NVIDIA vGPU Information & Specifications

Users demand a better experience when using graphics-intensive applications. While it might be okay for a timecard form to load slightly slowly, it’s not acceptable for a video to jitter.

Between the options of throwing more CPU at the problem or taking the vGPU route, the answer is obvious. And that’s is why we’re seeing a rapid adoption of vGPU in VDI environments.

With ControlUp 7.1 we have introduced vGPU metrics to the ControlUp console. Let’s take a look at what’s possible with these new additions:

  • Static VM level metrics like vGPU Model and architecturenumber of cores and driver version
  • Runtime VM level metrics like available memoryframe buffer size and amount allocatedvideo encoder and encoder usage.
  • Runtime process level metrics show how many vGPU resources each relevant process consumes.
  • Historical and trending vGPU metrics will be added to ControlUp Insights (coming soon), you’ll be able to answer questions like: “which applications are the most vGPU intensive?” Or ”At our current growth rate how long until we need to add more resources?”

In Computers view, we offer a detailed preset view of the NVIDIA GPU columns. This is a detailed explanation of the columns:

GPU Architecture

GPU's architecture name

GPU Available Memory NVIDIA GPU available memory in megabytes (MB)

GPU Frame Buffer Size

Size of memory assigned to the NVIDIA GPU

GPU Frame Buffer Usage NVIDIA GPU Frame Buffer usage in percents
GPU Memory Usage NVIDIA GPU memory usage in percents
GPU Model GPU name or NVIDIA GRID GPU profile type
GPU Number of Cores Number of NVIDIA CUDA cores
GPU NVIDIA driver version Current version of the installed NVIDIA GPU driver
GPU NVIDIA License Port Primary NVIDIA license server port
GPU NVIDIA License Server Primary NVIDIA license server name
GPU Usage NVIDIA GPU usage in percents
GPU Video Decoder Usage NVIDIA GPU video decoder usage in percents
GPU Video Encoder Usage NVIDIA GPU video encoder usage in percents


  • ControlUP Agent - must be installed on the machine where the GPU is installed.
  • NVIDIA WMI v2.30 - current version for 5.0,5.1,5.2.
  • GPU OS Support: 
    • Windows 7 64bit
    • Windows 8/8.1 64bit
    • Windows 10 64bit
    • Windows Server 2008 R2 64bit
    • Windows Server 2012 R2 64bit
    • Windows 2016 64bit
  • NVIDIA GRID Windows Driver Support: 
    • 5.0 - 385.41
    • 5.1 - 385.90
    • 5.2 - 386.09
  • Supported GPUs:
    • M6
    • M10
    • M60
    • P4
    • P6
    • P40
    • P100
  • Hypervisors Support:
    • All hypervisors supported by ControlUp 7.1

Known Issues:

After rebooting a machine the ControlUP Agent goes up before the NVIDIA driver and then information is not being pulled properly from Nvidia's API. 


  • Restart the ControlUp Agent by right clicking the machine > Agent Control > Restart Remote Agent. 

If you wish to change the Reg Keys for the number of retries to contact.

  • Registry path: HKEY_LOCAL_MACHINE\SOFTWARE\Smart-X\ControlUp\Agent\GPU (create GPU key)
  • Keys (DWORD):
    • NumberOfRetries - 5 by default
    • IntervalBetweenRetries - 60000ms by default


- Full utilization of your GPU can be shown in the command "nvidia-smi" (cmd)


- Our blog post about Managing vGPU in VDI



Was this article helpful?
0 out of 0 found this helpful
Have more questions? Submit a request
Powered by Zendesk