Kubernetes Networking for AI: Implementation

conceptstub

Manifests, Helm values, and scripts. Every script states whether it is read-only or changes the cluster.

Planned content

• Calico install (Tigera operator) with lab-safe defaults
• Multus thick install
• NADs: macvlan, host-device, SR-IOV with Whereabouts
• Soft-RoCE (rxe) for RDMA on the CPU lab
• NCCL env for multi-NIC pods