PC Rebooting itself for Some reason during gaming, and I can't figure out why

[Fuzzyy]

So, this issue has been happening over the last 8 months, and it is very annyoing. I cannot play a competitive game with my friends anymore... This issue become more frequent when I upgraded my Ram and SSD 8 months ago. (before that it happened like 3-4 times in whole 2 years of its lifetime)

 

Problem: When I am playing a game, everything gets frozen, and my pc reboots(can't see a blue screen or anytihng). This happens especially in Overwatch 2, where I can't even play a single game, it just crashes in 5-10 mins after launching the game, and in Apex Legends and CSGO it crashes in between 30mins to an hour(also experienced this on It Takes Two). For non-hardware intensive games(Cs:Source, l4d2) or just light web browsing/using windows, I have never experienced this problem. The temps are also in control, for CPU it is reaching max 75C and for GPU 67C during hardware intensive games. 

 

For troubleshooting, I have tried:

• Restoring every BIOS setting to default, disabled DOCP
• Updating to the latest BIOS
• Reinstalled Windows 11 many times
• Reinstalled Windows on a different SSD
• Updated Graphic Drivers
• Uninstalled any RGB software(turns out this is a thing, according to Bitwit)

 

Below are my PC Specs:

• Ryzen 5 5600X
• Asus B550-I Rog Strix
• 2x16GB 32GB Corsair Vengance LPX Ram
• EVGA XC3 Ultra 3070
• Corsair SF600 Platinum
• Samsung 980 PRO 1TB
• Noctua NH-L12S
• Case: FormD T1

 

My findings so far:

• I belive this issue can be caused by 3 components: Ram, SSD or GPU
• For RAM, I have disabled DOCP, and it did not work. I don't know how to make sure it is caused by RAM or not
• For the SSD, I have tried reinstalling Windows 11 on my older SSD(WD SN750) and it was good at first, but the issue started happening again a week later. I made sure that my new SSD(980 PRO) is getting enough cooling(I have a top down cooler, so it is getting direct airflow, so I don't think SSD is overheating).
• For the GPU, I have an intresting story: I lent it to my friend a year ago, and apperantly he tripped over the HDMI cable, breaking the HDMI port(no video outpuit anymore...), and this is only physical damage it took. The port does not look broken from the outside, but it does not work.
• I moved/shaked my case around to see if something loose was causing the problem, nothing happened.

 

Any idea what the cause is, or how can I find/fix it? Sorry for the long post, and thanks in advance.

[Dukesilver27-]

1 hour ago, Fuzzyy said:

• EVGA XC3 Ultra 3070
• Corsair SF600 Platinum

I would at least have a 650W non SFF PSU for a 3070, with all the transient spikes, going with SFF means you need higher capacity than needed. Besides, 600W is barely the minimum requirement for 3070, Nvidia suggest a 650W PSU.

And the fact that it happens during gaming, it only make sense that the PSU just gets tripped by the transient spikes of the GPU.

[PcBeExpensive]

1 hour ago, Fuzzyy said:

So, this issue has been happening over the last 8 months, and it is very annyoing. I cannot play a competitive game with my friends anymore... This issue become more frequent when I upgraded my Ram and SSD 8 months ago. (before that it happened like 3-4 times in whole 2 years of its lifetime)

 

Problem: When I am playing a game, everything gets frozen, and my pc reboots(can't see a blue screen or anytihng). This happens especially in Overwatch 2, where I can't even play a single game, it just crashes in 5-10 mins after launching the game, and in Apex Legends and CSGO it crashes in between 30mins to an hour(also experienced this on It Takes Two). For non-hardware intensive games(Cs:Source, l4d2) or just light web browsing/using windows, I have never experienced this problem. The temps are also in control, for CPU it is reaching max 75C and for GPU 67C during hardware intensive games. 

 

For troubleshooting, I have tried:

• Restoring every BIOS setting to default, disabled DOCP
• Updating to the latest BIOS
• Reinstalled Windows 11 many times
• Reinstalled Windows on a different SSD
• Updated Graphic Drivers
• Uninstalled any RGB software(turns out this is a thing, according to Bitwit)

 

Below are my PC Specs:

• Ryzen 5 5600X
• Asus B550-I Rog Strix
• 2x16GB 32GB Corsair Vengance LPX Ram
• EVGA XC3 Ultra 3070
• Corsair SF600 Platinum
• Samsung 980 PRO 1TB
• Noctua NH-L12S
• Case: FormD T1

 

My findings so far:

• I belive this issue can be caused by 3 components: Ram, SSD or GPU
• For RAM, I have disabled DOCP, and it did not work. I don't know how to make sure it is caused by RAM or not
• For the SSD, I have tried reinstalling Windows 11 on my older SSD(WD SN750) and it was good at first, but the issue started happening again a week later. I made sure that my new SSD(980 PRO) is getting enough cooling(I have a top down cooler, so it is getting direct airflow, so I don't think SSD is overheating).
• For the GPU, I have an intresting story: I lent it to my friend a year ago, and apperantly he tripped over the HDMI cable, breaking the HDMI port(no video outpuit anymore...), and this is only physical damage it took. The port does not look broken from the outside, but it does not work.
• I moved/shaked my case around to see if something loose was causing the problem, nothing happened.

 

Any idea what the cause is, or how can I find/fix it? Sorry for the long post, and thanks in advance.

