BSOD several times a day for no apparent reason

Nathan K

Reputable
Jul 29, 2015
25
0
4,540
Ok so my computer is crashing with the BSOD far too regularly lately. Several times a day sometimes. Always the same error code: WHEA_UNCORRECTABLE_ERROR
Google search has not helped, though it has led me to suspect *possibly* something to do with my processor cache? This is why I've posted this in the CPU thread. If this is the wrong place, my apologies.
Need to fix this as it's rather annoying. Usually I'm good with computers but haven't managed to fix this one yet.
Seems to happen completely randomly, and not specifically when my pc is under load.
How do I go about pinpointing the issue and fixing it?
What information do I need to provide you technical wizards with to help?
 

PreevBR

Commendable
Feb 22, 2016
163
0
1,710


Not CPU, but rather a hard drive/storage problem/failure. Did you try running chksdk? (Go to starting menu, search for CMD, open it and type "chkdsk /r" in the new window, then press enter and let the test check your hard drives. [It may also ask you to restart your computer to start the testings so if asked agree.])
 

Nathan K

Reputable
Jul 29, 2015
25
0
4,540


I have 4 drives in my pc - will this check all of them or just my C: drive?
UPDATE - ran it, seems to have only run on my C: drive but it reached 100% and then booted into windows. Doesn't seem to have stated any results of if it found or fixed any issues, guess I'll just have to wait and see if it blue screens again.
On a side note, my S.M.A.R.T values for all my drives seems fine and chkdsk without the /r suffix ran fine and reported no issues
 
WHEA_UNCORRECTABLE_ERROR is a tough one to nail down; it basically means that Windows thinks there's a hardware problem somewhere.

General steps to try, include checking HDDs with chkdsk, running a memory test via memtest86+, ensure you aren't overheating, and voltages are stable.
 

Nathan K

Reputable
Jul 29, 2015
25
0
4,540


I booted into the advanced startup options to use the command prompt and ran chkdsk /r on all 4 of my drives, every single one came back fine, no problems (though it took flippin ages on my 2TB HDD). So I think I can safely rule out my HDDs/SSDs.
I'm using a water cooler and even on stress tests I've never seen the temp go above 65C, so I think I can rule out overheating too.
Getting memtest86+ now to try that and see if it finds anything...

How do I check my voltages? The BIOS in my ASUS motherboard had an "EZ Tuning" feature which I used to overclock my processor (i5 4690k) and RAM and haven't made any changes to settings myself except to enable my onboard graphics as well as discrete to use a 4th monitor. I could plug this 4th monitor into my graphics card if that will help, pretty sure my GTX 970 supports 4 monitors. I just decided to use the 4th with my processor graphics because I wasn't sure if it would affect gaming performance (I only game on one monitor anyway, but have task manager, HWMonitor and ASUS GPU Tweak displayed on the other monitors while I game to keep an eye on temps and usage)
Only thing I can think of from google searching is that I've overclocked my CPU multiplier to 45x (reference clock 102MHz), but the processor cache is still at 38x. I read somewhere that it should be within 500MHz of my cpu core speed (4590 Mhz) or something? Could this possibly be the issue? The voltage is set to Auto for the cache and adaptive (1.275v) for the cores.

UPDATE: Ran memtest 86+, no errors found
 

Nathan K

Reputable
Jul 29, 2015
25
0
4,540


Well what confuses me here is it doesn't seem to happen when the system is under load, I can play the witcher 3 for hours which is quite CPU intensive and it doesn't crash, I've ran system benchmarks with no issue, yet I'll be listening to music and browsing the internet, or maybe running a virus scan or something, and poof BSOD
 

Nathan K

Reputable
Jul 29, 2015
25
0
4,540


It happens at both, though seemingly randomly. During a virus scan at high utilisation it's crashed, and basically just on the desktop at idle it's crashed too. It'll run fine for hours under high or low utilisation fine as well.

Only time it seems to consistently crash is if I run a full virus scan instead of a quick one - I'm using 360 total security and the last 3 times I've ran a full computer scan it's blue screened on me. This never happened before, and I've had the pc at these OC settings for about a year and been using this software for months as well. I always run a full scan once a month and it's never done this before but as I said now it's done it 3 times in a row today.
But it also happens at other times randomly.
Quite confused with this....

Update: once memtest86+ finishes running I'll restore bios to defaults (no overclock) and try running a full system scan in 360 to see if it crashes then.
 

Nathan K

Reputable
Jul 29, 2015
25
0
4,540


Opened Intel Extreme Tuning Utility, changed my baseclock back to 100mhz but left the multiplier at 45x, overclocked CPU Cache multiplier from 38x to 40x (to keep it within 500mhz of CPU core speed as I read somewhere) and increased the cache voltage to 1.25v adaptive. (Will probably reduce it in increments over the next few days til I find a lower stable voltage)
Seems to have fixed my issue, and I've only had to reduce my OC by 90MHz.
Couldn't say specifically which change fixed the issue but I suspect increasing the cache voltage is what actually did the trick. I could go back and change each setting one by one with tests to see if it blue screens, but who has that much time really?
All I know is the one consistent trigger I had (running a full virus scan) no longer results in BSOD.
If I get another blue screen crash over the next few weeks I'll post again.
 

Nathan K

Reputable
Jul 29, 2015
25
0
4,540


Turned voltages down to 1.25v for cores and 1.20v for cache and seems to still be stable, windows booted at least. Hopefully I can continue using these values and the system will stay stable