Unified KV Pooling to Accelerate Long-Context LLM Serving (opens in new tab)

Long-context LLM serving requires offloading KV caches to host-memory and SSDs, but existing mechanisms are not designed for such long contexts. We observe significant inefficiencies in current KV caching in long contexts: high serving latency ~30.7 s, exceeding the typical TTFT requirement of 10 s by more than 3x. Our in-depth analysis explains two major reasons: (1) retrieval is serialized through host-memory and SSD, leaving other host-memory...