Computer reboots randomly

[scipio_aemilius]

Hi everyone, I have a problem and I have no idea where to start to fix it.

Recently, I upgraded my graphic card (from a SAPPHIRE Pulse Radeon RX 560 to a NVIDIA GeForce RTX 2080). I also upgraded my power supply from a 750W to a 1000W, just to be on the safe side. And while I was at it, I added a second HDD of 20TB. My CPU (Intel Core i7-3770 - 3.40GHz) can probably take it. My motherboard is very old (10+ years). I'm also kind of a noob when it come to plugging stuff in my desktop.

Sadly, ever since then, I've experienced random reboots, only while playing games. The reboot seems like the computer is unplugged, and replugged almost immediately. No warning, no error message, just a plain reboot. Time to reboot is also random. I can play the same game and it will crash after 30 minutes before a reboot, or I can play a few hours without any problem. It seems totally random.

The games I play when my computer crash are not recent : Age of Empires 3 Complete (not even the Definitive version!) and Civilization 5, solo or multiplayer, and they are installed on my older 2TB HDD. I can also play other games, more recent games, installed on my new HDD. Unoptimized games like Life is feudal : Forest Village or Prison Architect will work my computer up much more than the previous two games, but will not cause a reboot.

At first I though maybe my computer overheated, but then I ran some 3DMark tests, and even the stress test for like 30 minutes, and nothing crashed, nothing rebooted, everything was fine and dandy.

Now, I am at a loss! Is my older HDD the culprit? Is my power supply not enough? Is my graphic card plugged incorrectly?

I need your advices, Ô sages of Steam!

If you need more info, or you want me to run some tests, I'll happily oblige!

And thank you for chipping in!

[_I_]

reboot is most likely psu or unstable oc

disable windows auto reboot on bsod
and check windows logs to see the errors was before the reboot

[scipio_aemilius]

By unstable oc, you mean overclock? I did not mess with anything other than plugging the hardware in, so no overclocking. And the PSU is totally new, I mean, I can happen out of the box, but man, that would suck...

Anyway, I'll try your Windows logs advice next time my computer reboots!

[_I_]

post a cpuz validation link
http://www.cpuid.com/softwares/cpu-z.html
cpuz -> validate button -> submit button
it will open a browser, copy the url (address) and paste it here

and psu brand/model?

[emoticorpse]

With all those hardware changes, I'd do a fresh Windows install to see if that helps anything. Sure there's a chance, it's still possible that it still randomly crashes, but you can at least know it's most likely hardware at that point and try a different avenue of troubleshooting.

If you really really don't like to format/re-install Windows though, then I suppose wait on re-installing.

Can you in run the Userbenchmark program also thing and post a link to your results? If you don't know what that is, google it and go to the website and download the free program then run it.

[scipio_aemilius]

Here's the CPUZ Validation page :
https://valid.x86.fr/65qglq

My new PSU is a Corsair RMe Series RM1000e

As for UserBenchmark, it rebooted my computer while testing the GPU. I just wanted to post it here before trying again.

[emoticorpse]

Originally posted by scipio_aemilius:

Here's the CPUZ Validation page :
https://valid.x86.fr/65qglq

My new PSU is a Corsair RMe Series RM1000e

As for UserBenchmark, it rebooted my computer while testing the GPU. I just wanted to post it here before trying again.

Keep trying. If it keeps rebooting the pc, that's good news. At least you have a reliable way to reproduce the problem.

[_I_]

core voltage looks low
should be closer to 1.2v, not under 1v

amd ram is cl11, kinda high for 1600(800mhz)
check the spd tab and see what the 1600 profile says cl should be
in bios, enable xmp and pick the 1600 profile

[scipio_aemilius]

I installed UserBenchmark on my newer HDD, and the test went through. Although I question some results... Here it is : https://www.userbenchmark.com/UserRun/67265026

Originally posted by _I_:

core voltage looks low
should be closer to 1.2v, not under 1v

amd ram is cl11, kinda high for 1600(800mhz)
check the spd tab and see what the 1600 profile says cl should be
in bios, enable xmp and pick the 1600 profile

I must admit, I didn't understand much of this answer... You need to tell me like I'm 5 years old, sadly.

