Contents

• 1. Overview
• 2. Initial Incident Report
• 3. GPU Node Triage
  • 3.1. Reporting a GPU Issue
  • 3.2. Understanding Xid Messages
  • 3.3. Running DCGM Diagnostics
  • 3.4. Running Field Diagnostics
  • 3.5. Network Testing
  • 3.6. Debugging Applications
• 4. Best Practices
  • 4.1. Collecting Node Metrics
  • 4.2. Catching Errors Before They Occur
• 5. Notices
  • 5.1. Trademarks