Question PC randomly shuts down while playing games

Aug 20, 2022
5
0
10
First of all, I apologize if this is in the wrong category or if a similar issue has been addressed before. The quick amount of Googling did not seem to get the results I needed.

To begin: I just recently got my hands on an RTX 3060Ti 8GB GPU and was excited to get to use it during my time off school. After installing it, I had noticed 4 short beeps sound from the motherboard upon startup (according to Google, the beeps relate to the motherboard CMOS battery). The computer continued to run fine so I did not pay much attention to it. About an hour in, while playing games while watching YouTube simultaneously, the computer forcefully restarts itself, without any indication of what happened. I have tried a couple of troubleshooting methods but without success. I tried to reseat the memory, reseat the GPU, and monitor the temperature. Reseating the components did not fix anything, and the temperature stays at around 60 degrees Celsius while doing my normal gaming, so I am not 100% sure on what the issue is.

I have a suspicion that I may need a new motherboard and CPU as these are two components that are extremely outdated. I got these two components from a friend and do not know their exact specifications, but what I do know is that both components are made by Intel, and the motherboard was made to fit the 2nd Gen Intel Core i7, which is what this PC currently uses. My PSU is about a year old and provides 750w of power, so I do not believe this would be the issue.

What do you guys think? Could it be an outdated motherboard/CPU? Could it be a faulty GPU? Or is there something else that I overlooked?
 
First of all, I apologize if this is in the wrong category or if a similar issue has been addressed before. The quick amount of Googling did not seem to get the results I needed.

To begin: I just recently got my hands on an RTX 3060Ti 8GB GPU and was excited to get to use it during my time off school. After installing it, I had noticed 4 short beeps sound from the motherboard upon startup (according to Google, the beeps relate to the motherboard CMOS battery). The computer continued to run fine so I did not pay much attention to it. About an hour in, while playing games while watching YouTube simultaneously, the computer forcefully restarts itself, without any indication of what happened. I have tried a couple of troubleshooting methods but without success. I tried to reseat the memory, reseat the GPU, and monitor the temperature. Reseating the components did not fix anything, and the temperature stays at around 60 degrees Celsius while doing my normal gaming, so I am not 100% sure on what the issue is.

I have a suspicion that I may need a new motherboard and CPU as these are two components that are extremely outdated. I got these two components from a friend and do not know their exact specifications, but what I do know is that both components are made by Intel, and the motherboard was made to fit the 2nd Gen Intel Core i7, which is what this PC currently uses. My PSU is about a year old and provides 750w of power, so I do not believe this would be the issue.

What do you guys think? Could it be an outdated motherboard/CPU? Could it be a faulty GPU? Or is there something else that I overlooked?
What psu you do have? could you list your pc full specs?
 
Aug 20, 2022
5
0
10
That's a hefty assumption, and generally incorrect, both in power of the psu and it not being an issue. This is one case where knowing the exact brand and model of psu is an absolute requirement.

CPU: Intel Core i7 2600 Sandy Bridge
GPU: Gigabyte GeForce RTX 3060Ti 8GB
Motherboard: Intel DH67CL Sandy Bridge Chipset
RAM: Intel-brand DDR3 4x4GB
PSU: EVGA Modular 750W

Wow I was totally unaware that I was still using DDR3 memory. That's a little bit outdated. I had also noticed that the motherboard's BIOS version dates back to 2018. I will look into getting that BIOS upgraded.
 
Last edited:

Karadjgne

Titan
Ambassador
CPU: Intel Core i7 2600 Sandy Bridge
GPU: Gigabyte GeForce RTX 3060Ti 8GB
Motherboard: Intel DH67CL Sandy Bridge Chipset
RAM: Intel-brand DDR3 4x4GB
PSU: EVGA Modular 750W

Wow I was totally unaware that I was still using DDR3 memory. That's a little bit outdated. I had also noticed that the motherboard's BIOS version dates back to 2018. I will look into getting that BIOS upgraded.

Neither is going to happen. It's Sandy Bridge, H67 chipset. The last actual bios update I believe was in 2013, but there were some actual chipset driver changes in 2017/2018 to address issues created by Microsoft when they revamped Windows10 into the Windows10 Creators Edition. That single update messed up everybody, including AMD and Nvidia, so fixes were issued to board partners for their chipset drivers. There hasn't been any updates since or bios updates because the mobo's are considered to old to be supported any longer.

