Multi-node Serving · Part 4

Multi-Node KV Cache Management

Treating KV cache as a first-class, hash-addressed, tiered cluster resource instead of private per-replica state, via Mooncake's architecture and the LMCache/NIXL layer it composes with.

07Prefix Caching and Tiered KV Offload Across Nodes: Mooncake's Architecture40 min08LMCache and NIXL: The Reusable KV Layer and the Transfer-Library Primitive25 min