Question PC reboots randomly without warning im out of ideas please help

Mar 28, 2020
47
0
30
0
Hardware is at the bottom and the images shown are kinda backwards so start from the last image to the first. Would appreciate anyone and everyone's help! thanks in advance.

Ok, so my issue is that my pc has been randomly restarting without any on screen warning and I honestly dont know why, I have done so many things to try and figure this out and none of them has worked, I will leave below what ive done followed by the hardware i have.

So, this started a week ago and happened for the first time while playing call of duty so at the time I thought nothing of it, got back on tried to load up cod and it happened again, so straight away I went to valley benchmark to check for overheating and power consumption, didnt really see anything wrong but just to point out I am a rookie and dont know what many numbers and words mean in apps like HWInfo or windows logs. So i went on a forumn and asked a few people for help which they then told me it was a power supply issue, after checking the windows logs it did show up as a kernal power critical but was told afterwards that this might not mean it is the issue, but before that I bought a new power supply, another RM850x and funnily enough the issue continues. Just to point out, I completely stripped the pc bare before getting my new PSU, I also applied new thermal paste to the cpu and applied new thermal paste to the gpu, no change.

I used to have extension cables for my PSU and now have been switched to default cabled, i have a 2080ti so it needs 2 8pin cables and a 6pin cable, i used to run 1 6+2 for the 2 8pins and the same again for the last but now I am running 3 6+2pin cables incase it was power related and maybe a cable issue im not sure but nothing changed. Finally to the point where I made progress, I uninstalled my nvidia drivers using DDU and my system was running games again but now it restarts around 4hours into playing games, usually between 2-4hours on average or if left on just the normal screen it lasts longer. I havent tried replacing my gpu or ram or cpu, same with the motherboard because i dont have the money. I have also reinstalled windows, tried a virus scanner, tried updating the bios and other drivers from the asus website, ran ccleaner to get rid of broken files and registry's, still no change as of yet. I have been checking windows logs but i have no clue what most of it says as there is so much information which is where i was hoping someone in the pc field could step in, I have attached a link which will be after this message View: https://imgur.com/gallery/fGZgDO4
it shows around 13 screenshots from windows logs which hopefully gives no personal info away but around an hour ago I had 3 back to back restarts within 10minutes, they were around the 11.30pm mark, so hopefully that does help, but as said before im honestly out of ideas and im hoping nothing needs to be replaced and its just an issue i can sort out in windows.

PS sometimes when the system restarts the gpu's fans speed up which is why i feel it could be a gpu issue, also i tried removing the ssd and hdd with no
change. I have also tried the memory diagnostic tool in windows and it came back with no errors in the logs.

Hardware:
1TB M.2 samsung evo ssd
500gb ssd
1tb hdd
cooling - NZXT x72 kraken
7 case fans in total
psu - RM850x 80w gold
32gb RAM - vengence pro
motherboard - asus rog Maximus xi hero (wifi)
gpu - rtx 2080 ti gaming x trio
cpu - i9 9900k 3.6GHz
hopefully thats all the needed hardware but just to add the motherboard has a digit display which shows any errors and its working normal so id assume its not a motherboard issue.
 
what all-core clock speeds and CPU temps are hit on a sample CPU-Z/bench/stress CPU run....?

Is MCE enabled on your board? (You'd know it was if all core clock speeds of 5 GHz are seen/maintained)

IF MCE were enabled, perhaps, you could tune down the clock speeds by 100-200 MHz, potentially avoiding any VRM thermal events, and/or, try blowing air across/onto the VRM circuitry

Is your GPU overclocked at all? (Try underclocking it 20-40 MHz on core/VRAM speeds)

Ultimately, a decent number of 2080/2080Ti's have failed out there, there was a rash of bad ones upon initial deployment...

If it's not your PSU, and not a RAM issue (try two different sticks, or, lower your XMP speeds to a more conservative 2666 MHz etc), then you can either test your GPU elsewhere, or, you'd need another one. (If your GPU can be tested elsewhere and acts up similarly there, that might help in the RMA process if it is eligible..
 
Mar 28, 2020
47
0
30
0
what all-core clock speeds and CPU temps are hit on a sample CPU-Z/bench/stress CPU run....?

Is MCE enabled on your board? (You'd know it was if all core clock speeds of 5 GHz are seen/maintained)

IF MCE were enabled, perhaps, you could tune down the clock speeds by 100-200 MHz, potentially avoiding any VRM thermal events, and/or, try blowing air across/onto the VRM circuitry

Is your GPU overclocked at all? (Try underclocking it 20-40 MHz on core/VRAM speeds)

Ultimately, a decent number of 2080/2080Ti's have failed out there, there was a rash of bad ones upon initial deployment...

If it's not your PSU, and not a RAM issue (try two different sticks, or, lower your XMP speeds to a more conservative 2666 MHz etc), then you can either test your GPU elsewhere, or, you'd need another one. (If your GPU can be tested elsewhere and acts up similarly there, that might help in the RMA process if it is eligible..
Apologies I was supposed to add HWInfo info with the pics but forgot to, ill add these below, but just to say on your other questions, the bios isnt overclocked, ive reset it and even updated the bios so that everything should be back to normal, the RAM also is 3200MHz so I should try bringing that down to see if the issue persists? also the gpu overclocks were taken off a few days ago but the problem still consists there, also im not sure what MRE is in all honesty but ill do a quick google search to have a look, but the speeds do still show up as 4700MHz for some reason instead of what id assume i should be getting being 3600 with it being a 3.6GHz cpu? so maybe it is on but hopefully the HWinfo pics saw if it is or not. View: https://imgur.com/gallery/zSJnmQ9
also just for the pics, all pics but the last were taken at the same time, the last pic was taken at a different time but all are when running valley benchmark. appreciate the help! im not too sure where mce is on the board but ive got food on which i need to see too so i will check once i get the chance.
 
