TideCache:透明缓存管理与逐层流式传输的高效混合LLM服务
TideCache: Efficient Hybrid LLM Serving with Transparent Cache Management and Layer-Wise Streaming
Jinbin Luo, Xiaozhe Yao, Shang Yang, Zhijian Liu, Song Han, Qinghao Hu
AI 速读 基于摘要
TideCache 通过透明缓存池、缓存流式器和请求调度器,在 CPU-GPU 超级芯片上高效管理混合架构 LLM 的异构缓存,提升服务吞吐并降低延迟。
cs.DC · cs.LG
查看原文摘要
Hybrid architectures have become mainstream among open-weight language models. Their global-attention and bounded-state layers maintain heterogeneous caches whose demands vary with context length and concurrency. We present TideCache, a serving system on CPU--GPU superchips with three components for allocating GPU memory among caches and managing their residency across GPU and host memory. The Transparent Cache Pool redistributes physical pages among cache components behind fixed virtual addresses, preserving cache layouts and backend interfaces. The Cache Streamer uses computation intervals provided by bounded-state layers to prefetch the next global-attention layer's prefix into GPU slots shared across global-attention layers, reducing GPU residency while overlapping layer-wise transfers with computation. Generated KV remains resident. The Request Scheduler selects resident or streamed execution and triggers capacity adjustments according to request size and queue state. Together, they let HBM released by streaming accommodate other cache components and additional requests. Implemented in SGLang and evaluated on GB300 and GH200 across seven model/platform configurations, TideCache completes up to 2.04x as many agent turns, reduces overall P50 per-token latency by 20--59%, and improves agent P50 TTFT by 1.40--3.05x compared with SGLang with HiCache.