I’ve been having an issue with my computer deadlocking after roughly 10-30 minutes of it being on. The simple test was when the computer froze up it didn’t register any inputs, caps lock didn’t change any lights on my keyboard, the computer fans still ran, sound just stopped instead of tonally rattling my headphones like a regular bsod.
I’m not entirely sure what could be causing it other than it’s a hardware issue. PC parts list below.
https://pcpartpicker.com/list/f2hPRV
The things I have tried are as follows:
Booting with my old windows 10 and Linux Mint Windows didn’t seem to record anything major in event viewer.
Switching out my ram cards one at a time and running a check on the sticks themselves. Neither stick appears to be the issue.
Checked my CPU seat for any damaged pins or debris.
Made sure all plugs are properly installed and all drivers are installed.
Flashed bios to both the recommended version and the most up to date version.
Set CPU, GPU, and internal SSD to gen 4 power in the bios.
I also ran a log check on Mint and the only error that it spat out was this:
[Dec 7 13:03] mce: [Hardware Error]: Machine check events logged
[ +0.000006] [Hardware Error]: Corrected error, no action required.
[ +0.000004] [Hardware Error]: CPU:0 (1a:44:0) MC14_STATUS[Over|CE|MiscV|AddrV|-|-|SyndV|CECC|-|-|-]: 0xdc2040000602010b
[ +0.000008] [Hardware Error]: Error Addr: 0x000000000008bdc0
[ +0.000001] [Hardware Error]: IPID: 0x000700b020347000, Syndrome: 0x000000262a1f2603
[ +0.000003] [Hardware Error]: L3 Cache Ext. Error Code: 2
[ +0.000001] [Hardware Error]: cache level: L3/GEN, tx: GEN, mem-tx: GEN
Looking this up seems to spit out a fault in the GPU, but this is way beyond my skill level here.
Thank you for any advise or tips for troubleshooting. I’m still within the return and replace period for my parts so if this indicates anything needing to be replaced, that’s always an option.
Thank you all very much!
Edited for formatting
Don’t put too much faith in Memtest - I’ve had systems that passed a 48-hour test, but would reliably BSOD before Windows could finish booting. A good tool for sure, but not the complete picture.
You certainly need more testing tools. Prime95 is excellent for identifying CPU, RAM, and PSU issues. Furmark for GPU.
The first thing you’ll need to do is further develop your steps to reproduce. Ideally, you’ll have something like “Boot to Windows, launch prime 95, test X. Locks within Y minutes”. You’ll need this to know if you’ve even fixed it.
You mention testing the RAM, but not the slots. Make sure you run your test plan with just RAM 1 in slot A, then just RAM 2 in slot B. It also looks like you have onboard video. Enable it, and test without the video card. Be careful with the results here - that’s probably stressing your PSU more than the entire rest of the system combined.
Finally, if you have a problem that just doesn’t make sense, it’s probably a defective PSU.
Also, due to CPU binning, it’s very unlikely you have a defective CPU. But you could have an incompatible one, and the X3D has had some very specific issues as well.
That’s a crash. A deadlock is something specific: https://en.wikipedia.org/wiki/Deadlock_(computer_science) (your crashes might be the result of a deadlock, but that’s unlikely; usually a deadlocked system is mildly responsive and logs the detected deadlock)
I would try running a stress test on the CPU and GPU and see if you can reliably reproduce it. Try swapping the gpu or using the igpu if possible.
Sorry about the semantics of the issue. It was my regurgitation of what I looked up and what seemed to fit the issue the best rather than “crash” which seems too vague.
I’ll try using the igpu or a really old GPU since those are all I have on hand. Stress isn’t really needed since it locks up while idle. Temps are fine and it just crashes regardless of what I’m doing. I’ve had it crash when I run a command on the terminal; I’ve had it crash when I open Firefox; I’ve had it crash when it’s hibernating.
First thing I do when there’s any weirdness on a new build is run a memtest. I use OCCT but in theory anything can work. Memory errors can cause weird errors that make it seem like it’s another problem. But at the same time CPU errors can make memory errors.
Before you run the mem test make sure your settings are correct. In the bios make sure all auto overclocking features are turned off, voltages are correct and, most importantly, the memory speeds are timing are correct. The auto settings are almost always universally wrong. Also make sure xmp/whatever AMDs xmp is called is turned on.
To my uninitiated understanding of advanced setups for computers, is there any recommended documentation for setting up the correct voltages for specific parts that isn’t just running the automatic settings? I was under the impression that it was fairly plug and play so if the voltages are wrong from the get go then that might significantly contribute to the problem.
For RAM it’s printed on the side of the stick usually. Take a picture of it with your phone and then punch those numbers into the bios. They don’t list all of the numbers which is annoying. ex:
cas: 1st number: 32
Trcdwr ras# pre 2nd number: 39
ras# act: 3rd number: 39 (usually this is a different value)
trc_sm: 4th number: 102
For CPUs that’s a bit more complicated since they don’t brag about that everywhere like ram. But with the latest bios that should use the correct voltage. Usually CPUs voltage is too high and that’s not necessarily a problem, it just runs hotter.
