1， 官方下载网址

    注意，本文并不使用nv预编译的包来安装，仅供参考：

NVIDIA Collective Communications Library (NCCL) | NVIDIA Developer

2，github网址

    这里是nv开源的nccl源代码，功能完整，不需要有任何疑虑：

GitHub - NVIDIA/nccl: Optimized primitives for collective multi-GPU communication

3，文档网址

    这里是官方教程，本文示例是根据其中的example改写的：

Using NCCL — NCCL 2.19.3 documentation

4，源码下载与安装

    4.1，下载

git clone --recursive https://github.com/NVIDIA/nccl.git

    4.2，编译

cd nccl
make -j src.build

或者为了节省编译时间和硬盘空间，可以指定gpu的架构，以sm_70为例：

make -j src.build NVCC_GENCODE="-gencode=arch=compute_70,code=sm_70"

效果图：

    4.3，打包

安装打包deb的工具：

sudo apt install build-essential devscripts debhelper fakeroot

打包：

make pkg.debian.build
ls build/pkg/deb/

 

    4.4，安装

其中，deb包的文件名中包含了cuda版本号，以自己生成的安装包的名字为准：

sudo dpkg -i  build/pkg/deb/libnccl2_2.19.4-1+cuda12.1_amd64.deb
sudo dpkg -i  build/pkg/deb/libnccl-dev_2.19.4-1+cuda12.1_amd64.deb

5，示例

这里的示例是单机单线程多卡的示例，本文使用了双gpu显卡为例，即，在一个进程中迭代操作两个gpu 显卡，实现 allreduce操作, 四个 float vector， S0,S1,R0,R1,数学效果如下：

R0 = S0 + S1；R[0] = S0[0] + S1[0];

R1 = S0 + S1；R[0] = S0[0] + S1[0];

对应代码，其中 ：

sendbuff[0] 是gpu-0 显存里边的要给vector，

sendbuff[1] 是gpu-1 显存里边的要给vector，

实现的数学效果为

显存 recvbuff[0] = sendbuff[0] + sendbuff[1];

显存 recvbuff[1] = sendbuff[0] + sendbuff[1];

本示例结构：

    5.1，源代码

#include <stdlib.h>
#include <stdio.h>
#include "cuda_runtime.h"
#include "nccl.h"
#include <time.h>
#include <sys/time.h>

#define CUDACHECK(cmd) do {                         \
  cudaError_t err = cmd;                            \
  if (err != cudaSuccess) {                         \
    printf("Failed: Cuda error %s:%d '%s'\n",       \
        __FILE__,__LINE__,cudaGetErrorString(err)); \
    exit(EXIT_FAILURE);                             \
  }                                                 \
} while(0)

#define NCCLCHECK(cmd) do {                         \
  ncclResult_t res = cmd;                           \
  if (res != ncclSuccess) {                         \
    printf("Failed, NCCL error %s:%d '%s'\n",       \
        __FILE__,__LINE__,ncclGetErrorString(res)); \
    exit(EXIT_FAILURE);                             \
  }                                                 \
} while(0)

void  get_seed(long long &seed)
{
  struct timeval tv;
  gettimeofday(&tv, NULL);
  seed = (long long)tv.tv_sec * 1000*1000 + tv.tv_usec;//only second and usecond;
  printf("useconds:%lld\n", seed);
}

void  init_vector(float* A, int n)
{
  long long seed = 0;

  get_seed(seed);
  srand(seed);
  for(int i=0; i<n; i++)
  {
    A[i] = (rand()%100)/100.0f;
  }
}

void print_vector(float* A, float size)
{
  for(int i=0; i<size; i++)
    printf("%.2f ", A[i]);

  printf("\n");
}

void vector_add_vector(float* sum, float* A, int n)
{
  for(int i=0; i<n; i++)
  {
    sum[i] += A[i];
  }
}

int main(int argc, char* argv[])
{
  ncclComm_t comms[4];

  printf("ncclComm_t is a pointer type, sizeof(ncclComm_t)=%lu\n", sizeof(ncclComm_t));
  //managing 4 devices
  //int nDev = 4;
  int nDev = 2;
  //int size = 32*1024*1024;
  int size = 16*16;
  int devs[4] = { 0, 1, 2, 3 };

  float** sendbuff_host = (float**)malloc(nDev * sizeof(float*));
  float** recvbuff_host = (float**)malloc(nDev * sizeof(float*));

  for(int dev=0; dev<nDev; dev++)
  {
    sendbuff_host[dev] = (float*)malloc(size*sizeof(float));
    recvbuff_host[dev] = (float*)malloc(size*sizeof(float));
    init_vector(sendbuff_host[dev], size);
    init_vector(recvbuff_host[dev], size);
  }

