PC Crashing Under Load/Games

[bigmaxnonions]

I have been having an issue with my custom built PC recently. My PC will freeze/sometimes BSOD during gaming. It only seems to happen during more demanding games, like Destiny 2 or larger parks in Planet Coaster. However, it runs fine when playing games like League of Legends (sad) and for everyday usage, like internet browsing and Microsoft Office.

 

My PC Specs are:

CPU: 5800X

Mobo: Asus ROG B550i

Ram: G.Skill Trident Z Neo DDR4 2x16GB 3600Mhz CL16

GPU: Asus TUF 3080

PSU: NZXT E850

SSD: Samsung 970 EVO+

Crucial MX500 SATA

 

When it crashes, it either will freeze and make a buzzing noise, which requires holding down the power button, or it will BSOD. When trying to turn it back on, it will cycle power and shut itself on and off, and will only turn on when the PSU is turned off and let sit for 10-15 seconds. Then, it will boot into safe mode, and seemingly work fine, until I try and run a demanding game again. The Event Viewer says it is a Kernel Power 41 Error (63). I have run Hard Disk Sentinel to check the drives, and I have run MemTest86 and have determined it is not the RAM. I can still boot and use the PC, so I don't think it is the CPU or GPU. I have updated the BIOS and have freshly reinstalled Windows 10 and all necessary drivers, so I don't believe it is software based. I did try and run OCCT, and it froze and I had to manually shut down the PC, so that is what is making me think it is the PSU. What is interesting is that the RGB I have in the case is moving, but when the computer freezes, it stops moving as well. It is an 850W PSU, and I was going to try swapping the connections for the GPU to see if that does anything. I swapped one of the GPU connections on my PSU and tried to run OCCT, and it immediately froze and BSOD, giving me SYSTEM\_SERVICE\_EXCEPTION and that cldflt.sys failed. Event Viewer says Kernel Power 43 again.

 

All this is what is leading me to believe that it is the PSU. I am looking for a second opinion because the holiday season is coming up and I don't want to spend more money than I need to, and this is going to be an unexpected purchase if I need to make it. I was planning on purchasing a Seasonic GX-850 or stepping up to a Corsair 1000 Watt PSU. 

 

Thanks for any help!

 

Edit: Temps are in the 60-70C range for CPU and 50-60C for GPU. 

I have tried a new power supply and it does not seem to have solved my issue. OCCT still crashes during the load test. 

[BobVonBob]

That PSU has a 10 year warranty. Have you considered warranty replacement instead of buying a new PSU?

[bigmaxnonions]

21 minutes ago, BobVonBob said:

That PSU has a 10 year warranty. Have you considered warranty replacement instead of buying a new PSU?

I wanted to make sure it was the PSU first, I’ve seen a number of people on LTT Forums and Reddit that have similar issues. 

[Applefreak]

48 minutes ago, bigmaxnonions said:

Hello! I think I've identified my PSU as the problem, looking for some second opinions.

 

I have been having an issue with my custom built PC recently. My PC will freeze/sometimes BSOD during gaming. It only seems to happen during more demanding games, like Destiny 2 or larger parks in Planet Coaster. However, it runs fine when playing games like League of Legends (sad) and for everyday usage, like internet browsing and Microsoft Office.

 

My PC Specs are:

CPU: 5800X

Mobo: Asus ROG B550i

Ram: G.Skill Trident Z Neo DDR4 2x16GB 3600Mhz CL16

GPU: Asus TUF 3080

PSU: NZXT E850

SSD: Samsung 970 EVO+

Crucial MX500 SATA

 

When it crashes, it either will freeze and make a buzzing noise, which requires holding down the power button, or it will BSOD. When trying to turn it back on, it will cycle power and shut itself on and off, and will only turn on when the PSU is turned off and let sit for 10-15 seconds. Then, it will boot into safe mode, and seemingly work fine, until I try and run a demanding game again. The Event Viewer says it is a Kernel Power 43 Error (63). I have run Hard Disk Sentinel to check the drives, and I have run MemTest86 and have determined it is not the RAM. I can still boot and use the PC, so I don't think it is the CPU or GPU. I have updated the BIOS and have freshly reinstalled Windows 10 and all necessary drivers, so I don't believe it is software based. I did try and run OCCT, and it froze and I had to manually shut down the PC, so that is what is making me think it is the PSU. What is interesting is that the RGB I have in the case is moving, but when the computer freezes, it stops moving as well. It is an 850W PSU, and I was going to try swapping the connections for the GPU to see if that does anything. I swapped one of the GPU connections on my PSU and tried to run OCCT, and it immediately froze and BSOD, giving me SYSTEM\_SERVICE\_EXCEPTION and that cldflt.sys failed. Event Viewer says Kernel Power 43 again.