You’ll want to comb over the motherboard’s settings in the BIOS. Some newer boards especially gaming focused ones have dynamic boost clocking on by default and this can push stuff out of spec.
I’ve been having an issue with my computer deadlocking after roughly 10-30 minutes of it being on. The simple test was when the computer froze up it didn’t register any inputs, caps lock didn’t change any lights on my keyboard, the computer fans still ran, sound just stopped instead of tonally rattling my headphones like a regular bsod.
What are the conditions at the time of the crash? Are you actively using it? Is the screen blank, or frozen showing the last activity? Does this occur every time the computer is on, or intermittently? Does it happen in both Windows and Linux? Does it happen if you boot a live OS and leave that running for awhile? Have you done any SMART checks?
Just to further clarify the symptoms, is it true that the following all appear to stop working at the same time?
- Video out
- Audio out
- Mouse in
- Keyboard in
Also what are your mouse and keyboard like? Are they wireless? Is there anything else like a USB hub between them and the motherboard? (or maybe an internal port extender inside the PC case for front panel ports or w/e?) Have you tried a different mouse/keyboard just to check the simple stuff? Is the mouse optical sensor still on after the crash (still getting power)?
It would be useful to strip your motherboard to just the CPU and boot to UEFI/BIOS with only the monitor and keyboard attached. (You might need to have 1 RAM module installed to get the motherboard to start - this depends on some specifics of the chipset and the POST test for your board - install the RAM only if you can’t get to UEFI without it). Does the crash occur in this state?
What are the conditions at the time of the crash? Are you actively using it? Is the screen blank, or frozen showing the last activity? Does this occur every time the computer is on, or intermittently? Does it happen in both Windows and Linux? Does it happen if you boot a live OS and leave that running for awhile? Have you done any SMART checks?
Both when actively using it and leaving it idle when I’m away doing anything else.
The screen freezes the last shown image on it.
Every time the computer is on it will eventually freeze.
Yes both windows and Linux.
Haven’t tried booting a live OS.
It’s likely not an SSD issue because I have booted from two different ssds.
Just to further clarify the symptoms, is it true that the following all appear to stop working at the same time?
Video out
Audio out
Mouse in
Keyboard in
Yes all of these stop working simultaneously.
Mouse and keyboard are both wired connections. And are still receiving power after the crash.
Headphones are wireless.
My next test is booting with no GPU and as barebones as possible, but as others have pointed out in the comments here it seems like it’s more of a CPU issue than not based on the error thrown, but this will at least give solid evidence of that being the case.
I also have to run OCCT as others have pointed out as well. Hopefully those tests aren’t interrupted by the crash.
Yeah it sounds like you’ve eliminated all of the most common possibilities. All of this narrows it down pretty far:
Both when actively using it and leaving it idle when I’m away doing anything else.
The screen freezes the last shown image on it.
Every time the computer is on it will eventually freeze.
Yes both windows and Linux.
Haven’t tried booting a live OS.
It’s likely not an SSD issue because I have booted from two different ssds.
If you’ve already swapped the SSD and reinstalled then the live OS would probably be a waste of time, it really just verifies that the fixed disk is not the source of the problem (although it could be something like the SATA bus, in which case swapping the drive would be irrelevant - low probability on that though).
The barebones test is definitely a good next step. The only other thing I’d suggest is checking the socket real carefully with a flashlight and a magnifying glass (I know you said you checked it already, not clear if that was just a quick look or a full examination). Also make sure the socket itself isn’t cracked or damaged in some way that might cause it to flex and not fully contact the CPU when you lock it down, and that all the pins lift up like they’re supposed to when the lever is folded down.
It also helps to know that my Windows 10 was booted from a SATA connection, but my Linux was booted from an M2 connection. So it’s not that either.
Do the monitors freeze or go black when the computer freezes up on you?
I had issues with a faulty PSU where, under high load but still below max, it would trip the overload safety and basically freeze the PC but fans and everything would stay “on” except the monitors.
Were the Win 10 and Mint separate hard drives? Both SSDs?
Monitors freeze up.
Could be something similar based on what you described. That sounds like the same symptoms.
Win 10 and Mint were on different drives and both SSDs, yeah.
There are a handful of stress tests you can run to have the CPU or GPU or whatever at max load.
A healthy system will run hot but otherwise be fine. A faulty system, especially a PSU, will likely run into the freeze in less than 30 seconds.
There’s also PSU testing hardware that’s basically plug in tests but if I’m being honest I’ve never really messed with them as this hasn’t come up very often for me.
So the crash still occurred when you were running off the older windows install? Just checking. But have you tried running in safe mode and seeing if the issue still occurs?
I would agree with you about a bad CPU. Call the store you purchased it from and ask what their replacement process is.
Try putting only one stick of ram in at a time
Can your PSU handle the hardware? Calculate all hardware max watt or check one of those online calculator (search for PSU calculator)
Check your memory by doing a memory test.
No overheating problems?
PSU should be able to handle it, it’s an 850 watt and the calculator says I should be hovering around 560 watts with no over clocking.
Memory test from windows said the cards were fine.
GPU heat is around 35C
CPU heat maxed out around 45C
