Unable to Launch VM Due to "Missing" Nvidia GPU

I am trying to launch a Talos VM that I have running Plex and Fileflows and suddenly it will no longer boot, and I get the following error.

[EFAULT] internal error: qemu unexpectedly closed the monitor: 
2025-04-26T06:01:18.913532Z qemu-system-x86_64: -accel kvm: warning: Number of SMP cpus requested (240) exceeds the recommended cpus supported by KVM (72)
2025-04-26T06:01:18.913643Z qemu-system-x86_64: -accel kvm: warning: Number of hotpluggable cpus requested (240) exceeds the recommended cpus supported by KVM (72)
2025-04-26T06:05:19.849045Z qemu-system-x86_64: -device {"driver":"vfio-pci","host":"0000:af:00.1","id":"hostdev0","bus":"pci.0","addr":"0x8"}: VFIO_MAP_DMA failed: Bad address
2025-04-26T06:05:19.889534Z qemu-system-x86_64: -device {"driver":"vfio-pci","host":"0000:af:00.1","id":"hostdev0","bus":"pci.0","addr":"0x8"}: vfio 0000:af:00.1: failed to setup container for group 16: memory listener initialization failed: Region pc.ram: vfio_dma_map(0x557cb07561e0, 0x100000000, 0x3f40000000, 0x7efd2be00000) = -14 (Bad address)

When I edit the GPU section of the VM it says the address is [0000:af:00.0] but the error refers to [0000:af:00.1], so the address changed?

It seems as though it is similar to this issue: https://forums.truenas.com/t/docker-apps-and-uuid-issue-with-nvidia-gpu-after-upgrade-to-24-10/22547
So based on that, and other reports I saw in Discord and Reddit I ran the following command: midclt call app.gpu_choices | jq

The only thing it shows is the Matrox GPU, not the NVIDIA. So I’m not sure what’s going on

{
  "0000:03:00.0": {
    "vendor": null,
    "description": "Matrox Electronics Systems Ltd. Integrated Matrox G200eW3 Graphics Controller",
    "vendor_specific_config": {},
    "pci_slot": "0000:03:00.0"
  }
}

I removed the GPU and the VM wil boot, but it will constantly throw nvidia related errors until it reboots (roughly every 90-120min).

At this point I have no idea what’s going on. Is the card there or not? How do I remove the references to the nvidia card so if nothing else the VM will stop rebooting itself.