SAC: Disaggregated KV Cache System for Sparse Attention LLMs with CXL
- 类型:arxiv
- 标识:2606.19746
- 链接:http://arxiv.org/abs/2606.19746v1
- 主分类:llm-infra
- 形态:method
- 被引:1
- 被引来源:Semantic Scholar
- S2被引:1
- OpenAlex被引:0
- 影响力被引:0
- TLDR:SAC is proposed, the first efficient disaggregated KV cache system optimized for sparse attention models, which leverages the low-latency, cache-line granularity load/store semantics of Compute Express Link (CXL), establishing CXL-based disaggregation as the superior infrastructure for emerging sparse attention models.
- OpenAlex ID:W7165402039
- OpenAlex DOI:10.48550/arxiv.2606.19746
- DOI:10.48550/arxiv.2606.19746
- DOI来源:OpenAlex
- 开放获取:green
- 开放获取链接:https://doi.org/10.48550/arxiv.2606.19746
- OpenAlex更新:2026-07-19
- 待LLM分类:否
- 标题中文:SAC:面向稀疏注意力 LLM 与 CXL 的解耦 KV Cache 系统
- TLDR中文:提出 SAC,这是首个针对稀疏注意力模型优化的高效解耦 KV Cache 系统,利用 CXL(Compute Express Link)的低延迟、cache-line 粒度 load/store 语义,将基于 CXL 的解耦确立为新兴稀疏注意力模型的优越基础设施。
- 来源文件:
- /inbox/tom/_candidates/2026-06-22-agent-rag-longcontext-candidates.json
- /inbox/tom/_candidates/2026-06-21-agent-rag-longcontext-candidates.json
- /inbox/tom/_candidates/2026-06-20-agent-rag-longcontext-candidates.json
- /inbox/tom/_candidates/2026-06-19-agent-rag-longcontext-candidates.json
- [S2 enrich]
- [OpenAlex backfill]