数据与计算发展前沿2026，Vol.8Issue(3)：15-28,14.DOI:10.11871/jfdc.issn.2096-742X.2026.03.002
HPC-AI融合下磁约束聚变集成建模系统中的可追溯性研究
From Building to Packaging:A Study on FAIR Data Traceability in HPC:AI-Driven Integrated Modeling of Magnetic Confinement Fusion
摘要
Abstract
[Background]With advances in magnetic confinement fusion research and computing capabilities,integrated fu-sion modeling is evolving from single-physics modules toward complex,multi-physics,multi-scale coupled sys-tems.The widespread adoption of artificial intelligence(AI)has led to an HPC-AI hybrid computing paradigm.[Objective]However,traditional HPC codes rely on stable system-level compilation environments,while AI ap-plications depend on dynamic Python ecosystems and containerization.Their fundamental differences in depen-dency management,build processes,and execution models make it difficult for the existing in-house integrated modeling framework(FuYun)to uniformly manage heterogeneous components,causing traceability gaps at the HPC-AI interface and threatening reproducibility and provenance.This study aims to address FAIR compliance challenges in software environment management and physics module execution within FuYun under HPC-AI inte-grated environments.[Methods]This study extends FuYun's scope from HPC"build"to AI"packaging,"intro-ducing a unified abstraction called the Computational Unit(CU)to encapsulate both traditional HPC programs and containerized AI applications.A cross-stack unique identifier(@pid)system and provenance tracking mecha-nism are designed.The module description schema is enhanced with AI-specific metadata fields(e.g.,model weights,training hyperparameters,random seeds)to ensure complete recording of critical information.[Results]The extend-ed framework successfully unifies management of heterogeneous components.Full-stack data provenance is achieved via the@pid system and enhanced tracking.Experiments show a 95%reproducibility rate and an 85%improvement in environment deployment efficiency over manual methods.This approach bridges the gap be-tween HPC modules and AI modules,allowing users to flexibly select different computational modules and orga-nize them into analysis workflows within an integrated platform,and to record the compute flow and result.关键词
数据管理/FAIR4RS/集成建模/容器/磁约束聚变/可追溯性Key words
data management/FAIR4RS/integrated modeling/containers/magnetic confinement fusion/provenance引用本文复制引用
刘晓娟,于治,张运动..HPC-AI融合下磁约束聚变集成建模系统中的可追溯性研究[J].数据与计算发展前沿,2026,8(3):15-28,14.基金项目
国家自然科学基金(12575217,12575272,12505263) （12575217,12575272,12505263）
国家重点研发计划(2024YFE03050002) （2024YFE03050002）
