Appendix B: GPU Resource Allocation
This section covers how VM’s vGPUs will be allocated across the available physical GPUs in a host.
vGPU Placement Policy
Integrated GPU-pass-through
Which vGPU Profiles are allowed on physical GPUs
Select the host server in the left resources pane. Navigate to the GPU tab. All physical GPU and vGPU resources available will display.
Select Edit under placement policy to modify GPU settings.
The options of maximum density or maximum performance can be viewed here for assigning the VMs to GPUs. The option to select an integrated GPU will also be displayed. Select the appropriate options for your deployment strategy.
In the right panel, select Edit Selected GPUs to see the resources allocated to each GPU and manage which vGPU profiles can be used on the host server.
How vGPUs should be allocated will be dependent on your deployment needs. Consult Citrix and your NVIDIA representative to determine the best configuration for your environment.