4 separate systems BSOD with the exact same issue. Video TDR?

[aaa tech]

Hi. I recently built 4 crypto miners (ya ya I get it, people dont line miners, moving on) and all 4 of them are having an incredibly strange blue screen issue. I've built over 30 crypto miners in the past without issue so this isn't my first time building these. Will include the minidumps below. These are the specs:

• AMD 3000G CPU
• MSI B450 or ASRock B450 motherboard
• 8gb RAM
• 240gb SSD

The issue occurs every 10-60 minutes regardless of if the system is idle or mining. The GPUs are a mix of basically every 30 series NVIDIA card on the market, LHR and non-LHR. Drivers are the most up to date drivers. Aside for the GPUs, each system is identical. I've also used this exact combination of hardware for other systems without issue. I'm 90% sure its software. I tried troubleshooting myself and keep seeing Video TDR or something. I tried updating drivers, disabling fast startup, making it so the disk and the display do not turn off, having a local machine maintain a remote connection, and a couple other things to see if I can get it to stop. I can rule out the following without question:

• Lack of power. I am using HP Platinum 1200w server power supplies via parallelminer on 240v. I am WAY below power utilization and the issue occurs even when it is idle.
• Build error. Risers are being powered via ATX, GPUs show up no problem in device manager, Motherboard BIOS is up to date, And I've built this exact configuration a dozen times before so for it to happen across 4 systems is weird.
• Power delivery from the AC outlet to the systems. I moved these systems to another building to test that for sure. Same issue in 3 separate locations

For some reason Windows is no longer generating minidumps but they've been turned off for a month and the issue hasn't changed so I am attaching the minidumps from a month ago. 1 of the dumps is like 700mb and this site wont let me attach it so here's a we transfer link: https://we.tl/t-bzJUJmMGVo. I can grab some more of the older dumps if it helps. There were also 2 other systems this issue was happening to but they aren't hooked up right now, I anticipate they will have the same issue.

 

Thank you so much.

113021-6921-01.dmp 112921-18671-01.dmp

[aaa tech]

Also I can rule out thermals. Right now the room the machines are in is 10 degrees celsius and nothing is mining.

[RageTester]

I don't think you need windows at all to do mining... if you believe it's software why not try linux?

 

[aaa tech]

5 minutes ago, RageTester said:

I don't think you need windows at all to do mining... if you believe it's software why not try linux?

 

I know I can use linux but I'd prefer not to for logistical reasons. Regardless the systems are 7 hours away right now so installing a new operating system remotely is not possible.

[IkeaGnome]

26 minutes ago, aaa tech said:

Drivers are the most up to date drivers.

Are these your most recent racks? If so, what are the chances they have a newer driver on them that is causing the issues. Have you tried older drivers?

[aaa tech]

1 minute ago, IkeaGnome said:

Are these your most recent racks? If so, what are the chances they have a newer driver on them that is causing the issues. Have you tried older drivers?

I've tried drivers  462.31, 472.47, and 497.29

[aaa tech]

Stupid question but what if I just disable the TDR detection? The errors I'm seeing are saying its happening when TDR resets. I'm seeing on microsoft's website I can disable it in registry. Any1 know if that will cause any unforseen side effects?
 

https://docs.microsoft.com/en-us/windows-hardware/drivers/display/tdr-registry-keys