矩阵LU分解的并行实现，文档中给出了不同矩阵规模和处理器数下的实验结果。-Matrix LU decomposition parallel implementation, the document gives a different matrix size and the number of processors under the experimental results.