一种应用于人工智能计算的分布式异构数据同步技术研究

打开文本图片集

doi：10.3969/J.ISSN.1672-7274.2025.06.008

中图分类号：TP393.08 文献标志码：A 文章编码：1672-7274（2025）06-0022-05

Research on Distributed Heterogeneous Data Synchronization Technology Applied to Artificial Intelligence Computing

ZHANG Keyi，WEN Hao，XING Shaosong （China Institute of Aeronautical SystemsEngineering，Beijing 1ooo2o，China）

Abstract： This article focuses on the development needs of artificial intelligence technology and proposes a data synchronization technology solution fordistributed heterogeneous data.This technologyachieves business collaboration and data fusion between traditional data systems and high-performance computing by building a unified identity authentication mechanism and data upload and download functions，designing and developing data synchronization interfaces.Through experimentalverification，this technical solutionhas shownexcellnt performance in data transmission eficiency and accuracy，and can effectively support high-performance computing platforms to carry out paralel computing of large-scale data.Itcan meet the needs of data management and algorithm training in artificial intelligence technology research with minimal modification costs.

Keywords： data system; high performance computing; data synchronization

近年来，航空专业领域信息化程度不断提高，飞机设计过程中需要进行的空气动力学设计、结构强度设计、零部件装配设计等诸多环节，已从传统的图纸设计和公式计算向基于计算机的数值模拟仿真演化，这种变化所带来的技术和能力革新主要从大数据和高性能计算两个方向进行，其代表性技术产物分别是数据系统和高性能计算平台。（剩余5324字）