I just built a new PC and am dealing with some incredibly frustrating recurring failures. I'd love to know what my next steps should be to isolate where my failure is; I'm not far from RMA'ing the whole build and going for a nice long walk.
First, stats:
OS: Windows 10 Pro (retail copy, not upgrade)
CPU: i7-6700K w/ Hyper 212 EVO
MOBO: Gigabyte Z170MX-Gaming 5 && MSI Z170M Mortar
RAM: Corsair Vengeance LPX 32gb (2x16) && G.Skill Ripjaws V 16gb (2x8) (both DDR4 @ 2400)
SSD: Samsung EVO 850 500gb
PSU: Corsair CX750M (one old used one, then another new from the store)
Case: Thermaltake Core V21
No wifi, No GPU, no optical drives, no additional peripherals.
Minidumps (this is a small sample from a few hours of working, but pretty consistent with pattern): https://www.dropbox.com/s/qhhdys3rlnr5kn2/dumps-11-15.zip
Local tests:
sfc /scannow (no invalid areas found)
chkdsk (no issues detected)
dism /restorehealth (went through the process, but made no changes)
History:
This was a mostly new build I knocked out over the weekend. It started with the Z170MX mobo, Corsair 2x16 RAM & my CX750m PSU from my previous PC (the PSU itself is a couple years old). All parts other than the PSU are brand new, opened-the-box-myself.
POST'ed on the first try with no problems, Windows 10 install from a USB key (created from Microsoft ISO download & Rufus bootable USB maker) went as expected. Within 5 minutes I had my first BSOD, and they've never stopped. I can turn the PC on and watch it bluescreen at random intervals all on its own, with varying messages. Occasionally I'll get maybe 10 minutes of zippy use out of it before it fails. I've been unable to isolate a particular app or action that triggers it, but there's no warning or slow-down. It's an immediate failure, dump & restart.
Common BSOD errors include:
- IRQL_NOT_LESS_OR_EQUAL
- SYSTEM_SERVICE_EXCEPTION
- MEMORY_MANAGEMENT
- SYSTEM_THREAD_EXCEPTION_NOT_HANDLED
... along with a handful of others. Most don't have associated files on the BSOD. BlueScreenViewer indicates most of them source from `ntoskrnl`, which I understand is core to Windows and probably a red herring.
I've tried all of the following:
- Reset Windows
- Refresh Windows
- Format & clean install Windows
- Replace the PSU (bought a brand new CX750M and replaced the PSU & cables)
- Swapped the motherboard (Gigabyte Z170MX -> MSI Z170M)
- Moved/reseated RAM
- Tried only one stick of RAM
- Moved one stick of RAM between slots.
- Enabled/disabled XMP to knock RAM to 2400
- Run ram at default speed (2133)
- Tried new RAM entirely (Corsair 2x16 -> G.Skill 2.8, both DDR4 @ 2400)
- Repeated all steps with the new RAM
- Update BIOS (on both boards, incrementally through updates to see if a particular rev stabilizes)
- Update drivers (manufacturer sourced, via Driver Verifier, etc. No unrecognized devices)
- Tried a few different SATA cables and moved to different SATA ports
- Swapping keyboards
- Swapping mice
- Swapping monitors (tearing my hair out at this point)
- Unplugging all case-related peripherals (USB font-ports, LEDs, etc)
- Plugging into a different surge protector
- Plugging into an entirely different circuit in a different part of the house
None of this has affected the errors: they're still the same, even after swapping out the mobos & RAM. The only portion of this build that I haven't attempted to replace at least once is the i7 itself & the SSD. I can definitely go grab another SSD if that's a possible point of failure, but I'm timeboxing this build to another week. If I'm not on track to resolve it by then I'm just going to buy a prebuilt from someone so I don't feel so bad about myself when it doesn't work.
I'm STUMPED. I know I've gotta be missing something obvious - I've never struggled with a build issue this badly. Does anyone have any tips/suggestions/thoughts about where else I should be debugging this? Nothing's too crazy to try, and I'm happy to provide whatever logs or output might help track down what's up Thanks for sticking with me!
