How to Request a Virtual GPU

The ITS Private Cloud vGPU offering includes the technology of NVIDIA TESLA P6 TDP GPUs and is supported by NVIDIA Licensing, encompassing both the NVIDIA RTX Virtual Workstation and NVIDIA Virtual Applications.

To tailor to various needs, we have made available a range of vGPU profiles as shown below:

 Instructions to Add vGPU profile

This how-to provides the steps to request a Virtual Graphic Processing Unit via the command line interface (vss-cli).

1. Run vss-cli configure to configure your VSS credentials. If you have already done this, please go to next step.

2. Verify you have access to compute cluster FD6. If you do not have access, please reach out to vss@eis.utoronto.ca

   vss-cli compute domain get FD6

3. Make sure your Virtual Machine is running on FD6, if it is, proceed to next step. If it is not, please execute the following command:

   vss-cli --wait compute vm set {vm-id-or-name} domain FD6

4. Shut down or power off your virtual machine:

   vss-cli --wait compute vm set {vm-id-or-name} state shutdown

5. Add the required vGPU profile, which in this example will create a grid_p6-4a, which is a 4GB for virtual applications:

6. If the requested profile is available, the request will succeed and you can proceed to power the virtua machine on:

7. Once the virtual machine powers on, refer to https://eis-vss.atlassian.net/wiki/spaces/VSSPublic/pages/1442578460 to complete the GPU driver installation and licensing.

 Instructions to remove vGPU profile

1. Run vss-cli configure to configure your VSS credentials. If you have already done this, please go to next step.

2. Shut down or power off your virtual machine:

3. Remove the existing GPU profile as follows:

4. Power on your virtual machine:

 Related articles