c++矩阵计算性能对比:Eigen和GPU解读

#include  #include  using namespace std; // 生成随机数 double GenerateRandomRealValue() {     std::random_device rd;     std::default_random_engine eng(rd());     std::uniform_real_distribution distr(1, 10);     return distr(eng); } int main() {         // 3d矩阵     double a[3][3];     for (int i = 0; i <3; ++i) {         for (int j = 0;  j <3; ++j) {             a[i][j] = GenerateRandomRealValue();         }     }     return 0; }

#include "Eigen/Dense" #include  using namespace std; using namespace Eigen; MatrixXd Generate2DMatrixByEigen() {         // 直接使用内置的Random，产生均匀分布随机矩阵     MatrixXd m = MatrixXd::Random(3,3);          // 也可以调用自定义的随机数生成函数填充数据     // MatrixXd m = MatrixXd::Zero(3,3).unaryExpr(std::bind(GenerateRandomRealValue));     return m; }

void CalcMatrixDotForLoop(const vector>& a, const vector>& b) {     std::chrono::high_resolution_clock::time_point t1 = std::chrono::high_resolution_clock::now();     if (a[0].size() != b.size()) {         cout << "error:" << a.size() << "," << b[0].size() << endl;         return;     }     vector> c;     vector c_row(b[0].size());     for (int i = 0; i  time_span = t2 - t1;     std::cout << "Loop takes " << time_span.count() << " ms\n";     // cout << "matrix c:\n";     // for (int i = 0; i 

void ModeEigen(const int a_row, const int a_col, const int b_row, const int b_col) {     std::chrono::high_resolution_clock::time_point t1 = std::chrono::high_resolution_clock::now();     auto c = a * b;     std::chrono::high_resolution_clock::time_point t2 = std::chrono::high_resolution_clock::now();     std::chrono::duration time_span = t2 - t1;     std::cout << "Eigen takes " << time_span.count() << " ms\n";     // cout << "matrix c:\n" << c << endl; }

auto t_begin = std::chrono::high_resolution_clock::now(); t1 = std::chrono::high_resolution_clock::now(); cudaMalloc((void**)&da,size); cudaMalloc((void**)&db,size); cudaMalloc((void**)&dc,size); t2 = std::chrono::high_resolution_clock::now(); time_span = t2 - t1; std::cout << "GPU malloc takes " << time_span.count() << " ms\n"; t1 = std::chrono::high_resolution_clock::now(); cudaMemcpy(da,a,size,cudaMemcpyHostToDevice); cudaMemcpy(db,b,size,cudaMemcpyHostToDevice); t2 = std::chrono::high_resolution_clock::now(); time_span = t2 - t1; std::cout << "cudaMemcpy takes " << time_span.count() << " ms\n"; t1 = std::chrono::high_resolution_clock::now(); dim3 dg(32,32); dim3 dbs((n+dg.x-1)/dg.x,(n+dg.y-1)/dg.y); mextix<<>>(da,db,dc,n); t2 = std::chrono::high_resolution_clock::now(); time_span = t2 - t1; std::cout << "gpu takes " << time_span.count() << " ms\n"; t1 = std::chrono::high_resolution_clock::now(); cudaMemcpy(c,dc,size,cudaMemcpyDeviceToHost); t2 = std::chrono::high_resolution_clock::now(); time_span = t2 - t1; std::cout << "cudaMemcpy back takes " << time_span.count() << " ms\n"; cudaFree(da); cudaFree(db); cudaFree(dc); auto t_end = std::chrono::high_resolution_clock::now(); time_span = t_end - t_begin; std::cout << "GPU total takes " << time_span.count() << " ms\n";