并行计算之路<1>——Hello, CUDA.

最新推荐文章于 2024-04-19 09:41:25 发布

xx116213

最新推荐文章于 2024-04-19 09:41:25 发布

阅读量1.5k

点赞数

分类专栏：并行计算文章标签： cuda 并行计算编程

本文链接：https://blog.csdn.net/xx116213/article/details/50663414

版权

并行计算专栏收录该内容

5 篇文章 2 订阅

订阅专栏

you will say goodbye to girls if you are saying “hello world”. 码猿有风险，入行需谨慎

第一个CUDA程序

参考《GPGPU编程技术——从GLSL、CUDA到OpenCL》的4.3节第一个CUDA程序，因为版本的不同所以对原书的代码进行修改。

修改后的代码

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "cuda_runtime.h"
#include "device_launch_parameters.h"

__global__ void my_first_kernel(float *x)
{
    int i = threadIdx.x + blockDim.x * blockIdx.x;

    *(x + i) = (float) threadIdx.x;
}

int main(int argc, char **argv)
{
    float *pfCPU = NULL;
    float *pfGPU = NULL;
    int nBlocks, nThreads, nSize, n;

    // 设置block数、每个block的线程数
    nBlocks = 2;
    nThreads = 8;
    nSize = nBlocks * nThreads;

    // 分配CPU和GPU存储空间
    pfCPU = (float *)malloc(nSize * sizeof(float));
    cudaMalloc((void **)&pfGPU, nSize * sizeof(float));

    // 数据初始化
    memset(pfCPU, 0, nSize * sizeof(float));
    cudaMemset(pfGPU, 0, nSize * sizeof(float));

    // 执行全局函数
    my_first_kernel<<< nBlocks, nThreads >>>(pfGPU);

    // 取回计算结果，并打印输出
    cudaMemcpy(pfCPU, pfGPU, nSize * sizeof(float), cudaMemcpyDeviceToHost);

    for (n = 0; n < nSize; n++)
    {
        printf("%d %f \n", n, *(pfCPU + n));
    }

    // 回收存储空间
    cudaFree(pfGPU);
    free(pfCPU);

    return 0;
}

代码分析

1) CUDA 初始化设备时默认使用0号设备，也可以通过 cudaSetDevice() 函数来启动其它GPU设备。
2) malloc() 分配内存空间
3) cudaMalloc() 分配显存空间
4) memset() 初始化内存空间
5) cudaMemset() 初始化显存空间
6) 限定符 __global__ 表示全局函数，从主机调用，在设备里执行。
7) cudaMemcpy() 将处理结果复制到目的内存。
8) free() 以及 cudaFree() 分别释放内存和显存的空间。

CUDA提供的demo

新建 CUDA Runtime的时候，会提供一个demo。当然比第一个严谨规范。

代码


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

#include <stdio.h>

cudaError_t addWithCuda(int *c, const int *a, const int *b, unsigned int size);

__global__ void addKernel(int *c, const int *a, const int *b)
{
    int i = threadIdx.x;
    c[i] = a[i] + b[i];
}

int main()
{
    const int arraySize = 5;
    const int a[arraySize] = { 1, 2, 3, 4, 5 };
    const int b[arraySize] = { 10, 20, 30, 40, 50 };
    int c[arraySize] = { 0 };

    // Add vectors in parallel.
    cudaError_t cudaStatus = addWithCuda(c, a, b, arraySize);
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "addWithCuda failed!");
        return 1;
    }

    printf("{1,2,3,4,5} + {10,20,30,40,50} = {%d,%d,%d,%d,%d}\n",
        c[0], c[1], c[2], c[3], c[4]);

    // cudaDeviceReset must be called before exiting in order for profiling and
    // tracing tools such as Nsight and Visual Profiler to show complete traces.
    cudaStatus = cudaDeviceReset();
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "cudaDeviceReset failed!");
        return 1;
    }

    return 0;
}

// Helper function for using CUDA to add vectors in parallel.
cudaError_t addWithCuda(int *c, const int *a, const int *b, unsigned int size)
{
    int *dev_a = 0;
    int *dev_b = 0;
    int *dev_c = 0;
    cudaError_t cudaStatus;

    // Choose which GPU to run on, change this on a multi-GPU system.
    cudaStatus = cudaSetDevice(0);
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "cudaSetDevice failed!  Do you have a CUDA-capable GPU installed?");
        goto Error;
    }

    // Allocate GPU buffers for three vectors (two input, one output)    .
    cudaStatus = cudaMalloc((void**)&dev_c, size * sizeof(int));
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "cudaMalloc failed!");
        goto Error;
    }

    cudaStatus = cudaMalloc((void**)&dev_a, size * sizeof(int));
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "cudaMalloc failed!");
        goto Error;
    }

    cudaStatus = cudaMalloc((void**)&dev_b, size * sizeof(int));
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "cudaMalloc failed!");
        goto Error;
    }

    // Copy input vectors from host memory to GPU buffers.
    cudaStatus = cudaMemcpy(dev_a, a, size * sizeof(int), cudaMemcpyHostToDevice);
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "cudaMemcpy failed!");
        goto Error;
    }

    cudaStatus = cudaMemcpy(dev_b, b, size * sizeof(int), cudaMemcpyHostToDevice);
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "cudaMemcpy failed!");
        goto Error;
    }

    // Launch a kernel on the GPU with one thread for each element.
    addKernel<<<1, size>>>(dev_c, dev_a, dev_b);

    // Check for any errors launching the kernel
    cudaStatus = cudaGetLastError();
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "addKernel launch failed: %s\n", cudaGetErrorString(cudaStatus));
        goto Error;
    }

    // cudaDeviceSynchronize waits for the kernel to finish, and returns
    // any errors encountered during the launch.
    cudaStatus = cudaDeviceSynchronize();
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "cudaDeviceSynchronize returned error code %d after launching addKernel!\n", cudaStatus);
        goto Error;
    }

    // Copy output vector from GPU buffer to host memory.
    cudaStatus = cudaMemcpy(c, dev_c, size * sizeof(int), cudaMemcpyDeviceToHost);
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "cudaMemcpy failed!");
        goto Error;
    }

Error:
    cudaFree(dev_c);
    cudaFree(dev_a);
    cudaFree(dev_b);

    return cudaStatus;
}

代码分析

其实不难发现，CUDA处理的流程如下。
1) 选择计算的GPU
2) 分配显存（或内存）空间
3) 初始化数据
4) 调用核函数
5) 处理结果数据
6) 释放显存（或内存）空间

SDK 和函数库

库名	说明
Thrust	一个类似于STL针对CUDA的C++模板库
NVPP	英伟达基本性能库
cuBLAS	GPU 的基本线性代数函数库
CUFFT	GPU 的快速傅里叶函数库
cuSparse	GPU 的稀疏矩阵数据的线性代数和矩阵操作库
Magma	一个用于数值计算和线性代数计算的函数库
GPU AI	GPU 路径规划函数库
CUDA Math lib	GPU 标准数学函数