arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2603.26469cs.DCcs.AIcs.LG
收藏

UNIFERENCE:一种用于开发分布式AI模型的离散事件模拟框架

UNIFERENCE: A Discrete Event Simulation Framework for Developing Distributed AI Models

Northwestern University(西北大学)

Doğaç Eldenk, Stephen Xia

AI总结:

UNIFERENCE通过轻量逻辑过程模拟设备和网络行为,实现分布式AI模型的开发、基准测试和部署,其在不同后端和硬件配置上达到98.6%的运行时准确性。

摘要:

开发和评估分布式推断算法仍然困难,因为缺乏标准化工具来建模异构设备和网络。现有研究常依赖于随意的测试床或专有基础设施,导致结果难以复现并限制了对假设硬件或网络配置的探索。我们提出了UNIFERENCE,一种离散事件模拟(DES)框架,用于在统一环境中开发、基准测试和部署分布式AI模型。UNIFERENCE通过轻量逻辑过程模拟设备和网络行为,仅在通信原语上同步,消除回滚同时保持因果顺序。它无缝集成PyTorch Distributed,使相同代码库能够从模拟过渡到真实部署。我们的评估显示,UNIFERENCE在不同后端和硬件配置上,与真实物理部署的运行时准确性高达98.6%。通过连接模拟和部署,UNIFERENCE提供了一个可访问、可复现的平台,用于研究分布式推断算法和探索未来系统设计,从高性能集群到边缘规模设备。该框架在https://github.com/Dogacel/Uniference开源。

英文摘要:

Developing and evaluating distributed inference algorithms remains difficult due to the lack of standardized tools for modeling heterogeneous devices and networks. Existing studies often rely on ad-hoc testbeds or proprietary infrastructure, making results hard to reproduce and limiting exploration of hypothetical hardware or network configurations. We present UNIFERENCE, a discrete-event simulation (DES) framework designed for developing, benchmarking, and deploying distributed AI models within a unified environment. UNIFERENCE models device and network behavior through lightweight logical processes that synchronize only on communication primitives, eliminating rollbacks while preserving the causal order. It integrates seamlessly with PyTorch Distributed, enabling the same codebase to transition from simulation to real deployment. Our evaluation demonstrates that UNIFERENCE profiles runtime with up to 98.6% accuracy compared to real physical deployments across diverse backends and hardware setups. By bridging simulation and deployment, UNIFERENCE provides an accessible, reproducible platform for studying distributed inference algorithms and exploring future system designs, from high-performance clusters to edge-scale devices. The framework is open-sourced at https://github.com/Dogacel/Uniference.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