[scipio_aemilius]

Originally posted by emoticorpse:

Keep trying. If it keeps rebooting the pc, that's good news. At least you have a reliable way to reproduce the problem.

Sadly, it's not that easy... I did the UserBenchmark test from my older HDD where it first crashed, and my desktop didn't reboot this time, and here's the result I got : https://www.userbenchmark.com/UserRun/67265072

[Phénomènes Mystiques]

That RAM says voltage should be 1.5v not CPU voltage RAM voltage.

CPU voltage should be at 1.2(5)v (Stock) towards 1.35v (Safe Max).

[_I_]

reset bios to defaults, and enable xmp

[Sigma957]

Did something shake loose? You might be shorting your motherboard.

[Illusion of Progress]

Is the PC restarting without power ever actually being cut? Or is power being cut momentarily (all fans, lights, etc. off, and then back on?) and being restored? I ask because that's potentially an important distinction.

If it's the former (PC restarting but power never actually gets cut), this sounds like the issue I had changing my video card (all other hardware remained the same), and it was the PC (CPU specifically) catching a machine check exception condition and signaling a restart as a result. PC hardware, CPUs (and RAM?) at least, have machine check architecture that can do this. The machine check exception, as they almost always are, was caused by a hardware fault; in my case the video card was bad. Did an RMA and it resolved it.

If power is actually being cut, it suggests more of a power/electrical issue (which could be PSU, motherboard, GPU, or even RAM, rarely CPU, but it might just be an improperly set voltage somewhere), and you did change the PSU. The new one is probably better and a higher wattage, yes, but new hardware can be faulty.

On the other hand, while it was working fine with no issues before any of these changes (?), you do have a very aged platform (CPU/motherboard) and that's always going to be a candidate. But for now, presume those aren't the issue if you didn't have issues before the recent changes and narrow it down and start at looking at the new stuff.

First things first, I'd reseat all connections, at least graphics cards, RAM, and power and data cables. You would have swapped those when changing the PSU (I hope if it's modular, you didn't reuse the cables for your old PSU) so all of those connections should be reseat to rule them out. Shouldn't need to bother reseating CPU or CPU cooling though.

If that doesn't stop it, check event viewer. There's likely to be clues here, BSOD or not.

Also, check for WHEA logs. Same story, and this is what helped me narrow my hunt down too.

Windows/LiveKernelReports/WHEA

Windows/LiveKernelReports/WATHCDOG

If those directories have logs that correspond to the time of these issues, either upload them somewhere, or download WinDbg and analyze them yourself and/or post results here.

If you want a better way to stress test, I'd also recommend OCCT. The free version has a 30 minute limit on tests (and a 10 second wait before each one) but it has a whole suite of tests you can do on demand to try and narrow things down.

Since you swapped three parts, you could narrow this down by minimizing the changes. The new HDD can be left out for now. The old PSU is likely enough for an RTX 2080. Using the new PSU and old GPU is also a possibility to rule out the PSU, but the old GPU might not pull enough power to expose it (though it sounds like this is tripped so easily in your case that maybe it would still happen even then if it was the PSU causing it).

[scipio_aemilius]

Originally posted by smallcat:

So , your old hdd could be the culprit

It came to my mind as well, but the more I think about it, the less likely it sounds. I mean, it worked quite well until I made all the changes, and it didn't reboot at all previously, unless I asked for a reboot...

Originally posted by Sigma957:

Did something shake loose? You might be shorting your motherboard.

It was another possible deduction. Maybe I didn't plug things correctly. Especially the motherboard. I didn't want to bend or break it, but I tried plugging the 24 pin cable to the best of my ability. I also could have mixed cables when plugging things up, especially the 4 pin cable to the motherboard, which I changed because I knew I didn't get the right one at first (I did use the ones that came in the box, though).

I doubt things shake loose, though. My desktop is on a stable desk, with nothing coming close to it. I'm also usually pretty calm when I play, so I don't tend to rock my desk much. But hey, maybe! I'll take another look.

Originally posted by _I_:

reset bios to defaults, and enable xmp

Will do!

Originally posted by Wynters:

That RAM says voltage should be 1.5v not CPU voltage RAM voltage.

CPU voltage should be at 1.2(5)v (Stock) towards 1.35v (Safe Max).

I guess I'll look at it when firing up my bios?

Originally posted by Illusion of Progress:

Is the PC restarting without power ever actually being cut? Or is power being cut momentarily (all fans, lights, etc. off, and then back on?) and being restored? I ask because that's potentially an important distinction.
...
If power is actually being cut, it suggests more of a power/electrical issue (which could be PSU, motherboard, GPU, or even RAM, rarely CPU, but it might just be an improperly set voltage somewhere), and you did change the PSU. The new one is probably better and a higher wattage, yes, but new hardware can be faulty.

Power is actually being cut, just like powering down my machine.

Originally posted by Illusion of Progress:

On the other hand, while it was working fine with no issues before any of these changes (?), you do have a very aged platform (CPU/motherboard) and that's always going to be a candidate. But for now, presume those aren't the issue if you didn't have issues before the recent changes and narrow it down and start at looking at the new stuff.

First things first, I'd reseat all connections, at least graphics cards, RAM, and power and data cables. You would have swapped those when changing the PSU (I hope if it's modular, you didn't reuse the cables for your old PSU) so all of those connections should be reseat to rule them out. Shouldn't need to bother reseating CPU or CPU cooling though.

Seems logical. Yes I did use the new PSU cables, even the electric cord.

Originally posted by Illusion of Progress:

If that doesn't stop it, check event viewer. There's likely to be clues here, BSOD or not.

Also, check for WHEA logs. Same story, and this is what helped me narrow my hunt down too.

Windows/LiveKernelReports/WHEA

Windows/LiveKernelReports/WATHCDOG

If those directories have logs that correspond to the time of these issues, either upload them somewhere, or download WinDbg and analyze them yourself and/or post results here.

There is a WATCHDOG report, but not a WHEA report. I'll post the watchdog here as soon as I can open it.

I did check Even Viewer, and there are a series of system critical Kernel-Power that seems to align with when my computer rebooted in the last two months.

Here's the one that happened yesterday :
- System
- Provider
[ Name] Microsoft-Windows-Kernel-Power
[ Guid] {331c3b3a-2005-44c2-ac5e-77220c37d6b4}
EventID 41
Version 8
Level 1
Task 63
Opcode 0
Keywords 0x8000400000000002
- TimeCreated
[ SystemTime] 2024-02-14T00:18:21.9945155Z
EventRecordID 521494
Correlation
- Execution
[ ProcessID] 4
[ ThreadID] 8
Channel System
Computer SPQR
- Security
[ UserID] S-1-5-18

- EventData
BugcheckCode 0
BugcheckParameter1 0x0
BugcheckParameter2 0x0
BugcheckParameter3 0x0
BugcheckParameter4 0x0
SleepInProgress 0
PowerButtonTimestamp 0
BootAppStatus 0
Checkpoint 0
ConnectedStandbyInProgress false
SystemSleepTransitionsToOn 0
CsEntryScenarioInstanceId 0
BugcheckInfoFromEFI false
CheckpointStatus 0
CsEntryScenarioInstanceIdV2 0
LongPowerButtonPressDetected false

Originally posted by Illusion of Progress:

If you want a better way to stress test, I'd also recommend OCCT. The free version has a 30 minute limit on tests (and a 10 second wait before each one) but it has a whole suite of tests you can do on demand to try and narrow things down.

Since you swapped three parts, you could narrow this down by minimizing the changes. The new HDD can be left out for now. The old PSU is likely enough for an RTX 2080. Using the new PSU and old GPU is also a possibility to rule out the PSU, but the old GPU might not pull enough power to expose it (though it sounds like this is tripped so easily in your case that maybe it would still happen even then if it was the PSU causing it).

I don't know about OCCT, but I did 3DMark tests, even a 30 mins stress test, and my computer managed to go through without much difficulty.

I'll try swapping my GPU if nothing else seems to work.

Again, thank you all for your help!

[Phénomènes Mystiques]

One time Windows Update set the feature "Turn off Hard Drive After..." to 5 minutes so the system shut off while paused in game, listening to music, or watching a movie after said 5 minutes.