nvidia 驱动程序在我的 VMWare 虚拟机上运行良好，但突然在安装了一些 tensorflow 和 cuda 之后，驱动程序无法连接到 GPU（我有两个 GPU），因此我卸载了该驱动程序（nvidia-driver-460）并尝试了几乎所有其他可用的驱动程序来解决问题，尝试更新内核，甚至将系统从 Ubuntu 20.04 升级到 Ubuntu 22.04，但没有成功，所以我希望有人能帮助我发现并解决问题……
当我跑步时nvidia-smi我得到了这个：
NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.
以下是来自 Kernal 的一些诊断消息dmesg | grep nvidia：
[47841.841925] nvidia-nvlink: Unregistered Nvlink Core, major device number 239
[47842.323471] nvidia-nvlink: Nvlink Core is being initialized, major device number 239
[47842.325793] nvidia 0000:02:00.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=none
NVRM: specific graphics driver download page at www.nvidia.com.
[47842.326439] nvidia: probe of 0000:02:00.0 failed with error -1
[47842.327022] nvidia 0000:02:03.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=none
NVRM: specific graphics driver download page at www.nvidia.com.
[47842.327549] nvidia: probe of 0000:02:03.0 failed with error -1
[47842.327836] nvidia-nvlink: Unregistered Nvlink Core, major device number 239
[47842.851441] nvidia-nvlink: Nvlink Core is being initialized, major device number 239
[47842.853698] nvidia 0000:02:00.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=none
NVRM: specific graphics driver download page at www.nvidia.com.
[47842.854318] nvidia: probe of 0000:02:00.0 failed with error -1
[47842.854911] nvidia 0000:02:03.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=none
NVRM: specific graphics driver download page at www.nvidia.com.
[47842.855439] nvidia: probe of 0000:02:03.0 failed with error -1
[47842.855714] nvidia-nvlink: Unregistered Nvlink Core, major device number 239
我尝试安装其他版本的 nvidia 驱动程序，并尝试将系统更新到 ubuntu 22.04，还尝试将其他驱动程序列入黑名单，但都没有任何效果。