That motherboard only fits DDR3. It cannot fit DDR4 or DDR5, which are physically different sizes and technologies. The only upgrade possible is to faster ram, which in DDR3 stopped at 2400MHz, but is of very little use and very expensive due to limited production. Best speed for 2nd/3rd gen Intel was 1866MHz Cas9 in 2 sticks or 1600MHz for 4 sticks with low timings like Cas7.

Still didn't address the psu. The only thing saying Modular means is it not an 80+ or White Evga as those were solid wires, not modular. So it's at least a Bronze unit, my guess would be a Evga B2 750w Bronze as those were relatively decent, especially for the budget price.

Your pc carries a power draw total of @ 350w, and even including a possible 200w power spike from the gpu, still puts you within normal protective boundaries for that psu. On one condition, group regulated psus like the B2 series are slower to respond (analog) than the newer DC-DC designs (digital), so it's entirely possible the gpu is pulling too much, too fast for the psu to compensate for and its short-changing the cpu voltages as they are on the same 12v rail.

If you think of it like boxing, your psu just threw 2 big, slow haymakers with a ton of power, but the gpu just smacked you in the face with 3x short, lightning fast jabs. What are the chances that 2nd haymaker actually landed. Pretty slim. Therein lies the issue, 2nd haymaker misses, cpu gets no power, bluescreen.
 
Aug 20, 2022
5
0
10
Neither is going to happen. It's Sandy Bridge, H67 chipset. The last actual bios update I believe was in 2013, but there were some actual chipset driver changes in 2017/2018 to address issues created by Microsoft when they revamped Windows10 into the Windows10 Creators Edition. That single update messed up everybody, including AMD and Nvidia, so fixes were issued to board partners for their chipset drivers. There hasn't been any updates since or bios updates because the mobo's are considered to old to be supported any longer.

That motherboard only fits DDR3. It cannot fit DDR4 or DDR5, which are physically different sizes and technologies. The only upgrade possible is to faster ram, which in DDR3 stopped at 2400MHz, but is of very little use and very expensive due to limited production. Best speed for 2nd/3rd gen Intel was 1866MHz Cas9 in 2 sticks or 1600MHz for 4 sticks with low timings like Cas7.

Still didn't address the psu. The only thing saying Modular means is it not an 80+ or White Evga as those were solid wires, not modular. So it's at least a Bronze unit, my guess would be a Evga B2 750w Bronze as those were relatively decent, especially for the budget price.

Your pc carries a power draw total of @ 350w, and even including a possible 200w power spike from the gpu, still puts you within normal protective boundaries for that psu. On one condition, group regulated psus like the B2 series are slower to respond (analog) than the newer DC-DC designs (digital), so it's entirely possible the gpu is pulling too much, too fast for the psu to compensate for and its short-changing the cpu voltages as they are on the same 12v rail.

If you think of it like boxing, your psu just threw 2 big, slow haymakers with a ton of power, but the gpu just smacked you in the face with 3x short, lightning fast jabs. What are the chances that 2nd haymaker actually landed. Pretty slim. Therein lies the issue, 2nd haymaker misses, cpu gets no power, bluescreen.

I apologize, I just now realized that I had actually replaced my PSU with a different one, the EVGA was my previous PSU before this one.

The PSU in my computer is the CORSAIR RMX Series, RM850x, 850 Watt, Fully Modular Power Supply, 80+ Gold. So it is actually 850w, not 750w as I had previously stated. This was purchased brand new in October last year.

I should also mention that there is no bluescreen of death when the issue arises. The computer just kills itself completely, then reboots after a couple of seconds. It is as if I had forcefully shut down the computer by holding the power button for 5 seconds.
 

Karadjgne

Titan
Ambassador
Oh. Well that's an excellent psu, shouldn't be any worries there, it's over-provisioned like crazy lol.

Cmos is a listing of all the stuff that's currently in the ram when you shutdown via windows or sleep or hibernate (should be disabled, it's a desktop not a laptop). Also contains all the previous bios settings and equipment addresses, so on boot the bios doesn't need to go searching for that stuff, like with a reset. You boot, cmos fills in everything, windows loads, quick and easy. That includes any time settings.

If there's a cmos battery issue, boot takes much longer as bios has to start from scratch. The system time will also be wrong as the battery didn't maintain the time. Cmos battery does not engage unless there's an absense of power, like if you pull the plug out. Pc's that are plugged in are not 'off' as such, they are in Standby mode basically, you'll still see motherboard lights, still have USB access etc.

I'd replace the battery. If you already got beeps, it's not going to hurt and it is several years old at this point.

That won't affect instant shutdowns. There's a conflict elsewhere that's doing that. Check the windows event viewer for any critical (red flagged) errors or any warnings (yellow flagged) that repeatedly show up.
 
Aug 20, 2022
5
0
10
Oh. Well that's an excellent psu, shouldn't be any worries there, it's over-provisioned like crazy lol.

Cmos is a listing of all the stuff that's currently in the ram when you shutdown via windows or sleep or hibernate (should be disabled, it's a desktop not a laptop). Also contains all the previous bios settings and equipment addresses, so on boot the bios doesn't need to go searching for that stuff, like with a reset. You boot, cmos fills in everything, windows loads, quick and easy. That includes any time settings.

If there's a cmos battery issue, boot takes much longer as bios has to start from scratch. The system time will also be wrong as the battery didn't maintain the time. Cmos battery does not engage unless there's an absense of power, like if you pull the plug out. Pc's that are plugged in are not 'off' as such, they are in Standby mode basically, you'll still see motherboard lights, still have USB access etc.

I'd replace the battery. If you already got beeps, it's not going to hurt and it is several years old at this point.

That won't affect instant shutdowns. There's a conflict elsewhere that's doing that. Check the windows event viewer for any critical (red flagged) errors or any warnings (yellow flagged) that repeatedly show up.
I thought about replacing the battery, but the beeps are not present if I put my old GTX 960 back in (or if there's no card at all). I did confirm that the system board was sounding 4 short beeps and not 3 short + 1 long beeps.

For the mean time, I do have some new parts on the way. I have an AMD Ryzen 5 5600x CPU, ROG Strix B550-A system board, and Corsair Vengeance LPX 32GB (2X16GB) DDR4 3200 ordered. Even if this does not fix the sudden shutdowns, I would rather have a system that is mostly up to date. I will keep updates as they come across!
 

Karadjgne

Titan
Ambassador
Backup any data you want or need to keep. With the new parts you'll need to reinstall windows completely, amd and Intel don't get along, and the info windows need to run the board correctly won't be discovered without a reinstallation.

Which will wipe out any existing data.

Also, disconnect any drives other than the C drive or you'll have issues later.
 
I thought about replacing the battery, but the beeps are not present if I put my old GTX 960 back in (or if there's no card at all). I did confirm that the system board was sounding 4 short beeps and not 3 short + 1 long beeps.

For the mean time, I do have some new parts on the way. I have an AMD Ryzen 5 5600x CPU, ROG Strix B550-A system board, and Corsair Vengeance LPX 32GB (2X16GB) DDR4 3200 ordered. Even if this does not fix the sudden shutdowns, I would rather have a system that is mostly up to date. I will keep updates as they come across!
some oem boards have bad compability with gpu. hence why gpu a wont work, but gpu b would work. IIRC the 3060Ti uses a UEFI GOP, that might be the problem.
 

Karadjgne

Titan
Ambassador
H67, that's iffy if it's Legacy or UEFI, but generally shouldn't be an issue, the biggest issue was with the hybrid boards, usually 3rd party OEM like HP and Dell, who tried implementing a UEFI substructure in a Legacy bios. It was a mess for a few years after the Maxwell cards and full UEFI vbios was implemented (gtx750/750ti). If the 960 worked, a 3060ti should too, vbios is the same at a basic level.
 
Aug 20, 2022
5
0
10
UPDATE: With my new motherboard, CPU, and RAM upgrades, my video card seems to no longer give me any issues. Haven't done any heavy testing yet, but I was able to use the computer as I normally do without any random shut off. If I had to guess, I think the processor in the previous board was not getting enough power since it only took a 4-pin connector to power it, whereas my new mobo requires an 8 or 10-pin connector.

I learned a valuable lesson: Make sure some of the core components of the computer are somewhat up to date before trying to throw in a video card like an RTX 3060.