Your question is GPU Cluster Networking. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
Design a network to support a cluster of 5,000 GPUs for Lambda's AI workloads.
Explain the topology, network fabrics, addressing, routing, bandwidth, and failure domains you would choose for distributed training and inference. Cover the data plane, control plane, observability, capacity planning, rollout strategy, and how you would validate performance and recover from failures.