In the last pic above, CPU usage is quite low, yet, 4700 MHz clocks are shown across all cores....; are you in Performance mode for a power plan, or, is there a fixed 47X multiplier specified for manual mode in the BIOS?

Again , absolutely ruling out a potential RAM issue, I'd test with one stick at time, or, slap in a known good RAM kit, or, run reduced clocks, or combinations of the above; WIndows memory test might pass when you run it, meaning it was good in that time frame, but, could have issues. Most generally accept Memtest86 run from a USB flashdrive for a few passes over a few hours to be a generally accepted test. Additionally, running Prime95 in default blended mode also tests memory pretty thoroughly.... Your temps looked 'ok' but, there was not really a load. Can you run Prime95/small FFT (custom, disable AVX/AVX2), and not HWMonitor temps/stability after 10 minutes

Additionally, what core voltage is shown in the top part of HWMonitor? Vcore, and VID 0 - 7?

4700 MHz is the all-core turbo default, whereas 3600 MHz would be a minimum base clock used by some boards to keep power consumption down....; naturally, most prefer 4700 MHz as a reasonable 'default', even if it does perhaps bust the 95 W TDP...
 
Mar 28, 2020
47
0
30
0
In the last pic above, CPU usage is quite low, yet, 4700 MHz clocks are shown across all cores....; are you in Performance mode for a power plan, or, is there a fixed 47X multiplier specified for manual mode in the BIOS?

Again , absolutely ruling out a potential RAM issue, I'd test with one stick at time, or, slap in a known good RAM kit, or, run reduced clocks, or combinations of the above; WIndows memory test might pass when you run it, meaning it was good in that time frame, but, could have issues. Most generally accept Memtest86 run from a USB flashdrive for a few passes over a few hours to be a generally accepted test. Additionally, running Prime95 in default blended mode also tests memory pretty thoroughly.... Your temps looked 'ok' but, there was not really a load. Can you run Prime95/small FFT (custom, disable AVX/AVX2), and not HWMonitor temps/stability after 10 minutes

Additionally, what core voltage is shown in the top part of HWMonitor? Vcore, and VID 0 - 7?

4700 MHz is the all-core turbo default, whereas 3600 MHz would be a minimum base clock used by some boards to keep power consumption down....; naturally, most prefer 4700 MHz as a reasonable 'default', even if it does perhaps bust the 95 W TDP...
in HWmonitor the Vcore is averaging 0.8V idle, max its hit is 1.2V and lowest 0.6V but its changing pretty frequently like going from 0.7 to .8 .9 1v back to .8 .7 etc also ill try doing what youve said now
 
Mar 28, 2020
47
0
30
0
In the last pic above, CPU usage is quite low, yet, 4700 MHz clocks are shown across all cores....; are you in Performance mode for a power plan, or, is there a fixed 47X multiplier specified for manual mode in the BIOS?

Again , absolutely ruling out a potential RAM issue, I'd test with one stick at time, or, slap in a known good RAM kit, or, run reduced clocks, or combinations of the above; WIndows memory test might pass when you run it, meaning it was good in that time frame, but, could have issues. Most generally accept Memtest86 run from a USB flashdrive for a few passes over a few hours to be a generally accepted test. Additionally, running Prime95 in default blended mode also tests memory pretty thoroughly.... Your temps looked 'ok' but, there was not really a load. Can you run Prime95/small FFT (custom, disable AVX/AVX2), and not HWMonitor temps/stability after 10 minutes

Additionally, what core voltage is shown in the top part of HWMonitor? Vcore, and VID 0 - 7?

4700 MHz is the all-core turbo default, whereas 3600 MHz would be a minimum base clock used by some boards to keep power consumption down....; naturally, most prefer 4700 MHz as a reasonable 'default', even if it does perhaps bust the 95 W TDP...
also for power plan im running with ultimate performance , ill give the bios a check and get back to you from my phone
 
Mar 28, 2020
47
0
30
0
In the last pic above, CPU usage is quite low, yet, 4700 MHz clocks are shown across all cores....; are you in Performance mode for a power plan, or, is there a fixed 47X multiplier specified for manual mode in the BIOS?

Again , absolutely ruling out a potential RAM issue, I'd test with one stick at time, or, slap in a known good RAM kit, or, run reduced clocks, or combinations of the above; WIndows memory test might pass when you run it, meaning it was good in that time frame, but, could have issues. Most generally accept Memtest86 run from a USB flashdrive for a few passes over a few hours to be a generally accepted test. Additionally, running Prime95 in default blended mode also tests memory pretty thoroughly.... Your temps looked 'ok' but, there was not really a load. Can you run Prime95/small FFT (custom, disable AVX/AVX2), and not HWMonitor temps/stability after 10 minutes

Additionally, what core voltage is shown in the top part of HWMonitor? Vcore, and VID 0 - 7?

4700 MHz is the all-core turbo default, whereas 3600 MHz would be a minimum base clock used by some boards to keep power consumption down....; naturally, most prefer 4700 MHz as a reasonable 'default', even if it does perhaps bust the 95 W TDP...
These are my bios settings and values on the side, assuming that the x47 could be bc everythings on auto and it could be defaulted? Dont have anything on manual as far as im aware tho Bios View: https://imgur.com/gallery/HMOyvpW
 

ASK THE COMMUNITY