Azure Monitor for containers now supports monitoring GPU usage on Azure Kubernetes Service (AKS) GPU-enabled node pool. Use it to monitor containers requesting and using GPU resources in AKS clusters. The collection will automatically happen if you have GPU-enabled nodes starting with agent version ciprod03022019. We currently support monitoring two GPU vendors:
Use GPU monitoring to:
See the available GPU nodes, GPU memory usage and pods requesting GPU and status.
Visualise through the built-in workbook available in the workbook gallery.
Write alerts on pod status.
Read more information about GPU monitoring.
Learn more about Azure Monitor for containers
