GPU Monitoring
  • 13 May 2025
  • 2 Minutes to read
  • Dark
    Light
  • PDF

GPU Monitoring

  • Dark
    Light
  • PDF

Article summary

You can use ControlUp to monitor the performance of your GPUs in real-time, whether on data is on the machine, session, or process level.

In addition to metrics for NVIDIA GPUs, ControlUp collects data for all GPU models by using Windows built-in GPU metrics. This means that you can display GPU metrics in the Real-Time DX Console and the VDI Web App just as you would by running the Task Manager on the machine with the GPU installed.

System Prerequisites

The Real-Time DX Console supports a broad range of graphic card manufacturers. The monitored machine running the GPU must meet one of the following requirements:

  • Non-NVIDIA GPU models require the Windows Display Driver Model (WDDM).
  • Physical Windows machine or virtual Windows machine with passthrough devices.
  • Any other Windows/VM configuration where data is visible from the Task Manager.

Monitor GPU Data

To enable GPU monitoring:

  1. Deploy the ControlUp Agent on each machine that runs the GPU.
  2. On each of the same machines, set the following registry values:

Registry Key: HKEY_LOCAL_MACHINE\SOFTWARE\Smart-X\ControlUp\Agent\GPU\
Value Name: IsGPUDisabled
Value Type: REG_DWORD
Value Data: 0

Registry Key: HKEY_LOCAL_MACHINE\SOFTWARE\Smart-X\ControlUp\Agent\GPU\
Value Name: EnableNvidiaGPUCollection
Value Type: REG_DWORD
Value Data: 0

Tip for adding registry values on multiple machines

Use the Controllers pane to set the registry keys on multiple machines simultaneously.

Collecting Data from NVIDIA API (Optional)

Windows is the default data source for collecting GPU metrics. To switch back to collecting data from the NVIDIA API, set the following registry values on the machine with the NVIDIA GPU installed:

Registry Key: HKEY_LOCAL_MACHINE\SOFTWARE\Smart-X\ControlUp\Agent\GPU\
Value Name: EnableNvidiaGPUCollection
Value Type: REG_DWORD
Value Data: 1

GPU Metrics

ControlUp provides the following GPU-related metrics to help you monitor your machines' GPUs:

Metric NameDescriptionTab
Average GPU Frame Buffer UsageAverage frame buffer usage for all GPUsFolders
Average GPU UsageAverage usage of all GPUsFolders
GPU ArchitectureGPU architectureMachine
GPU Available MemoryGPU available memory in megabytes (MB)Machine
GPU CPU UtilizationGPU CPU UtilizationSession
GPU Decoder UtilizationGPU decoder UtilizationSession/Process
GPU Driver versionCurrent version of the installed GPU driverSession
GPU Encoder UtilizationGPU encoder UtilizationSession/Process
GPU Frame Buffer Memory UtilizationGPU Frame Buffer Memory UtilizationSession/Process
GPU Frame Buffer SizeSize of memory assigned to the GPUMachine
GPU Frame Buffer UsageUsed size of the physical graphics card frame buffer memory in percentMachine
GPU License PortPrimary license server portMachine
GPU License ServerPrimary license server nameMachine
GPU Memory UsageGPU memory usage in %Machine
GPU ModelGPU name or GRID GPU profile typeMachine
GPU Number of CoresNumber of CUDA coresMachine
GPU UsageGPU usage in %Machine
GPU UtilizationGPU UtilizationSession/Process
GPU Video Decoder UsageGPU video decoder usage in %Machine
GPU Video Encoder UsageGPU video encoder usage in %Machine
Machines with GPUNumber of machines with GPUsFolders

Was this article helpful?