Advanced

Congestion Control for AI Fabrics

Collective traffic is the worst case a network can see: hundreds of senders transmit to the same receiver at the same instant (incast), every flow is large and long-lived, and RDMA cannot tolerate a dropped packet. Congestion control is the set of mechanisms (PFC, ECN with DCQCN, adaptive routing, packet spraying) that keep queues from overflowing without stalling the fabric. On plain Ethernet a busy all-reduce can fall to about 60% of link rate; with a tuned control loop it holds above 90%. Reading the counters that show which one you have is the on-call skill.

Unlock the full curriculum — ₹2,000 / $25every concept + every answer · 6 months · no auto-renew

RELATED CONCEPTS

RDMA, InfiniBand and RoCEv2 Rail-Optimized and Fat-Tree Fabrics NCCL and Collective Algorithms Debugging a Slow All-Reduce NVLink and Fabric Faults Ring vs Tree All-Reduce

LESSONS THAT TEACH THIS

Fleet and ReliabilityIsolation is what you sell, and the levels guarantee different things15 min→

PRACTICE THIS IN REAL QUESTIONS

Networking, Interconnects & StorageWhat is incast, why does it break an RDMA fabric, and what do ECN and priority flow control each do about it?→Networking, Interconnects & StorageWhat is Ultra Ethernet trying to fix, and does it change how you would choose a fabric today?→Networking, Interconnects & StorageInfiniBand or RoCE version 2 for a new GPU training cluster. Make the call and say what would reverse it.→Networking, Interconnects & StorageAfter a switch firmware update, all-reduce bandwidth across the cluster fell 20 percent. Every link is up and no errors are logged. Find it.→Networking, Interconnects & StorageSeveral unrelated jobs slow down at once and nothing is dropping packets. How do you confirm a pause cascade and find where it started?→GPU & Accelerator ArchitectureExplain GPUDirect RDMA. What does it remove from the data path, and when does it change performance?→