As said above, try undervolting the card to try to prevent higher power spikes and see if that helps. Try deleting GPU drivers with DDU and re-installing. Try one ram stick in each slot.

[Fuzzyy]

8 hours ago, Dukesilver27- said:

I would at least have a 650W non SFF PSU for a 3070, with all the transient spikes, going with SFF means you need higher capacity than needed. Besides, 600W is barely the minimum requirement for 3070, Nvidia suggest a 650W PSU.

And the fact that it happens during gaming, it only make sense that the PSU just gets tripped by the transient spikes of the GPU.

 

8 hours ago, PcBeExpensive said:

As said above, try undervolting the card to try to prevent higher power spikes and see if that helps. Try deleting GPU drivers with DDU and re-installing. Try one ram stick in each slot.

I highly doubt it is because of the PSU, some people even ran 3080 with this PSU with a more power hungry cpu. And I have used the same system for over 2 years, and did not have an issue at all before. I also run my card with an undervolt as well, but it did not work. Here is a video of optimum tech that also pushed me to get the SF600:

8 hours ago, PcBeExpensive said:

one ram stick in each slot.

I haven't tried this before, so will try it this weekend definitely.

[Dukesilver27-]

Well, based on the troubleshooting you have done, it points to hardware issue. And the most obvious one seems to be the PSU.

The only way to be sure is to swap out each part to pinpoint the issue, and if it were me, I'd first try the PSU.

[Fuzzyy]

6 minutes ago, Dukesilver27- said:

Well, based on the troubleshooting you have done, it points to hardware issue. And the most obvious one seems to be the PSU.

The only way to be sure is to swap out each part to pinpoint the issue, and if it were me, I'd first try the PSU.

Thanks, will also try with another PSU as well

[PcBeExpensive]

1 hour ago, Fuzzyy said:

 

I highly doubt it is because of the PSU, some people even ran 3080 with this PSU with a more power hungry cpu. And I have used the same system for over 2 years, and did not have an issue at all before. I also run my card with an undervolt as well, but it did not work. Here is a video of optimum tech that also pushed me to get the SF600:

 

It could be luck or something but the 30 series cards are known for there power spikes.

[Fuzzyy]

2 hours ago, Dukesilver27- said:

Well, based on the troubleshooting you have done, it points to hardware issue. And the most obvious one seems to be the PSU.

The only way to be sure is to swap out each part to pinpoint the issue, and if it were me, I'd first try the PSU.

 

46 minutes ago, PcBeExpensive said:

It could be luck or something but the 30 series cards are known for there power spikes.

I want to mention one last thing, today I have ran Furmark on QHD preset, and I have maxed out the GPU power limit in afterburner(112%), and my 3070 pulled a little over 270 watts. I have also run the CPU burner in the background with 11 threads(PBO enabled), and it pulled another 100 watts, and my system did not shut down. I ran the tests for an hour. After this, would you still says a low wattage PSU could cause this?

[Fuzzyy]

Also, I do not have another higher wattage PSU at the moment, but I have a RTX 3060 in my htpc, so I will try testing with a 3060, and if it does not shutdown, I guess it is either my GPU or PSU. 

[Dukesilver27-]

The potential main issue here is not the total power draw, it's the transient spikes, sudden increase in power draw that could trip the OCP feature of a PSU, which would shutdown the PC instantly.

This usually happens during gaming or something that uses both CPU and GPU intensively, not synthetic benchmarks.

Hardware also degrades overtime, something could fail out of nowhere, just because it has been fine for 2 years, does not mean it would've been fine for 10 years right?

[Fuzzyy]

19 hours ago, Dukesilver27- said:

The potential main issue here is not the total power draw, it's the transient spikes, sudden increase in power draw that could trip the OCP feature of a PSU, which would shutdown the PC instantly.

This usually happens during gaming or something that uses both CPU and GPU intensively, not synthetic benchmarks.

Hardware also degrades overtime, something could fail out of nowhere, just because it has been fine for 2 years, does not mean it would've been fine for 10 years right?

Yes I agree with you. Did some more testing, ordered a 750w psu from amazon(cm gold v750 sfx) and the issue still persist, I started a return request for the PSU. But I think it might be because of my riser cable as well, I have the Louqe Cobalt riser, and I have read some comments that a faulty one can lead to random pc restarts, and it is a very common thing. Will do a open test bench setup without the riser tomorrow, and hopefully it works

[silencer12]

@Fuzzyy

 

 

New bios update released on August 28. One day ago. ROG STRIX B550-I GAMING BIOS 3301

 

1. Recommended for vital update to mitigate the potential security vulnerabilities
2. Improve system stability
 

[Dukesilver27-]

If PSU is fine, next would be motherboard. You're very lucky to be living in a country where returns are as easy as it is.

[T487]

You can test the RAM with MemTest86.

 

 

[Fuzzyy]

14 hours ago, Dukesilver27- said:

If PSU is fine, next would be motherboard. You're very lucky to be living in a country where returns are as easy as it is.

Lets say thanks to Amazon for making my 3rd world country better lol 

[Fuzzyy]

Today I did an open testbench build on top of my Motherboard box, and now I don't get any reboots or something, I think it was really because of the PCIe riser, maybe I have a faulty one.

[Dukesilver27-]

Well, certainly is possible, I haven't had bad experience with PCIE riser so I can't say for certain.