[SOLVED] Feels like a unfixable issue currently, does anyone know if this is a windows issue?

[xMecki]

Ok, so my PC has been triggering tf out of me the past 3weeks, ive been onto multiple forums asking for help, asked close tech savy friends for help and no one has a clue whats wrong with my PC its confused everyone.

Hardware
Geforce rtx 2080 ti gaming x trio
850w gold 80 corsair psu
Kraken x72 water cooling
M.2 1tb samsung evo
H700i case
7 mounted case fans
I9 9900k 3.6GHz cpu
2tb hdd 500gb sdd
Maximus hero xi mb
System is 64bit

Ok so to sum this up ill list everything ive tried, my issue being that my system restarts with no error, just boots back up into the log in page. Ive spent 3 damn weeks troubleshooting this with help and nothings worked and just to say where I am at right now, I am confident this is a Windows issue but my minds been changing bc everytime i fix something it continues but im honestly kinda clueless and done with this system rn.

- To start this off I figured it was a PSU ossue due to kernal power logs in ky windows logs so i bought a new PSU, installed everything from scratch, once booted the issues continued

- Next, my GPU needs 2x 8pins and a 6pin cable, so instead of using 1 6+2 to cover the 2 8pins and an additional 6+2 for the last 6 pin I added 3x 6+2 cables incase it was power related still. No success.

- Then, I reinstalled Windows using the dowbload tool online so my files and apps etc were still safe, again didnt work, restarts started again.

- Nvidia drivers, I uninstalled them all using ddu then reinstalled them back on, made progress tbh, the system didnt restart within 5mins of starting up a game, but restarted over a number of hours later, if i ran stress tests then it would crash within 15mins using valley benchmark.

- Temps, volts etc I checked these using HWInfo, HWmonitor, core temps etc the temps were fine while using valley benchmark along with the power, volts.

- Overclocks, I actually did this earlier on but forgot to add it, everything was set to default, mem clocks, gpu clocks, bios clocks, all reset, turbo was on in bios and ran auto at 4.7GHz, caused restarts still, i disabled turbo and forced 3.6GHz/x36 to be ran, still restarted. Sometimes restarts would go into a state where the gpu fans would power upto 100% as if its using power or overheating.

- Ran virus programmes, also ram ccleaner to fic registry and broken files, no viruses found and no fix to the restarts.

- Wanted to get the RAM out of the way so flashed Memtest86 onto a usb and ran it for 5 hours getting 0 errors from all 4 tests and all 4 sticks.

- I have also reapplied thermal paste to both the CPU and GPU.

- drives, I disconnected my hdd and sdd while leaving in the m.2, still crashed, put in the hdd and toon out the m.2 still restarted. (I still have windows on my hdd due to it being what i used before upgrading, I rarely use the hdd or connect it).

- Ran Prime95 - where i saw issues, cpu was running at 95° after 15mins of stress testing, found out that my kraken radiator fans where connected to different slots, fixed it and got the temps down. Still crashes with low temps.

- Downloaded Pop Os onto a flash drive so i could stress test in linux to rule out a windows issue, ran the stress test for 5 hours without any crashes, closed the stress test and went to sleep leaving idle, woke up and still no crashes on the system, boot back into windows and oh hey lemme crash the system again. (Reason why i believe this has to be a windows issue as ive already troubleshooted every piece of hardware apart from the mb)

- also done alot of things in cmd which i cant fully remember, but ive done sfc scans, ive done drive scans, ive installed drive software.

With the linux test going smoothly i feel like my last option is to fully reset windows and get rid of all my apps etc and hope it fixes, Im not sure what else I can really do, the 2 things I think i have it pinned down to are windows or my gpu, I dont think itll be the cpu as it passed a cpu stress test for 5hours on linux while it crashed on windows and there were no bent pins so everything was good there, also sometimes say after running valley it lasted 30mins in valley before restarting, then when it came back on the restarts happened every few mins on idle, gpu fans powered upto 100% so that also what makes me wonder if its a gpu temp issue, but in HW the temps didnt pass 80° so im confused.

Im out of ideas and would appreciate anyones input on this because rn I feel like ive done everything apart from resetting windows compmetely and changing the gpu for longer (i ran the test with a roomates gpu and it got to 30mins without crashing before i closed it but id need to extend that to make sure its not my gpu thats an issue)

 

[boju]

Wow, ya tried many things, have to applaud you on that.

Hmm, it's possible Windows may be the problem. Try disable automatic device driver updates in system > advanced settings > hardware. Then reinstall motherboard drivers and see how you go.

