Ready Cohorts：界定LLM智能体控制中的GPU机会与避免主机往返
原标题：Ready Cohorts: Bounding GPU Opportunity and Avoiding Host Round Trips in LLM-Agent Control
AI 摘要
该论文研究LLM智能体服务中的GPU控制门控问题，通过两个研究定义了可测量的GPU控制门：截止期限可行的队列供应和观测放置。研究分析了并发队列调度和端上路由与主机重新调度的对比，发现设备驻留路径在36种配置中全部更快，速度提升1.19倍至2.39倍。论文还提出了固定分区份额、精确离线份额等指标，并通过动态规划精确计算离线份额。
正文节选
Ready Cohorts: Bounding GPU Opportunity and Avoiding Host Round Trips in LLM-Agent Control Abstract Two studies define measurable GPU control gates for LLM-agent services by analyzing concurrent cohort scheduling and on-device routing versus host redispatch. LLM-agent services repeatedly execute small deterministic transitions between model and tool calls: route an outcome, update state, and emit the next effect. We ask when this control path exposes enough concurrent work for GPU execution, and