资源列表
MPI
- 使用MPI并行编程，矩阵相乘程序测试及 LAPACK 的使用-The use of MPI parallel programming, matrix multiplication program testing and the use of LAPACK
mpi
- 搭建MPI集群，实现矩阵运算，对比不同线程不同规模允许时间，对比加速比。-MPI cluster structures, implement matrix operations, comparison of different threads of different sizes to allow time, compared to speedup.
BP_Hidden
- 双隐层BP神经网络实现，功能为时间序列的预测。-Double hidden layer BP neural network, the function of time series prediction.
transpose.cpp.tar
- parallel traspose of matrix using mpi
Source3
- 采用改进的三维Hasin准则，考虑基体和纤维损伤，以及复合材料分层损伤- U91C7 u7539 u659 u8F09 u7F4 u7R4 u7354 u03 U6599 u5206 u5C42 u635F u4F24
cuda+mpi
- 并行计算中矩阵相乘的mpi和cuda两种方法，已完成检验，有结果输出(In parallel computing, two methods of matrix multiplication, MPI and CUDA, have been tested and output with results)
MPI Enumeration sorting
- 通过MPI并行集群的搭建，形成并行开发环境，编程并实现枚举排序功能(The parallel development environment, programming and enumeration sorting function is formed through the construction of parallel cluster of MPI)
fft
- 基于MPI的并行快速傅里叶变换算法，通过多核并行计算提升计算性能。(the fast Fourier transform algorithm based on MPI coding)
wghf
- 在流体力学中对圆柱绕流进行网格划分和计算用的初步生成。(Preliminary generation of mesh and calculation of flow past a cylinder in fluid mechanics.)
01_查询GPU设备
- cuda代码查询设备信息，基础实用的小例子代码(the code of cuda for querying the device)
检测CPU和GPU处理图片的时间对比
- 检测CPU和GPU处理图片的时间对比,需要拥有支持CUDA的硬件设备(Time comparison between detecting CPU and GPU processing pictures)
kernel
- 1,使用VS创建一个工程 2，复制文件代码 3、使用内核cuda函数使用内核cuda函数调试专用(Use kernel CUDA function to use kernel CUDA function to debug special-purpose)