3.2 AI Network Design Use Cases (such as Machine Learning, Large Language Models, and Pattern Recognition) 400-007 Practice Questions (Page 6)

Questions 26–27

1. 26foundation · easy

   Which design feature is most important for ensuring resilience in an AI network?

   • ADisabling routing protocols to avoid convergence delays
   • BUsing only static routes to simplify troubleshooting
   • CA single path with high bandwidth to reduce cost
   • DRedundant paths and fast failover mechanisms
   Reveal Solution Discussion Select an answer first
2. 27application · medium

   A research university is deploying a distributed machine learning training cluster. The network team wants to monitor the cluster's network performance to identify bottlenecks during training jobs. The team needs to correlate network metrics with GPU utilization to understand the impact of network congestion on training throughput. Which telemetry approach is most effective?

   • ADeploy packet capture on the core links and manually analyze the captures to identify network congestion
   • BUse a unified monitoring platform that collects streaming telemetry from switches and GPU metrics from servers with synchronized timestamps
   • CCollect network metrics from switches using SNMP and GPU metrics from the servers using a separate monitoring system
   • DUse a network performance monitoring tool that measures latency and jitter between servers
   Reveal Solution Discussion Select an answer first