Do you have a spare hdd to test a fresh Windows install on? Id try this first if you can.

 

[xMecki]

Wow, ya tried many things, have to applaud you on that.

Hmm, it's possible Windows may be the problem. Try disable automatic device driver updates in system > advanced settings > hardware. Then reinstall motherboard drivers and see how you go.

Do you have a spare hdd to test a fresh Windows install on? Id try this first if you can.

Sadly not but my plan tomorrow is to completely wipe my hdd if i can and store any files i need on there then keep it to one side while completely restarting and installing a new windows on my m.2 to see if it works, havent had much exp with it so itll be something ill have to research, but with what you said first ill give it a try and see what happens.

Did forget to mention also as i missed out a few things that ive done, but ive installed a new bios/updated version, i thought i installed new drivers firmware etc but it may have had issues im unsure, but ill give it another try, usually when I dbl tap the asusinstall tool it just shows a loading wheel then nothing afterwards so i dont know it its a background process which i dont see or if i should be getting something to pop up.

 

[boju]

It's recommended to flash bios from a usb stick using ezflash utility in the bios itself. I don't trust doing this from Windows if that's what you did.

After bios update, load bios defaults to make sure cmos is fully updated.

Install chipset, sata, and anything else looking important from the motherboard's website driver page.

 

[xMecki]

It's recommended to flash bios from a usb stick using ezflash utility in the bios itself. I don't trust doing this from Windows if that's what you did.

After bios update, load bios defaults to make sure cmos is fully updated.

Install chipset, sata, and anything else looking important from the motherboard's website driver page.

Yeah thats how i did it sady, flashed bios onto my usb then installed the others from the site

 

[boju]

Yeah thats how i did it sady, flashed bios onto my usb then installed the others from the site

Fair enough.

Yeah try load bios defaults if you hadn't yet since the bios update. Disable device driver updates and try reinstall motherboard drivers again. If it still crashes, full wipe and reinstall probably be the last thing to try.

 

[xMecki]

Fair enough.

Yeah try load bios defaults if you hadn't yet since the bios update. Disable device driver updates and try reinstall motherboard drivers again. If it still crashes, full wipe and reinstall probably be the last thing to try.

Out of curiosity with the driver updates, should I be seeing a dialog box at all when installing or is it kinda like a invisible download if you get what i mean, like background process which i wont see, also any tips for force uninstalling nvidia drivers? Everytime i click uninstall on control panel nothing happens or pops up

 

[boju]

I'd imagine there should be a dialog box yeah where you accept the terms of installation and stuff. User Account Control could be getting in the way if set too strong. Try turn it off.

DDU should be enough to wipe Nvidia drivers but using the custom clean install option should be good enough. DDU is really only necessary if changing brands, AMD to Nvidia etc.

 

[onespeedbiker]

When you are installing windows, do you have any other drives attached? If yes, that is likely your problem. Disconnect all the drives except the target drive and then load windows. Further, any time you have two or more drives with active partitions and the same OS, you'll often run into booting problems, regardless of your boot order.

 

[boju]

Yes, Windows 10 tends to put the boot loader else where if multiple drives are connected. It's fine though but if were to remove one of those drives or it died, the boot loader goes with it, then yeah, Windows wont boot. Disconnecting all but destination drive when installing is good advice.

 

[xMecki]

I'd imagine there should be a dialog box yeah where you accept the terms of installation and stuff. User Account Control could be getting in the way if set too strong. Try turn it off.

DDU should be enough to wipe Nvidia drivers but using the custom clean install option should be good enough. DDU is really only necessary if changing brands, AMD to Nvidia etc.

Yeah I think it must have bugged out or something the first time as im getting diag boxes now when it was just skipping at first, just installed everything so ill most likely run prime95 fpr a quicker stress test to see if it restarts, in the past it took maybe 10mins for a restart

 

[xMecki]

When you are installing windows, do you have any other drives attached? If yes, that is likely your problem. Disconnect all the drives except the target drive and then load windows. Further, any time you have two or more drives with active partitions and the same OS, you'll often run into booting problems, regardless of your boot order.

I wish it was because of that but I only have my m.2 attached, I didnt need the other drives so I disconnected them a few days ago, so the reinstalls of windows have been with only my main drive connected, the boots happened quite a lot in safe mode recently when I was attempting to run checks on my driver and on windows using the sfc scannow cmd

 

[xMecki]

Yes, Windows 10 tends to put the boot loader else where if multiple drives are connected. It's fine though but if were to remove one of those drives or it died, the boot loader goes with it, then yeah, Windows wont boot. Disconnecting all but destination drive when installing is good advice.

