Nmax, the matrix size N that gives the highest flops.
得出最高 flop 时所使用的矩阵大小 N。
Binning also can reduce the overall matrix size, thus allowing higher frame rates.
绑定也可以减少总矩阵大小,因此可以获得更高的帧率。
Matrix LU decomposition parallel implementation, the document gives a different matrix size and the number of processors under the experimental results.
说明: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。
应用推荐