i5-6600k build, plagued by mystery issues

[Karn-]

hi all, here's a bit of a rundown - thanks for taking the time to read if you stick around

Specs:

i5-6600k
Noctua NH-d14 w/1151 mounting kit
MSi z170a pc mate w/latest bios
Corsair LPX 2x8gb DDR4 2133
EVGA GTX 970 ACX 2.0
Corsair Force GT 120 SSD w/up to date firmware
1tb WD black drive
1tb WD green drive
Corsair TX750 psu
Corsair 500r case

I went from an AMD 8320 and an asus M5A97 to this intel setup about a month ago, and for 3+ weeks I didn't have a single issue - clean install of windows, DDU graphics driver install, chipset drivers and bios and firmwares all up to date. played games for hours and day to day use was 100% stable.

This was at a 4.5ghz overclock, with the ram running at 2400 with slightly lowered timings. it was fantastic.

Then, seemingly out of nowhere, I started getting bluescreens with IRQL_NOT_LESS_OR_EQUAL.

I couldn't figure it out, and then it went away for a couple days so I left it alone.

Then they came back, and worse, the computer would just hard lock and not bluescreen, and the reset button would not work. I had to hard power off and boot back up.

I spent some time looking through dump files with whocrashed, looking at eventviewer and msinfo32, all to no avail. The dumps always said the same thing.

Eventually, upon a hard lockup, the computer would go into an "overclocking failed" boot loop, even if I set the bios to optimized defaults.

This is where it got weird, because at this point I decided to start fully troubleshooting which meant turning off all overclock or such related settings and starting to crackdown on these crashes.

Didn't matter, still got the "overclocking settings failed, please run setup" boot loops, of which I could eventually get into windows and the system would then run for an arbitrary time before crashing.

So onward to testing

Memtest passed with flying colors for both sticks, together and individually.

Stress tests and the intel cpu diagnostic program all passed. Nary a single hiccup or temperature issue or lockup.

And then I would crash while reading a reddit thread listening to music, with no games or stress tests running.

I cannot for the life of me figure out what this is. I've reinstalled windows, I've done all the tests, I don't have any strange software or USB devices hooked up, my ram passed memtest and is on the compatibility list for my motherboard, my temperatures OC or not are 100% fine (4.5ghz overclock never passed 60c under full load, thanks noctua!) and all my bios/firmware/drivers are clean and up to date.

This is becoming a serious issue because the crashing can go days without happening, and then happen within minutes of booting for an entire day with no relent, regardless of overclock or devices or anything else.

Here is some of what whocrashed has spit out amongst crashes, if you guys need anything else please let me know, I would really love to get this figured out as I don't really have the option to just start replacing parts.
____________________________________________________
On Mon 5/30/2016 9:24:35 PM GMT your computer crashed
crash dump file: C:\WINDOWS\Minidump\053016-6578-01.dmp
This was probably caused by the following module: ntkrnlmp.exe (nt!KeBugCheckEx+0x0)
Bugcheck code: 0xC000021A (0xFFFFC000C4863700, 0xFFFFFFFFC0000428, 0x0, 0x15C91210000)
Error: STATUS_SYSTEM_PROCESS_TERMINATED
Bug check description: This means that an error has occurred in a crucial user-mode subsystem.
There is a possibility this problem was caused by a virus or other malware.
The crash took place in the Windows kernel. Possibly this problem is caused by another driver that cannot be identified at this time.

____________________________________________________
On Sat 6/4/2016 7:07:34 PM GMT your computer crashed
crash dump file: C:\WINDOWS\Minidump\060416-6109-01.dmp
This was probably caused by the following module: ntkrnlmp.exe (nt!KeBugCheckEx+0x0)
Bugcheck code: 0xA (0xFFFFF803207A0A10, 0x2, 0x1, 0xFFFFF8033051FB10)
Error: IRQL_NOT_LESS_OR_EQUAL
Bug check description: This indicates that Microsoft Windows or a kernel-mode driver accessed paged memory at DISPATCH_LEVEL or above.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
The crash took place in the Windows kernel. Possibly this problem is caused by another driver that cannot be identified at this time.


____________________________________________________
On Sat 6/4/2016 7:07:34 PM GMT your computer crashed
crash dump file: C:\WINDOWS\memory.dmp
This was probably caused by the following module: ntkrnlmp.exe (nt!KeBugCheckEx+0x0)
Bugcheck code: 0xA (0xFFFFF803207A0A10, 0x2, 0x1, 0xFFFFF8033051FB10)
Error: IRQL_NOT_LESS_OR_EQUAL
Bug check description: This indicates that Microsoft Windows or a kernel-mode driver accessed paged memory at DISPATCH_LEVEL or above.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
The crash took place in the Windows kernel. Possibly this problem is caused by another driver that cannot be identified at this time.