Yeahh, so basically I just opened up Prime 95 while having core temps open and literally instantly the pc restarted and then getting straight back in to the sign in page within like 20sec and trying to open event viewer to see the logs it instantly crashed again, its a confusing one but im hoping its just a dodgy windows install, also out of curiosity theres no way it could be a virus right? Undetectable type which would lodge itself into my bios as ive already ran virus tests

 

[boju]

Possible but unlikely if you reinstalled Windows. But if Windows wasn't a fresh install it's possible i guess.

Have you tried testing Prime 95 with just 1 memory stick installed? Just in case it's something not even memtest is picking up. Think the first dimm is the 4th slot for 1 stick operation. Just to see what happens.

 

[onespeedbiker]

How did this start? Was it the install of the M.2 ssd? Have you tried using a different drive? Also, did you buy your RAM as a 4 stick kit?

 

[xMecki]

How did this start? Was it the install of the M.2 ssd? Have you tried using a different drive? Also, did you buy your RAM as a 4 stick kit?

Funnily enough it started when I was playing the new call of duty warzone, just a random restart and then thought nothing of it, tried to open the game up again and then restarted straight away. I had my m.2 for atleast 6months before this happened, gpu for a few months also, rest of the hardware i had for over a year and a half, no software updates or installs a week prior to the crash apart from the usual geforce experience driver updates and windows updates. So it was a pretty confusing one which is why i figured hey seems like somethings overheating or the powers being completely drained, checked windows logs, had kernal power logs at the time it was my first time using event viewer so i bought a new psu and the issue kept occuring, all in all the whole system is under 2 years old

 

[xMecki]

How did this start? Was it the install of the M.2 ssd? Have you tried using a different drive? Also, did you buy your RAM as a 4 stick kit?

Also forgot to mention but the ddr4 was bough in 2 packs of 2 but are both 3200MHz vengence pro sticks, ran dual channel at first then switched to using all 4

 

[xMecki]

Possible but unlikely if you reinstalled Windows. But if Windows wasn't a fresh install it's possible i guess.

Have you tried testing Prime 95 with just 1 memory stick installed? Just in case it's something not even memtest is picking up. Think the first dimm is the 4th slot for 1 stick operation. Just to see what happens.

With 1 its currently running prime, what length of time would you say is fair to run for? (Aslong as it doesnt crash within the usual time for myself) I may try running memtest86 again afterwards but like you said do 1 stick at a time

 

[boju]

Probably within the time it usually crashes, 10~15mins. Try each stick like you said then move on to two sticks.

 

[xMecki]

Probably within the time it usually crashes, 10~15mins. Try each stick like you said then move on to two sticks.

Can RAM affect cpu temps at all? First 40mins for the first ram stick was ok, now im hitting 100° after 15mins using 2 sticks

https://imgur.com/a/ASmiJLG

View: http://imgur.com/gallery/ASmiJLG

 

[xMecki]

Probably within the time it usually crashes, 10~15mins. Try each stick like you said then move on to two sticks.

I think my cooling fan wasnt actually turning on bc nzxt cam didnt start up gunna double check

 

[boju]

Yeah check cooling, ram can raise temps with the cpu's memory controller but not by much. What prime 95 version are you running?

Edited if didn't see.

 

[xMecki]

Yeah check cooling, ram can raise temps with the cpu's memory controller but not by much. What prime 95 version are you running?

Edited if didn't see.

Will current thermal paste be fine after hitting those temps or would you advice a change? Using mx-4 currently also prime95 i got from the extremeoverclocking site so the version there is 29.8 build 6 on 64 bit and as i say that the system just restarted at idle, will try switching the ram again

 

[boju]

Paste is fine. It's the later prime 95 versions above v26.6 using AVX instructions by default in the tests and that heats up cpu much more than necessary. There is a way to turn off AVX in newer versions or run v26.6.

Guy there explaining how to edit local.txt file to disable AVX.
View: https://www.reddit.com/r/intel/comments/adpen2/does_prime95_use_avx_by_default/

 

[xMecki]

Yeah check cooling, ram can raise temps with the cpu's memory controller but not by much. What prime 95 version are you running?

Edited if didn't see.

Also just tested 4/4 RAM sticks alone in the pc in 1 slot and all 4 times now using valley benchmark the systems crashed within 5mins, now im also getting a new error happening, the motherboard seems to be restarting over and over every second, shows a red led then green resets red then green resets and so on, the screen is pure black with the backlight on but you can see the screen flashing black, this is before the logo even comes on