CUDA 矩陣乘法

使用global memory

報錯

• 錯誤 17 error : no instance of overloaded function “cudaMalloc” matches the argument list E:\Niki\MVDR_BTR\MVDR_BTR\MatrixMulOnDevice.cu 16
  修正：把CUDAcudaMalloc(&Nd, size);改成cudaMalloc((void**)&Nd, size);

• 錯誤 17 error : argument of type “float” is incompatible with parameter of type “void *” E:\Niki\MVDR_BTR\MVDR_BTR\MatrixMulOnDevice.cu 17
  修正：把float Md, *Nd, *Pd; 改成float

  Md, Nd, Pd;

編譯通過的程式碼

#include "cuda_runtime.h"
#include "device_launch_parameters.h"

#include <stdio.h>

__global__ void MatrixMulKernel(float* Md, float* Nd, float* Pd, int Width);

void MatrixMulOnDevice(float* M, float* N, float* P, int Width)
{
   int size = Width * Width * sizeof(float
 
); 
    float  *Md, *Nd, *Pd;

// Allocate and Load M, N to device memory 
    cudaMalloc((void**)&Md, size);
    cudaMemcpy(Md, M, size, cudaMemcpyHostToDevice);

     cudaMalloc((void**)&Nd, size);
     cudaMemcpy(Nd, N, size, cudaMemcpyHostToDevice);

     // Allocate P on the device
    cudaMalloc((void
 
**)&Pd, size);



// Kernel invocation code – to be shown later
    // Setup the execution configuration
    dim3 dimBlock(Width, Width);
    dim3 dimGrid(1, 1);

    // Launch the device computation threads!
    MatrixMulKernel<<<dimGrid, dimBlock>>>(Md, Nd, Pd,Width);


// Read P from the device
      cudaMemcpy(P, Pd, size, cudaMemcpyDeviceToHost);

       // Free device matrices
      cudaFree(Md); cudaFree(Nd); cudaFree (Pd);
}

// Matrix multiplication kernel – per thread code

__global__ void MatrixMulKernel(float* Md, float* Nd, float* Pd, int Width)
{
    // 2D Thread ID
    int tx = threadIdx.x;
    int ty = threadIdx.y;

    // Pvalue is used to store the element of the matrix
    // that is computed by the thread
    float Pvalue = 0;
    for (int k = 0; k < Width; ++k)
    { 
         float Melement = Md[ty * Width + k];
         float Nelement = Nd[k * Width + tx];
         Pvalue += Melement * Nelement;
    }
    // Write the matrix to device memory;
    // each thread writes one element
    Pd[ty * Width + tx] = Pvalue;
}

測試

新增主函式程式碼：

#include "cuda_runtime.h"
#include "device_launch_parameters.h"

#include <stdio.h>
#include <iostream>
using namespace std;

void MatrixMulOnDevice(float* M, float* N, float* P, int Width);

int main()
{
   int Width;
   Width = 8;
   float M[64];
   float N[64];
   float P[64];

   int i,j;

   for(i=0;i<Width;i++)//row
   {
       for(j=0;j<Width;j++)//colume
       {
           M[i*Width+j]= i+j+1;
           N[i*Width+j]= i+1;
       }
   }

   //float *Mp = M;
   //float *Np = N;
   //float *Pp = P;

    /*MatrixMulOnDevice(Mp,Np, Pp, Width);*/
   MatrixMulOnDevice(&M[0],&N[0], &P[0], Width);
   printf("矩陣相乘結果為：\n");
   for(i=0;i<Width;i++)//row
   {
       for(j=0;j<Width;j++)//colume
       {
           printf(" %f \t", P[i*Width+j]);
       }
       printf("\n");
   }

    return 0;
}