I've also attached the two dumps, just not the memory.dmp because its 650mb.

Hope you all are having a great weekend, thanks for reading.

 

[Araxie]

IRQL_NOT_LESS_OR_EQUAL always tend to point to system RAM issues, the cases where this isn't the issue, then it's a Daily OS kernel corruption. so which it's exactly where the crash dump its pointing as "the crash took place in the windows kernel". so at this point your options are,

1) Problem with RAM Stability.

2)OS Kernel Corruption.

3) Any device with driver corruption.

So that mean start from Zero with fully clean install of windows and re-download of drivers.. and if the problems persist then the only left factor would be RAM..

 

[Karn-]

I've done fully clean installs with freshly downloaded drivers 3 times, and I've passed my RAM through memtest.

I guess I should just contact corsair and/or get it replaced and see if it still happens, regardless of the sticks passing?

 

[killianss]

I wonder if it would help to try running each stick individually and seeing if it crashes with one or the other

 

[primetime]

underclock your cpu and see what happens before going any further on ram troubleshooting. I have seen way to many skylake threads where it turned out the cpu needed rma'ing. If the problem goes away when sysem is underclocked then i would check into returning it or rmaing thru intel.

You should be able to downclock the ram as well if downclocking the cpu has no effect. One or the other should give some results. What i can say is these modern cpus are no where near as durable from the generations im still using......stuff like prime95 and what not can damage those new ones with out even knowing it and lead to it giving the instability issues you have.

Have a look at this thread...Windows 10 blue Screen Loop

 

[N4CR]

Win7?

Had practically same issue on a 6700k box. Did some updates it helped but it's a driver issue and the fact that win7 doesn't support properly skylake.

Going to migrate to win10 as I CBF going through all the drivers, as the driver diagnostic tool crashes out windows every time.

Memtest stable for 2 full passes.

 

[Nenu]

Sounds like a memory issue.
Some ram cannot run at rated speed on certain motherboards.
BIOS updates may help.

I had the same trouble with Corsair LPX DDR4 and my 6600K on an ASRock Z170 Extreme 4+ and an MSI Z170a Gaming Pro.
I had to run the memory slower than rated speed, no other option to get it stable.
Corsair told me I had to pay for shipping to Holland from UK to get a replacement.

 

[Araxie]

I've done fully clean installs with freshly downloaded drivers 3 times, and I've passed my RAM through memtest.

I guess I should just contact corsair and/or get it replaced and see if it still happens, regardless of the sticks passing?

as was recommended above, try with each stick individually, maybe one is faulty.. sometimes even could be a motherboard RAM slot faulty, but at this point to me its RAM related the issue, at what voltage are the RAM working? sometimes with just a little 0.05V bump problem are fixed.

 

[Karn-]

Thanks everyone for the replies, I'm on W10 to clear that up.

At this point, I have not crashed and I'm pretty sure it was a (sometimes?) faulty stick of RAM.

I had done the overnight memtest on both sticks, together and individually, and all passed. I actually did the individual test twice per stick. Passed.

A day after making this thread I redid the individual test and one of the sticks repeatedly threw up a slew of errors and actually locked up memtest, forcing a hard reboot - just like I would experience in windows.

I've been running the machine for about 36 hours on the other, working stick of ram, and I've not had an issue, but I'm still not convinced as I have gone up to 9 days previously without a crash while using the supposed "faulty" stick of ram.

I've done an RMA through Amazon and they are going to overnight me a replacement set, which I will verify and send them the old ones.

I had worried about having to RMA the cpu, and/or dealing with underclocking, but considering just how stable the machine seemed to be under the overclocks and voltages, outside of the absolutely random crashes, I really am curious if it all comes down to that finnicky stick of ram that only wanted to work 2/3rds of the time, and crash the other 3rd.

I've heard of bad ram, but never something so inconsistent as this one.

I'll consider the problem fixed for now, but I will come back and update the thread once I get the other set of ram and verify that its stable and have gone without a crash, just in case anyone else has the problem and finds this thread.

Thanks again to everyone for the suggestions, I guess the people who claim memtest needs to run for 72 hours aren't really kidding, are they? l: