GraptikBlizzard

Prominent
Jul 3, 2021
32
1
535
So for the last 6 months after building my PC, I've been experiencing random shutdowns and reboots (with no BSOD) while at idle or opening programs. Never happens with gaming or stress test. Well, I just figured out the issue and it is extremely weird.

I noticed that I can replicate the shutdowns sometimes by opening the Oculus app. It doesn't shut down the PC every time, but sometimes it will. I noticed my PC (not the speakers, the actual inside of the PC) also makes a small but noticeable buzzing sound every time I open the Oculus app. Stay with me here, because it gets weirder.

I reset my BIOS to default settings, and it turns out, it only crashes and makes this sound when the core ratio setting is changed from "Auto" to anything else. Doesn't matter if I put it on 4.8GHz all core (which is what my CPU runs at stock). As long as that single setting is changed, it will crash eventually opening the app and makes a short buzzing sound every time that app is opened. I can have my RAM overclocked to the ceiling and it will be fine, but change that setting from Auto and all bets are off.

I have no idea what to do about this. I'm assuming it's the motherboard? I'm probably going to order a new one off of Amazon to test it. Do you guys think it could be the CPU? The system is rock solid stable at 4.9GHz and will run Prime95 for hours along with every other stress test, but when opening certain programs with that settings changed from auto, it's an instant reboot.

Specs:
Motherboard: MSI MAG Tomahawk Z490
CPU: i9 10900K
RAM: 32GB @3900MHz
GPU: 2080ti EVGA FTW3 Ultra
Cooler: Noctua NH-D15S
 
Solution
Yeah, I'm suspicious that perhaps it IS simply a board issue. The new board will tell us a lot and if it isn't the problem, we can move on from there. For what it's worth, at first, try using ONLY a single DIMM, installed in the A2 slot and no other, and if the problem still happens like that then it's clearly not the board.

Also, I suppose it should go without saying that you MIGHT need to do another clean install when you install the new board IF it is a different model than what you currently have. You may or may not know it, but Intel has stopped offering non-manufacturer specific chipset drivers since the last few generations, so all chipset drivers for Intel platforms tend to be motherboard specific these days, even if the...
Have you checked to see if you have the MOST up to date stable motherboard BIOS version?

What version IS installed?

What is your EXACT power supply model number? Yes, it might be relevant.

Did you do a full CLEAN install of Windows, removing all existing partitions on the drive and installing to the unpartitioned space? And if you used a new drive to install Windows on, did you make sure to remove all, even the hidden partitions used by the system, on the old drive, if it is also attached?

Have you installed the MOST recent drivers for the chipset, audio and network adapters (LAN and WiFi, if applicable) from the motherboard product page? Graphics card drivers, preferably using the DDU?
 
I checked the health of it and it is apparently at 82%.
I'd be somewhat concerned about this, but not too much. It's probably due to TBW and not actual drive condition. Is this the same drive Windows is on? If so, then clearly a clean install wasn't done when you built the new system and THAT is probably your biggest issue, unless by "programs" you are talking about game files. In general, most applications need to be installed on the same drive the OS is installed on, and when they are not and you install Windows on a new drive, you break all the dynamic links. Perhaps a bit more information is necessary in trying to resolve this issue.
 

GraptikBlizzard

Prominent
Jul 3, 2021
32
1
535
Have you checked to see if you have the MOST up to date stable motherboard BIOS version?

What version IS installed?

What is your EXACT power supply model number? Yes, it might be relevant.

Did you do a full CLEAN install of Windows, removing all existing partitions on the drive and installing to the unpartitioned space? And if you used a new drive to install Windows on, did you make sure to remove all, even the hidden partitions used by the system, on the old drive, if it is also attached?

Have you installed the MOST recent drivers for the chipset, audio and network adapters (LAN and WiFi, if applicable) from the motherboard product page? Graphics card drivers, preferably using the DDU?
Motherboard BIOS is updated to the latest version (7C80v1A).
Power Supply is a brand new RM750x, I have tried three different power supplies hoping to fix this issue.
Windows was clean installed and all of my drives were completely wiped including all partitions in the installation disk manager.
Used DDU to uninstall Nvidia drivers and then clean installed them and also installed the chipset drivers from the MSI website.
 

GraptikBlizzard

Prominent
Jul 3, 2021
32
1
535
I'd be somewhat concerned about this, but not too much. It's probably due to TBW and not actual drive condition. Is this the same drive Windows is on? If so, then clearly a clean install wasn't done when you built the new system and THAT is probably your biggest issue, unless by "programs" you are talking about game files. In general, most applications need to be installed on the same drive the OS is installed on, and when they are not and you install Windows on a new drive, you break all the dynamic links. Perhaps a bit more information is necessary in trying to resolve this issue.
I did a clean install of Windows when I first built the PC and have done so a couple times since, each time reinstalling all of my programs and apps onto my SSD (same SSD Windows is on). The 82% figure is just from a program that checks SSD health with regards to how many read/write cycles it has left.
 
Windows was clean installed and all of my drives were completely wiped including all partitions in the installation disk manager.
The Oculus (along with all of my other programs) app is on my SSD that I've had for a couple years. I checked the health of it and it is apparently at 82%.
These would seem to be contradictory statements, unless of course I'm misunderstanding and you ONLY mean that you've had the SSD for a couple of years, but that the programs and OS currently on it were all recently reinstalled on there.
 

GraptikBlizzard

Prominent
Jul 3, 2021
32
1
535
Ok, so this problem has been happening since day one of assembling this system?
Yeah, right from the start. It's taken me a while to narrow the cause down because it was only happening about once a week until I realised that certain programs cause it to happen more often. I also only just figured out that it only happens when the core ratio setting is changed from anything other than "Auto", even if it's to a lower clock speed than stock.
 
In this case, since you've already done the majority of the basic troubleshooting steps I'd recommend, I would HIGHLY recommend that you remove the CPU cooler, remove the CPU and check for bent pins on the motherboard. Check VERY closely. Even just a SINGLE bent pin, which can be very hard to see and often requires the use of a magnifying glass or high powered reading glasses, can cause ANY problem you might encounter on any system. As can a power supply issue, but I think you've probably already adequately determined that the PSU is not to blame as that RM750x is plenty for that system so long as there is no significant overclocking of the CPU or graphics card happening AND so long as the other two power supplies you tried are of equivalent or similar quality, condition and capacity.

Make sure you first have a tube of thermal paste to replace what's already on there and some 90% or higher isopropyl alcohol to clean the mating surfaces of the CPU heat spreader and heatsink base before reapplying TIM and reinstalling.

Also, what is the EXACT model of your memory kit, how many DIMMs are in the kit and which slots are they installed in, exactly. Yes, it can certainly make a difference.

Also, while I understand you've reinstalled applications and Windows, you didn't address the question regarding whether you'd gone to the product page for your motherboard and downloaded, and installed, the latest drivers for your chipset, networking and audio, all of which NEED to be done. Do not rely on any automatic driver utilities, ever, and don't rely on Windows native drivers, which are only basic universal type drivers in most cases and don't provide the latest or most fully featured drivers for most motherboards.
 

GraptikBlizzard

Prominent
Jul 3, 2021
32
1
535
In this case, since you've already done the majority of the basic troubleshooting steps I'd recommend, I would HIGHLY recommend that you remove the CPU cooler, remove the CPU and check for bent pins on the motherboard. Check VERY closely. Even just a SINGLE bent pin, which can be very hard to see and often requires the use of a magnifying glass or high powered reading glasses, can cause ANY problem you might encounter on any system. As can a power supply issue, but I think you've probably already adequately determined that the PSU is not to blame as that RM750x is plenty for that system so long as there is no significant overclocking of the CPU or graphics card happening AND so long as the other two power supplies you tried are of equivalent or similar quality, condition and capacity.

Make sure you first have a tube of thermal paste to replace what's already on there and some 90% or higher isopropyl alcohol to clean the mating surfaces of the CPU heat spreader and heatsink base before reapplying TIM and reinstalling.

Also, what is the EXACT model of your memory kit, how many DIMMs are in the kit and which slots are they installed in, exactly. Yes, it can certainly make a difference.
I have a new motherboard arriving in a few hours for me to test and see if it fixes the issue, while I'm installing it I will check my current motherboard for any bent pins. My memory is: Corsair Vengeance CMK16GX4M2B3000C15.
I have 4 sticks of it but the issue persists no matter what configuration I have them in, one stick, two sticks, four sticks. Overclocked or at 2133MHz. The only factor that changes whether it shuts down or not is the core ratio setting.
 
And when you change the core ratio, what else are you also changing along with it?

And, for what purpose are you CHANGING the core ratio? I think we need a little more information on this, as normally the core ratio is only changed when overclocking. And if you are overclocking but not also making OTHER significant changes, to things like core voltage, among other settings, then the problem is likely due to instability. Are you making changes in the BIOS or are you using a desktop app like XTU to make these changes?

There is always, albeit very slim, possibility of a bad CPU. That would be extremely rare though without some kind of user abuse like being dropped, etc., being involved. CPUs rarely fail and even more rarely COME to you failed, unless somebody had it before you.
 

GraptikBlizzard

Prominent
Jul 3, 2021
32
1
535
And when you change the core ratio, what else are you also changing along with it?

And, for what purpose are you CHANGING the core ratio? I think we need a little more information on this, as normally the core ratio is only changed when overclocking. And if you are overclocking but not also making OTHER significant changes, to things like core voltage, among other settings, then the problem is likely due to instability. Are you making changes in the BIOS or are you using a desktop app like XTU to make these changes?
I am changing the core ratio to 4.9GHz all core along with the voltage. The 10900k has an all core boost of 4.9GHz anyway, I just change the core ratio to make sure it runs at that all the time. However, as I said before, it doesn't matter what I change the clock to, I could change it to 4.5GHz (300MHz slower than stock) and still the issue persists. It seems it only matters that the setting is changed, not what it is changed to. 4.9GHz is rock solid and can run Prime95 for hours and hours along with any other stress test. Even if I supply it with way more voltage than it needs to be completely stable, the issue still persists. I just cloned my C: drive to my brand new NVME drive to see if the drive was failing but nope, still the exact same. My new motherboard should arrive within the hour so that will be my next test.
 
Yeah, I'm suspicious that perhaps it IS simply a board issue. The new board will tell us a lot and if it isn't the problem, we can move on from there. For what it's worth, at first, try using ONLY a single DIMM, installed in the A2 slot and no other, and if the problem still happens like that then it's clearly not the board.

Also, I suppose it should go without saying that you MIGHT need to do another clean install when you install the new board IF it is a different model than what you currently have. You may or may not know it, but Intel has stopped offering non-manufacturer specific chipset drivers since the last few generations, so all chipset drivers for Intel platforms tend to be motherboard specific these days, even if the majority of it is the same behind the curtains there are board specific driver changes from brand to brand and even in some cases from board to board, as well as for sure a need for different drivers based on the specific network adapters and audio chipsets that are installed. So, I'd probably recommend starting from scratch IF at first you don't see the problem resolved or if DIFFERENT problems surface.
 
Solution

GraptikBlizzard

Prominent
Jul 3, 2021
32
1
535
Yeah, I'm suspicious that perhaps it IS simply a board issue. The new board will tell us a lot and if it isn't the problem, we can move on from there. For what it's worth, at first, try using ONLY a single DIMM, installed in the A2 slot and no other, and if the problem still happens like that then it's clearly not the board.

Also, I suppose it should go without saying that you MIGHT need to do another clean install when you install the new board IF it is a different model than what you currently have. You may or may not know it, but Intel has stopped offering non-manufacturer specific chipset drivers since the last few generations, so all chipset drivers for Intel platforms tend to be motherboard specific these days, even if the majority of it is the same behind the curtains there are board specific driver changes from brand to brand and even in some cases from board to board, as well as for sure a need for different drivers based on the specific network adapters and audio chipsets that are installed. So, I'd probably recommend starting from scratch IF at first you don't see the problem resolved or if DIFFERENT problems surface.
I don't want to speak too soon but I think the Motherboard swap seems to have solved it. I am also getting much better temperatures (10-20c lower). I am running at 5GHz now and there is no strange sound when I open the Oculus app and I haven't experienced a shutdown/reboot yet.
I find it really peculiar that the other Motherboard shuts down if you simply changed the status of one setting on it, what an extremely odd issue. I also wonder what that buzzing sound was, I hope it wasn't anything dangerous for my CPU.
 
This wouldn't be terribly unusual, and, unfortunately, the Tomahawk boards mostly don't have the same great record overall as the B450 Tomahawk boards, in general. So, hard to say, but if the board swap fixes the problem, that's pretty hard to argue with.

In general, and you can ask anyone here, I'm just not a fan of MSI at all. They have a very SHORT record of things getting better, but a very long record of things being crap in the past, and in the present, and in general.

And then when you add in the behavior of a company like MSI, it makes it really hard to try and care about them as a company.

And if you are not familiar with that, well, might watch this anyhow.

 

GraptikBlizzard

Prominent
Jul 3, 2021
32
1
535
This wouldn't be terribly unusual, and, unfortunately, the Tomahawk boards mostly don't have the same great record overall as the B450 Tomahawk boards, in general. So, hard to say, but if the board swap fixes the problem, that's pretty hard to argue with.

In general, and you can ask anyone here, I'm just not a fan of MSI at all. They have a very SHORT record of things getting better, but a very long record of things being crap in the past, and in the present, and in general.

And then when you add in the behavior of a company like MSI, it makes it really hard to try and care about them as a company.

And if you are not familiar with that, well, might watch this anyhow.

Coincidentally my friend just had his MSI Z370-a Pro motherboard die on him a couple days ago. Looks like I'll be staying away from MSI boards in the future. For this new board I went with Gigabyte since I've personally had many good experiences with them in the past. Looks like this new board definitely fixed the issue. Running at 5GHz and 3600MHz RAM and not had any shutdowns or reboots and there is no strange sound when opening Oculus. I wonder what it was about the Oculus app that seemed to exacerbate the problem? Either way, thanks for your help on this issue, it is much appreciated.
 

Zerk2012

Titan
Ambassador
Coincidentally my friend just had his MSI Z370-a Pro motherboard die on him a couple days ago. Looks like I'll be staying away from MSI boards in the future. For this new board I went with Gigabyte since I've personally had many good experiences with them in the past. Looks like this new board definitely fixed the issue. Running at 5GHz and 3600MHz RAM and not had any shutdowns or reboots and there is no strange sound when opening Oculus. I wonder what it was about the Oculus app that seemed to exacerbate the problem? Either way, thanks for your help on this issue, it is much appreciated.
Not surprised on the Z370 A board they had poor Vram that can be a lack of research on his part.
Buy the cheapest Z boards made and you get cheap parts.

View: https://m.youtube.com/watch?v=cZcqw-K1drY