MKL 进行矩阵向量运算

/* C source code is found in dgemm_example.c */


#define min(x,y) (((x) < (y)) ? (x) : (y))


#include
#include
#include "mkl.h"


int main()
{
    float *A, *B, *C;
    int m, n, k, i, j;
    int l,r;
    double s_initial,s_elapsed;
    int LOOP_COUNT =100;
    float alpha, beta;
    double sum;


    printf ("\n This example computes real matrix C=alpha*A*B+beta*C using \n"
            " Intel(R) MKL function dgemm, where A, B, and  C are matrices and \n"
            " alpha and beta are float precision scalars\n\n");


    m = 2000, k = 200, n = 1000;
    printf (" Initializing data for matrix multiplication C=A*B for matrix \n"
            " A(%ix%i) and matrix B(%ix%i)\n\n", m, k, k, n);
    alpha = 1.0; beta = 0.0;


    printf (" Allocating memory for matrices aligned on 64-byte boundary for better \n"
            " performance \n\n");
    A = (float *)mkl_malloc( m*k*sizeof( float ), 64 );
    B = (float *)mkl_malloc( k*n*sizeof( float ), 64 );
    C = (float *)mkl_malloc( m*n*sizeof( float ), 64 );
    if (A == NULL || B == NULL || C == NULL) {
      printf( "\n ERROR: Can't allocate memory for matrices. Aborting... \n\n");
      mkl_free(A);
      mkl_free(B);
      mkl_free(C);
      return 1;
    }


    printf (" Intializing matrix data \n\n");
    for (i = 0; i < (m*k); i++) {
        A[i] = (float)(i+1);
    }


    for (i = 0; i < (k*n); i++) {
        B[i] = (float)(-i-1);
    }


    for (i = 0; i < (m*n); i++) {
        C[i] = 0.0;
    }


    printf (" Computing matrix product using Intel(R) MKL dgemm function via CBLAS interface \n\n");
    cblas_sgemm(CblasRowMajor, CblasNoTrans, CblasNoTrans, 
                m, n, k, alpha, A, k, B, n, beta, C, n);
    printf ("\n Computations completed.\n\n");


    printf (" Top left corner of matrix A: \n");
    for (i=0; i       for (j=0; j         printf ("%12.0f", A[j+i*k]);
      }
      printf ("\n");
    }


    printf ("\n Top left corner of matrix B: \n");
    for (i=0; i       for (j=0; j         printf ("%12.0f", B[j+i*n]);
      }
      printf ("\n");
    }
    
    printf ("\n Top left corner of matrix C: \n");
    for (i=0; i       for (j=0; j         printf ("%12.5G", C[j+i*n]);
      }
      printf ("\n");
    }




/* C source code is found in dgemm_with_timing.c */


    printf (" Making the first run of matrix product using Intel(R) MKL dgemm function \n"
            " via CBLAS interface to get stable run time measurements \n\n");
    cblas_sgemm(CblasRowMajor, CblasNoTrans, CblasNoTrans, 
                m, n, k, alpha, A, k, B, n, beta, C, n);


    printf (" Measuring performance of matrix product using Intel(R) MKL dgemm function \n"
            " via CBLAS interface \n\n");
    s_initial = dsecnd();
    for (r = 0; r < LOOP_COUNT; r++) {
        cblas_sgemm(CblasRowMajor, CblasNoTrans, CblasNoTrans, 
                    m, n, k, alpha, A, k, B, n, beta, C, n);
    }
    s_elapsed = (dsecnd() - s_initial) / LOOP_COUNT;


    printf (" == Matrix multiplication using Intel(R) MKL dgemm completed == \n"
            " == at %.5f milliseconds == \n\n", (s_elapsed * 1000));


/* C source code is found in matrix_multiplication.c */


    printf (" Making the first run of matrix product using triple nested loop\n"
            " to get stable run time measurements \n\n");
    for (i = 0; i < m; i++) {
        for (j = 0; j < n; j++) {
            sum = 0.0;
            for (l = 0; l < k; l++)
                sum += A[k*i+l] * B[n*l+j];
            C[n*i+j] = sum;
        }
    }


    printf (" Measuring performance of matrix product using triple nested loop \n\n");
    s_initial = dsecnd();
    for (r = 0; r < LOOP_COUNT; r++) {
        for (i = 0; i < m; i++) {
            for (j = 0; j < n; j++) {
                sum = 0.0;
                for (l = 0; l < k; l++)
                    sum += A[k*i+l] * B[n*l+j];
                C[n*i+j] = sum;
            }
        }
    }
    s_elapsed = (dsecnd() - s_initial) / LOOP_COUNT;
    
    printf (" == Matrix multiplication using triple nested loop completed == \n"
            " == at %.5f milliseconds == \n\n", (s_elapsed * 1000));




    printf ("\n Deallocating memory \n\n");
    mkl_free(A);
    mkl_free(B);
    mkl_free(C);


    printf (" Example completed. \n\n");
    return 0;
}