构建首个从原理出发的机器学习系统建模框架，实现跨硬件的快速架构分析。
MLSYSIM: First-Principles Infrastructure Modeling for Machine Learning Systems
- 基于需求-供给抽象，分离计算需求与硬件供应，确保单位一致性。
- 整合22类系统瓶颈与28个可组合模型，支持亚秒级设计空间探索。
- 适合芯片设计、系统优化等需快速评估全栈性能的研究者使用。
随着机器学习从实验室走向关键基础设施，支撑它的系统跨度极大，从亚毫瓦级微控制器到千兆瓦级数据中心集群。跨尺度推理困难：实验测量需实际硬件，而周期精确模拟每配置需数小时，缺乏快速全栈架构分析工具。本文提出MLSYSIM（机器学习系统基础设施建模），一个基于第一性原理的分析框架，将系统‘物理规律’形式化为严格维度的Python引擎。该框架采用需求-供给抽象，解耦计算需求与硅基供应及环境因素，并在运行时强制单位完整性，杜绝隐式单位转换错误。所有输入均来自带类型和来源追踪的注册表，确保每个数值都有明确出处。在此基础上，我们归纳了22种‘系统墙’，并开发了28个可组合的模型与求解器，实现亚秒级的设计空间探索，识别关键约束，并生成整个机器学习系统生命周期的理想硬件规格。
原文摘要 · Abstract (English)
As machine learning shifts from laboratory curiosity to critical infrastructure, the systems that sustain it span an extraordinary range, from sub-milliwatt microcontrollers to multi-gigawatt datacenter fleets. Reasoning across this range is hard: empirical profiling requires the target hardware in hand, while cycle-accurate simulation costs hours per configuration, leaving no tool for rapid, full-stack architectural reasoning. We present MLSYSIM (Machine Learning Systems Infrastructure Modeling), a first-principles analytical framework that formalizes the "physics of systems" into a dimensionally-strict Python engine. MLSysim is built on a demand-supply abstraction that decouples computational demand from silicon supply and environmental context, and it enforces unit integrity at runtime so the silent conversion errors that plague ad-hoc modeling cannot occur. Every input is drawn from a typed, provenance-tracked registry, so no number enters an analysis without a documented source. On this engine we codify a taxonomy of 22 "Systems Walls" resolved by 28 composable models and solvers, enabling sub-second design-space exploration that identifies binding constraints and synthesizes ideal hardware specifications across the entire ML systems lifecycle.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。