Question PC shuts down while gaming 1440p (possible hardware issue?) No overheating.

[BooNDocKk]

So my PC is shutting down after about 5-10 min while gaming in 1440p not in 1080p. There is no overheating that I'm seeing on any monitoring programs. Also boots and works properly browsing web and through windows. I have tried almost everything I know to do. Originally thought it was driver issue but now I'm thinking it is hardware related (PSU).
Things I've tried: Reinstalled windows 11. Then reinstalled windows 10 on a separate drive to see if it was 11 related. Updated all drivers for everything. completely wiped video drivers with DDU. All Bios settings are default. unplugged every unnecessary accessory and RGB controller. I've disassembled my PC completely looking for anything that would be an issue. No loose connections . AIO looks good. Pump is working with good thermal paste spread. Pins on motherboard and CPU look fine. Really my only option now is to throw parts at it. I'm thinking PSU. unfortunately all I have here is a 650 watt PSU and its not sufficient for my system. So ive ordered a new one of those. But just wondering if you guys have any ideas or any advice for me on this issue. Thanks.

Specs:
Intel core i7 12700k
NZXT Kraken x53 240mm aio
Asus Strix Z690 - Gaming Wifi
Corsair Vengeance 32Gb DDR5 5200
MSI Gaming trio rtx 3080
EVGA 850watt GA PSU
Samsung 980 1Tb SSD
samsung 970 1Tb SSD
case: Corsair 5000D

 

[DSzymborski]

So my PC is shutting down after about 5-10 min while gaming in 1440p not in 1080p. There is no overheating that I'm seeing on any monitoring programs. Also boots and works properly browsing web and through windows. I have tried almost everything I know to do. Originally thought it was driver issue but now I'm thinking it is hardware related (PSU).
Things I've tried: Reinstalled windows 11. Then reinstalled windows 10 on a separate drive to see if it was 11 related. Updated all drivers for everything. completely wiped video drivers with DDU. All Bios settings are default. unplugged every unnecessary accessory and RGB controller. I've disassembled my PC completely looking for anything that would be an issue. No loose connections . AIO looks good. Pump is working with good thermal paste spread. Pins on motherboard and CPU look fine. Really my only option now is to throw parts at it. I'm thinking PSU. unfortunately all I have here is a 650 watt PSU and its not sufficient for my system. So ive ordered a new one of those. But just wondering if you guys have any ideas or any advice for me on this issue. Thanks.

Specs:
Intel core i7 12700k
NZXT Kraken x53 240mm aio
Asus Strix Z690 - Gaming Wifi
Corsair Vengeance 32Gb DDR5 5200
MSI Gaming trio rtx 3080
EVGA 850watt GA PSU
Samsung 980 1Tb SSD
samsung 970 1Tb SSD
case: Corsair 5000D

PSU is definitely the first place to start with this GPU. After EVGA stopped having Super Flower make PSUs because of cost, they saved money with Andyson, hec, and FSP and the GA and GQ PSUs use a cheaper, less advanced topology that has issues with the transient loads of high-end 30 series cards. If going with EVGA, it's best to either go for a G6 (made by SeaSonic) or go somewhere else like a Corsair RMx. This has been a common problem with 3080s and 3090s and your symptoms -- crashing suddenly when the GPU needs to suddenly do more robust things -- are consistent with this behavior.

 

[BooNDocKk]

That makes sense. I used this PSU from a previous build. Lesson learned never cheap out on a power supply. Thanks btw. I didnt know this. EVGA just lost a customer.

 

[DSzymborski]

That makes sense. I used this PSU from a previous build. Lesson learned never cheap out on a power supply. Thanks btw. I didnt know this. EVGA just lost a customer.

Just to play devil's advocate, the 30 series were still a ways off when EVGA changed partners, so they likely didn't expect to have these kinds of issues. Active clamp reset forward is cheaper to implement and tends to be noisier because of hard switching and less efficient, but they would have been OK with the power spikes of most GPUs until that series came out. Even SeaSonic had some surprising problems with some of their older Focus PSUs. Not that I like those EVGA PSUs, but it turned out worse than they reasonably expected at the time.

As such, whenever someone's building with a recent, power-hungry GPU, we always recommend the top-tier PSUs only. Those transients make it much harder to skate by with a good budget PSU. Like you wouldn't ideally go back five years and run a 1080 Ti on a Corsair CX 650W, but it would probably be fine.

 

[BooNDocKk]

Fair enough. Thanks for the knowledgeable info. Have a corsair RM1000x in route. I dont have an actual PSU plug in tester. But what do you think about the software OCCT PSU tester? With my circumstances would it be safe to run that?

 

[DSzymborski]

PSU testing isn't that useful without specific equipment.

It should be fine to use. Just make sure to not reuse any modular cables!

 

[Flayed]

Fair enough. Thanks for the knowledgeable info. Have a corsair RM1000x in route. I dont have an actual PSU plug in tester. But what do you think about the software OCCT PSU tester? With my circumstances would it be safe to run that?

If you already have a new PSU coming, it may not matter but you can test to see if this is the issue by using MSI Afterburner to reduce the power level to your graphics card and see if the shutdown still happens.