compositing window manager would be enough.

    Did you try underclocking the CPU+RAM/tweaking the Vcore slightly up or down?
    Quote Originally Posted by QuackBot View Post
    Idk about that, and i'm fucking stupid.

  14. #14
    Ancient Ways's Avatar
    Join Date
    July 2, 2011
    Posts
    108
    I have tried running off only one drive (both the new ssd and the hybrid), both had issues with crashing programs. I'll try messing with underclocking when I get home. Is that through BIOS voltage, never done it before.

    Edit: I see there is a couple programs to underclock or through bios; which is safer and what are your recommended voltages?
    Last edited by Ancient Ways; January 19 2018 at 01:04:00 AM.

  15. #15
    Daneel Trevize's Avatar
    Join Date
    April 10, 2011
    Location
    T L A
    Posts
    12,378
    I'd expect it's best to do it via the BIOS, and it'll depend on the CPU, mobo and RAM you have as to what their common quirks are. It should be quite easy to google the default values, ensure you're not above those, then drop the Ghz >10%. It'll depend on the combo as to whether it'd be better to keep power and timing values otherwise stock, or to ease them off too, in order to create more tolerance of desynchronisation.
    If such settings then cure your issue, you can present that to your manufacturers & see where they point the blame & suggest how to test/RMA next, and/or keep the underclocked setup if it's performant enough.
    Quote Originally Posted by QuackBot View Post
    Idk about that, and i'm fucking stupid.

  16. #16
    Ancient Ways's Avatar
    Join Date
    July 2, 2011
    Posts
    108
    Ok, finally got a chance to underclock; these are all the settings I changed (noticed RAM speed was not see to auto, so set it to that) cpu voltage set to offset and auto and then cpu ratio from 34 to 33. And just like that the FFXIV benchmark completed for the first time. Event log had no errors as well. Going to try loading Prey (only other gpu/cpu intensive thing installed right now) and see if it crashes or is stable. Many thanks for help so far.


    Edit: Well, so far so good. FFXIV benchmark completed twice with no shutdowns, Prey ran fine for the 5 minutes or so I played it and Prime95 has been chugging away for 40 minutes with no errors and multiple self-tests passed. Now to shut everything down and try to reinstall my SSD.
    Last edited by Ancient Ways; January 19 2018 at 05:42:46 PM.

  17. #17
    Daneel Trevize's Avatar
    Join Date
    April 10, 2011
    Location
    T L A
    Posts
    12,378
    Quote Originally Posted by QuackBot View Post
    Idk about that, and i'm fucking stupid.

  18. #18
    Ancient Ways's Avatar
    Join Date
    July 2, 2011
    Posts
    108
    System seems stable. Was able to play 2 rounds in WoWS without problems and installed a couple programs with no errors before I went to work last night. I'll test and install a few more things, but things are looking up, many thanks for all the help.

  19. #19
    Daneel Trevize's Avatar
    Join Date
    April 10, 2011
    Location
    T L A
    Posts
    12,378
    Maybe that'll be the last Intel CPU you buy?
    Quote Originally Posted by QuackBot View Post
    Idk about that, and i'm fucking stupid.

  20. #20
    Ancient Ways's Avatar
    Join Date
    July 2, 2011
    Posts
    108
    Quote Originally Posted by Daneel Trevize View Post
    Maybe that'll be the last Intel CPU you buy?
    Haha, we'll see when its time to upgrade again. First time I've had a problem with an Intel CPU and I've been building my own gaming pc's for 15 years or so. Motherboards I've had die a couple times though.

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •