Nvidia Cheatsheet

Triton & DCGM Integration

Learn how our NVIDIA DCGM and Triton integrations help you monitor the health and performance of your GPUs and AI models.

This Datadog cheatsheet provides:

• Measure various metrics like power and resource consumption for our DCGM and Triton integrations
• A quick-start guide to using Datadog to collect metrics and status information to monitor and visualize NVIDIA GPU performance
• Metrics like GPU temperature to determine if workloads overload your hardware
• The ability to correlate GPU and CPU utilization alongside the overall inference load of your Triton server