Introduction

This dashboard displays GPU metrics collected from NVIDIA dcgm-exporter via a metric endpoint added to Prometheus. A separate endpoint is added to Prometheus via a Service Monitor.

Quickstart

helm install stable/prometheus-operator --generate-name \
    --set "prometheus.prometheusSpec.serviceMonitorSelectorNilUsesHelmValues=false"

kubectl create -f https://raw.githubusercontent.com/NVIDIA/gpu-monitoring-tools/2.0.0-rc.9/dcgm-exporter.yaml
kubectl create -f https://raw.githubusercontent.com/NVIDIA/gpu-monitoring-tools/2.0.0-rc.9/service-monitor.yaml

More information on the dcgm-exporter README.