📚 Resources

• DCGM Exporter + Prometheus + Grafana for GPU observability: https://github.com/NVIDIA/dcgm-exporter
• Grafana DCGM Exporter dashboard: https://grafana.com/grafana/dashboards/12239-nvidia-dcgm-exporter-dashboard/
• Cost/utilization right-sizing figures for GPU fleets: https://www.spheron.network/blog/gpu-monitoring-for-ml/
• Setting up Prometheus — NVIDIA GPU Telemetry: https://docs.nvidia.com/datacenter/cloud-native/gpu-telemetry/latest/kube-prometheus.html
• vLLM Production Metrics: https://docs.vllm.ai/en/stable/usage/metrics/

Week 16 · Capstone Project & Course Review

Notion ID: 392c08fd-0278-8194-84c1-d7742d175166 Notion URL: https://app.notion.com/p/392c08fd0278819484c1d7742d175166