First, stats:
OS: Windows 10 Pro (retail copy, not upgrade)
CPU: i7-6700K w/ Hyper 212 EVO
MOBO: Gigabyte Z170MX-Gaming 5 && MSI Z170M Mortar
RAM: Corsair Vengeance LPX 32gb (2x16) && G.Skill Ripjaws V 16gb (2x8) (both DDR4 @ 2400)
SSD: Samsung EVO 850 500gb
PSU: Corsair CX750M (one old used one, then another new from the store)
Case: Thermaltake Core V21
No wifi, No GPU, no optical drives, no additional peripherals.
Minidumps (this is a small sample from a few hours of working, but pretty consistent with pattern): https://www.dropbox.com/s/qhhdys3rlnr5kn2/dumps-11-15.zip
Local tests:
sfc /scannow (no invalid areas found)
chkdsk (no issues detected)
dism /restorehealth (went through the process, but made no changes)
History:
This was a mostly new build I knocked out over the weekend. It started with the Z170MX mobo, Corsair 2x16 RAM & my CX750m PSU from my previous PC (the PSU itself is a couple years old). All parts other than the PSU are brand new, opened-the-box-myself.
POST'ed on the first try with no problems, Windows 10 install from a USB key (created from Microsoft ISO download & Rufus bootable USB maker) went as expected. Within 5 minutes I had my first BSOD, and they've never stopped. I can turn the PC on and watch it bluescreen at random intervals all on its own, with varying messages. Occasionally I'll get maybe 10 minutes of zippy use out of it before it fails. I've been unable to isolate a particular app or action that triggers it, but there's no warning or slow-down. It's an immediate failure, dump & restart.
Common BSOD errors include:
- IRQL_NOT_LESS_OR_EQUAL
- SYSTEM_SERVICE_EXCEPTION
- MEMORY_MANAGEMENT
- SYSTEM_THREAD_EXCEPTION_NOT_HANDLED
... along with a handful of others. Most don't have associated files on the BSOD. BlueScreenViewer indicates most of them source from `ntoskrnl`, which I understand is core to Windows and probably a red herring.
I've tried all of the following:
- Reset Windows
- Refresh Windows
- Format & clean install Windows
- Replace the PSU (bought a brand new CX750M and replaced the PSU & cables)
- Swapped the motherboard (Gigabyte Z170MX -> MSI Z170M)
- Moved/reseated RAM
- Tried only one stick of RAM
- Moved one stick of RAM between slots.
- Enabled/disabled XMP to knock RAM to 2400
- Run ram at default speed (2133)
- Tried new RAM entirely (Corsair 2x16 -> G.Skill 2.8, both DDR4 @ 2400)
- Repeated all steps with the new RAM
- Update BIOS (on both boards, incrementally through updates to see if a particular rev stabilizes)
- Update drivers (manufacturer sourced, via Driver Verifier, etc. No unrecognized devices)
- Tried a few different SATA cables and moved to different SATA ports
- Swapping keyboards
- Swapping mice
- Swapping monitors (tearing my hair out at this point)
- Unplugging all case-related peripherals (USB font-ports, LEDs, etc)
- Plugging into a different surge protector
- Plugging into an entirely different circuit in a different part of the house
None of this has affected the errors: they're still the same, even after swapping out the mobos & RAM. The only portion of this build that I haven't attempted to replace at least once is the i7 itself & the SSD. I can definitely go grab another SSD if that's a possible point of failure, but I'm timeboxing this build to another week. If I'm not on track to resolve it by then I'm just going to buy a prebuilt from someone so I don't feel so bad about myself when it doesn't work.
I'm STUMPED. I know I've gotta be missing something obvious - I've never struggled with a build issue this badly. Does anyone have any tips/suggestions/thoughts about where else I should be debugging this? Nothing's too crazy to try, and I'm happy to provide whatever logs or output might help track down what's up Thanks for sticking with me!