中文摘要：

基于已完成重大升级改造的北京正负电子对撞机（BEPCII）和北京谱仪（BESIII）的数据分析工作正在全面展开。作为数据分析的基础和关键的BESIII离线物理软件,它的性能决定着数据分析的进度和结果。针对BESIII实验物理数据的并行计算处理,在多核CPU硬件结构基础上,采用多进程编程技术进而实现文件级别的多进程BESIII离线数据并行处理。性能测试表明多进程并行处理可以减少资源消耗而且能够有效地提高数据分析的运算速度,从而为BESIII实验早出快出高质量的物理成果奠定坚实的基础。

英文摘要：

The update of BEPCII（ Beijing Electron-Positron Collider Upgrade） and BESIII（ Beijing Spectrometer III）has been completed and massive data is being analyzed. BESIII offline software is the foundation of the data analysis,and its performance determines the progress and physics results. This paper researched on the parallel computing of the BESIII experiment data. Based on the multi-core CPU architecture,multi-process programming technique was adopted to achieve the parallel computing on file-level of BESIII offline data. Performance test results show that multiple processes in parallel processing can improve the computing speed and reduce resource consumption which helps BESIII experiment to achieve high level physics results faster and earlier.