arXiv：cs.AI· Wonmi Choi, Minuk Park, Zhixiong Niu, Yongqiang Xiong, Chuck Yoo, Gyeongsik Yang·· 21 小时前
并非所有 AI 智能体都相同：解析资源与性能动态
Not All AI Agents Are Equal: Characterizing Resource and Performance Dynamics
AI 导读
研究对检索增强问答、网页搜索和软件编程三类 LLM 智能体任务做了资源动态与延迟刻画，发现同一工具在不同任务下的资源行为差异显著，并发请求会暴露 CPU、磁盘 I/O、内存等任务相关瓶颈，更快的 LLM 响应或更多 CPU 核心并不总能加速智能体。基于此提出的 CPU 感知工具准入与任务感知 CPU 分配，使 CPU 敏感任务延迟改善约 5.4 倍，多任务平均延迟降低约 32%。
正文
Abstract:LLM-based AI agents process user requests through iterative reasoning and tool execution, often involving the invocation of remote LLM APIs with local tool containers. This execution model can make the optimization of agent serving difficult because latency, local resource demand, and container bottlenecks inter-mix across requests. However, the current agent ecosystem runs without much consideration of resource dynamics, which results in significant waste of the precious resources. This paper analyzes the resource inter-mix of AI agents for three representative tasks: retrieval-augmented question answering, web search, and software coding. To this end, we characterize the latency with respect to the resource dynamics of processing multiple requests and tasks concurrently. Our measurements show that agents have a wide range of behaviors depending on tasks, so that even the same tool can differ substantially in resource dynamics. We also find that running multiple requests concurrently exposes task-dependent bottlenecks in resource dynamics such as CPU, disk I/O, and memory. Furthermore, we uncover that faster LLM responses or more CPU cores do not always accelerate agents. Based on these observations, we demonstrate new optimization opportunities that exploit the resource dynamics of tasks: CPU-aware tool admission and task-aware CPU allocation. Our results show that the latency of CPU-sensitive agent tasks improves $\sim$5.4$\times$, and the average latency across multiple tasks is reduced $\sim$32% compared to native agents.
| Subjects: | Artificial Intelligence (cs.AI); Performance (cs.PF) |
| Cite as: | arXiv:2609.19947 [cs.AI] |
| (or arXiv:2609.19947v2 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2609.19947 arXiv-issued DOI via DataCite |
Submission history
From: Wonmi Choi [view email]
[v1]
Thu, 17 Sep 2026 09:20:53 UTC (1,534 KB)
[v2]
Mon, 12 Oct 2026 03:56:41 UTC (1,525 KB)
来源：arXiv：cs.AI · arxiv.org