LMCache: Supercharge Your LLM with the Fastest KV Cache Layer
vLLM’s reference system for K8S-native cluster-wide deployment with community-driven performance optimization