Chia_Recompute_Server crashing on nvdxgdmal64.dll when using OpenCL

[AtomicInternet2]

I have two 5700G CPUs serving my Gigahorse farmer each with a chia_recompute_server to access the OpenCL iGPU resources.
One of these machines also has an nVidia 1030 serving an official Chia client.

The 5700G with no nVidia GPU has a recompute server working fine.
The 5700G with the 1030 has a recompute server that runs for a few seconds and then crashes with an nVidia driver error.

I previously used CHIAPOS_VISIBLE_DEVICES=5 to force the recompute to ignore the CUDA card entirely, and have recently added the CHIAPOS_MAX_CUDA_DEVICES=0 but it still crashes on nVidia driver issues, despite telling it to only look at OpenCL.

The command line crash is below:

C:\Status>SET CHIAPOS_MAX_CUDA_DEVICES=0

C:\Status>SET CUDA_VISIBLE_DEVICES=5

C:\Status>chia_recompute_server.exe
[chiapos] Using 16 / 16 CPU threads
[chiapos] Using 0 / 0 CUDA devices
[chiapos] Using 1 / 1 OpenCL devices (AMD Accelerated Parallel Processing)
Using 16 / 16 CPU threads
Using gfx90c (AMD Accelerated Parallel Processing)
ReComputeServer listening on 0.0.0.0 port 11989

C:\Status>

The Event Viewer error is below:

Faulting application name: chia_recompute_server.exe, version: 0.0.0.0, time stamp: 0x6677c753
Faulting module name: nvdxgdmal64.dll_unloaded, version: 0.0.0.0, time stamp: 0x665ba50a
Exception code: 0xc0000005
Fault offset: 0x0000000000008080
Faulting process id: 0x0x268C
Faulting application start time: 0x0x1DAC630BA0E46F8
Faulting application path: C:\Status\chia_recompute_server.exe
Faulting module path: nvdxgdmal64.dll
Report Id: 150f027b-4a72-48c4-a6c1-3f7362854be2
Faulting package full name:
Faulting package-relative application ID:

[AtomicInternet2]

Figured out my system drive was dying, so that file was actually corrupt. I've restored from backup and am running with no issues, so closing this.