AtumAI: A Principled Framework for Agentic Generation of Datacenter Control-Plane Policies
- ID: c4f04a1b
- 原文链接: https://arxiv.org/abs/2608.02569
- PDF: https://arxiv.org/pdf/2608.02569v1
- 作者: Qiushi Lin, Chaojie Zhang, Íñigo Goiri, Aditya Akella, Ricardo Bianchini, Jovan Stojkovic
- 日期: 2026-08-03
- 更新: 2026-08-03
- 分类: agents
- 来源类型: paper
- 标签: agentic-ai, datacenter, control-plane, policy-search
- 质量评分: 4/5
- 抓取时间: 2026-08-05T12:25:56.344927+00:00
中文导读
AtumAI 把数据中心控制平面策略设计拆成两部：从自然语言目标编译出可机器检查可搜索的目标约束决策变量和评估方法；再用 LLM扩散模型进化算法和代理模型组成系统化搜索循环摘要称其在工作负载放置资源伸缩和功耗管理三类任务上超过专家基线
为什么值得关注
AtumAI formalizes datacenter policy design for agentic search and refinement.
这条记录适合进入 AAIF 的 agent / AI 系统设计观察池：论文把问题限定在可检查的任务规范、搜索空间和评估方法上，再讨论如何让 LLM、扩散模型、进化算法或强化学习式记忆机制服务于可复现的系统改进。下面内容只基于 arXiv 元数据、摘要和条目既有摘要整理，未补充摘要之外的实验细节。
关键信息
- 论文标题：AtumAI: A Principled Framework for Agentic Generation of Datacenter Control-Plane Policies
- 作者：Qiushi Lin, Chaojie Zhang, Íñigo Goiri, Aditya Akella, Ricardo Bianchini, Jovan Stojkovic
- arXiv：https://arxiv.org/abs/2608.02569
- 发布时间：2026-08-03
- 更新时间：2026-08-03
- arXiv 分类：cs.AI, cs.DC, cs.OS
- 关联标签：agentic-ai, datacenter, control-plane, policy-search
English Abstract
The efficiency of a datacenter rests on its control plane policies. Designing these policies is increasingly hard: the hardware-software stack grows fast, the design space is vast and interdependent, and prototyping a single policy takes months. Agentic AI promises to automate this search. Off the shelf, however, it falls short on three fronts. It is not formal: with no structured, searchable statement of the problem, the search has little structure to exploit and hard constraints are not guaranteed. It is not transferable: each task is solved from scratch, so nothing learned on one task carries to the next. Finally, it is not systematic: relying on the LLM as the sole source of candidates, it explores a narrow slice of the design space and settles into local optima. We introduce AtumAI, a framework that generates datacenter control-plane policies with agentic AI, making the process formal, transferable, and systematic. From a goal stated in plain language, AtumAI autonomously proposes, tests, and refines candidate policies until one satisfies the request. It does so through two components. The Datacenter Task Compiler automates problem formulation: it compiles the request into a formal, machine-checkable, and searchable specification of the task's objectives, constraints, decision variables, and evaluation methodology. The Evolutionary Design Discovery Loop then searches this specification, expanding the search beyond the LLM itself via a diffusion model, an evolutionary algorithm, and a surrogate model. Together, they reduce onboarding a new task from months of engineering to writing its description. We evaluate AtumAI on three control-plane tasks with distinct problem scopes, design spaces, and trade-offs: workload placement, resource scaling, and power management. Across all tasks, the policies generated by AtumAI consistently outperform expert-engineered baselines.
English Summary
The efficiency of a datacenter rests on its control plane policies. Designing these policies is increasingly hard: the hardware-software stack grows fast, the design space is vast and interdependent, and prototyping a single policy takes months. Agentic AI promises to automate this search. Off the shelf, however, it falls short on three fronts. It is not formal: with no structured, searchable statement of the problem, the search has little structure to exploit and hard constraints are not guaranteed. It is not transferable: each task is solved from scratch, so nothing learned on one task carries to the next. Finally, it is not systematic: relying on the LLM as the sole source of candidates, it explores a narrow slice of the design space and settles into local optima....
Obsidian Notes
- 内容由
opencli arxiv paper 2608.02569 -f json拉取 arXiv 元数据与摘要生成。 - 中文导读与价值判断锚定在条目已有摘要、论文摘要、作者、日期与分类信息上。