 

All this is what is leading me to believe that it is the PSU. I am looking for a second opinion because the holiday season is coming up and I don't want to spend more money than I need to, and this is going to be an unexpected purchase if I need to make it. I was planning on purchasing a Seasonic GX-850 or stepping up to a Corsair 1000 Watt PSU. 

 

Thanks for any help!

 

Edit: Temps are in the 60-70C range for CPU and 50-60C for GPU. 

Before replacing the PSU, DDU your graphics driver and remove all RGB and other stuff that runs in the background except for AV and system tools. Also try running the games on a single stick of RAM at a time. Usually PSU issues occur under load the same way, not just in some games unless certain games. It is possible that PSU has an issue but check other things first. The windows error code means that the system suddenly lost power as a result safe mode and power cycling on reboot is not unheard of. Has the error been occurring recently or has it been ongoing for a while? If the problematic games are on a certain SSD, try replacing it's cable and avoid bending it.

[bigmaxnonions]

17 minutes ago, Applefreak said:

Before replacing the PSU, DDU your graphics driver and remove all RGB and other stuff that runs in the background except for AV and system tools. Also try running the games on a single stick of RAM at a time. Usually PSU issues occur under load the same way, not just in some games unless certain games. It is possible that PSU has an issue but check other things first. The windows error code means that the system suddenly lost power as a result safe mode and power cycling on reboot is not unheard of. Has the error been occurring recently or has it been ongoing for a while? If the problematic games are on a certain SSD, try replacing it's cable and avoid bending it.

I have tried running the games in question with a single stick and it still crashes. I’ve tested both sticks in memtest86, and both passed after 4 runs. The problem has only started occurring in the past 2-3 weeks, and had not happened before that. The only games I’ve tested that have crashed are destiny 2, Planet Coaster, and Overwatch 2. League of legends runs just fine. The games are all installed on an NVME, so no cables to bend. The drive registers at 98% health, but seems to run a little warm at around 50C. The only RGB I have is my water block, running through Aura Sync, and my RAM, running through whatever G.Skills RGB software is called. I will try DDUing the drivers and reinstalling them, but I wiped and reinstalled windows 10 and reinstalled all drivers after doing so, so I don’t think that will necessarily help. 

[Applefreak]

2 minutes ago, bigmaxnonions said:

I have tried running the games in question with a single stick and it still crashes. I’ve tested both sticks in memtest86, and both passed after 4 runs. The problem has only started occurring in the past 2-3 weeks, and had not happened before that. The only games I’ve tested that have crashed are destiny 2, Planet Coaster, and Overwatch 2. League of legends runs just fine. The games are all installed on an NVME, so no cables to bend. The drive registers at 98% health, but seems to run a little warm at around 50C. The only RGB I have is my water block, running through Aura Sync, and my RAM, running through whatever G.Skills RGB software is called.

Try DDU the drivers as well but it is quite possible that you may need a different PSU. Maybe opt for a 1000W instead. Do you have another PSU to test with? you probably would need a weaker GPU as well as the 3080 will need too much power for an older unit and crash as well. 

Probably not relevant on a newer system but make sure your PSU is fairly dust free and get's enough air from underneath your case. 

[bigmaxnonions]

4 hours ago, Applefreak said:

Try DDU the drivers as well but it is quite possible that you may need a different PSU. Maybe opt for a 1000W instead. Do you have another PSU to test with? you probably would need a weaker GPU as well as the 3080 will need too much power for an older unit and crash as well. 

Probably not relevant on a newer system but make sure your PSU is fairly dust free and get's enough air from underneath your case. 

Just got home from work and am going to try to DDU my drivers. Should I try upgrading to Windows 11 for shits and giggles lmao?

[bigmaxnonions]

DDUed drivers and reinstalled, updated some new AMD drivers as well, ram OCCT and after 2:30 mins, system froze again. Temps were 77C on CPU and 50C on GPU, so I don’t think it was a heat issue. Gonna pick up a 1000W power supply Sunday and RMA my current one. 

[bigmaxnonions]

So I picked up a Corsair RM1000X and installed it into my system and I am still having the same issues. OCCT will run for about 2-3 mins and then my computer will BSOD. Its been flashing so quickly I can't read what it says. When my computer restarts, the QLED for the RAM is always on, and takes a few restarts to make it go away. I have tested both sticks of RAM and they both say they have no issues. Would it be something to do with my motherboard? I am starting to run out of ideas for this.