diff --git a/README.md b/README.md index 5dccf2e..7b0100c 100644 --- a/README.md +++ b/README.md @@ -32,7 +32,7 @@ You can follow this guide if you have a vGPU supported card from [this list](htt 15. [Adding a vGPU to a Proxmox VM](#adding-a-vgpu-to-a-proxmox-vm) 16. [Licensing](#licensing) 17. [PSA for Pascal (and older) GPUs like the P4, GTX 1080...](#psa-for-pascal-and-older-gpus-like-the-p4-gtx-1080) -18. [Pascal GPU Compatibility with Volta Spoofing]() +18. [Pascal GPU Compatibility with Volta Spoofing](#pascal-gpu-compatibility-with-volta-spoofing) 18. [Common problems](#common-problems) 19. [Support](#support) 20. [Feed my coffee addiction ☕](#feed-my-coffee-addiction-) @@ -663,22 +663,34 @@ Starting from driver version 17.0, nvidia in their infinite wisdom dropped suppo ### Pascal GPU Compatibility with Volta Spoofing -As of the latest updates, the current drivers are not compatible with Pascal GPUs. To ensure proper functionality, you need to spoof the Pascal GPU as a Volta GPU. This involves installing driver version 550.127.06 with the appropriate patch on the host system and modifying the virtual machine configuration file. For both Windows and Linux clients, driver version 550.127.05 is required. - -It is important to note that, at the time of writing, video encoding is not compatible with the 16.x drivers. +It is important to note that, at the time of writing, video encoding is not compatible with the 16.x drivers. To ensure proper functionality, you need to spoof the Pascal GPU as a Volta GPU. This involves installing driver version 550.127.06 with the appropriate patch on the host system and modifying the virtual machine configuration file. For both Windows and Linux clients, driver version 550.127.05 is required. To edit the configuration file, first replace `number_qm` with the actual number of your virtual machine in the file path `/etc/pve/qemu-server/number_qm.conf`. Then use the following command to open the file: ```bash -nano /etc/pve/qemu-server/number_qm.conf +nano /etc/pve/qemu-server/numero_qm.conf ``` -Add the following line to the configuration file: +For an NVIDIA Tesla P4, add the following line to the configuration file: ```bash hostpci0: 0000:03:00.0,device-id=0x1DB6,mdev=nvidia-63,sub-device-id=0x12BF,sub-vendor-id=0x10de,vendor-id=0x10de ``` +To identify the PCIe address for other models of the same generation, use the `lspci` command to list all PCI devices on your system: + +```bash +lspci +``` + +Look for the NVIDIA GPU in the list. The output will include the PCIe address (e.g., `03:00.0`) and the device ID. Ensure the device ID matches the model you are configuring. For the Tesla P4, the device ID is `1DB6`. Replace the PCIe address in the configuration line with the address identified. + +For example, if your GPU is listed at `04:00.0`, update the configuration line as follows: + +```bash +hostpci0: 0000:04:00.0,device-id=0x1DB6,mdev=nvidia-63,sub-device-id=0x12BF,sub-vendor-id=0x10de,vendor-id=0x10de +``` + ## Common problems Most problems can be solved by reading the instructions very carefully. For some very common problems, read here: