Abstract: As there are several factors that affecting the performances of the embedded systems，the article particularly analyzes the factors that affect the performances of the embedded systems，and proposes a point that hardware/software partitioning based on multi-target.Bring IP and software framework reuse into hardware/software partitioning method.Mapping the complex embedded system configuration into DAG（Direct Acyclic Graph）.Firstly proposes the concept of performance priority，and speeds up the calculating of the algorithm through the import of operation research and data preprocess.

Key words: multi-target evaluation, SoC design, hardware/software partitioning, algorithm optimization, IP core reuse, capability priority

摘要： 针对存在多种因素影响嵌入式系统综合性能的实际情况，详细分析了影响嵌入式系统性能的各项性能指标，提出了一种基于多性能指标评价的软硬件协同划分思想。利用SoC可重用的特性，将IP核复用及软件架构重用引入到软硬件划分算法当中。通过功能模块层的抽象，将复杂的嵌入式系统构成映射到数学上的DAG（Direct Acyclic Graph）之上。提出了性能指标优先级的概念，并通过在算法中加入对给定的参数数据预先处理及引入运筹学中分支定界的思想，优化了算法的求解，加快了算法的收敛速度，较之单纯的整个空间的条件遍历更优。

关键词: 多性能指标评价, SoC系统设计, 软硬件划分, 算法优化, IP核复用, 性能指标优先