  //sigma(sendbuff_host[i]); i = 0, 1, ..., nDev-1
  float* result = (float*)malloc(size*sizeof(float));
  memset(result, 0, size*sizeof(float));

  for(int dev=0; dev<nDev; dev++)
  {
    vector_add_vector(result, sendbuff_host[dev], size);

    printf("sendbuff_host[%d]=\n", dev);
    print_vector(sendbuff_host[dev], size);
  }
  printf("result=\n");
  print_vector(result, size);

  //allocating and initializing device buffers
  float** sendbuff = (float**)malloc(nDev * sizeof(float*));
  float** recvbuff = (float**)malloc(nDev * sizeof(float*));
  cudaStream_t* s = (cudaStream_t*)malloc(sizeof(cudaStream_t)*nDev);

  for (int i = 0; i < nDev; ++i) {
    CUDACHECK(cudaSetDevice(i));
    CUDACHECK(cudaMalloc(sendbuff + i, size * sizeof(float)));
    CUDACHECK(cudaMalloc(recvbuff + i, size * sizeof(float)));
    CUDACHECK(cudaMemcpy(sendbuff[i], sendbuff_host[i], size*sizeof(float), cudaMemcpyHostToDevice));
    CUDACHECK(cudaMemcpy(recvbuff[i], recvbuff_host[i], size*sizeof(float), cudaMemcpyHostToDevice));
    CUDACHECK(cudaStreamCreate(s+i));
  }

  //initializing NCCL
  NCCLCHECK(ncclCommInitAll(comms, nDev, devs));

  //calling NCCL communication API. Group API is required when using
  //multiple devices per thread
  NCCLCHECK(ncclGroupStart());
  printf("blocked ncclAllReduce will be calleded\n");
  fflush(stdout);

  for (int i = 0; i < nDev; ++i)
    NCCLCHECK(ncclAllReduce((const void*)sendbuff[i], (void*)recvbuff[i], size, ncclFloat, ncclSum, comms[i], s[i]));

  printf("blocked ncclAllReduce is calleded nDev =%d\n", nDev);
  fflush(stdout);
  NCCLCHECK(ncclGroupEnd());

  //synchronizing on CUDA streams to wait for completion of NCCL operation
  for (int i = 0; i < nDev; ++i) {
    CUDACHECK(cudaSetDevice(i));
    CUDACHECK(cudaStreamSynchronize(s[i]));
  }

  for (int i = 0; i < nDev; ++i) {
    CUDACHECK(cudaSetDevice(i));
    CUDACHECK(cudaMemcpy(recvbuff_host[i], recvbuff[i], size*sizeof(float), cudaMemcpyDeviceToHost));
  }

  for (int i = 0; i < nDev; ++i) {
    CUDACHECK(cudaSetDevice(i));
    CUDACHECK(cudaStreamSynchronize(s[i]));
  }

  for(int i=0; i<nDev; i++) {
    printf("recvbuff_dev2host[%d]=\n", i);
    print_vector(recvbuff_host[i], size);
  }

  //free device buffers
  for (int i = 0; i < nDev; ++i) {
    CUDACHECK(cudaSetDevice(i));
    CUDACHECK(cudaFree(sendbuff[i]));
    CUDACHECK(cudaFree(recvbuff[i]));
  }

  //finalizing NCCL
  for(int i = 0; i < nDev; ++i)
      ncclCommDestroy(comms[i]);

  printf("Success \n");
  return 0;
}

    5.2，编译

    参考Makefile中的如下一条：

single_thread_allreduce: single_thread_allreduce.cpp
    g++ -g $< -o $@ $(LD_FLAGS)

Makefile:

LD_FLAGS := -lnccl -L/usr/local/cuda/lib64 -lcudart -I/usr/local/cuda/include

MPI_FLAGS := -I /usr/lib/x86_64-linux-gnu/openmpi/include -L /usr/lib/x86_64-linux-gnu/openmpi/lib -lmpi -lmpi_cxx

EXE := single_thread_allreduce oneServer_multiDevice_multiThread mpi_test

all: $(EXE)

single_thread_allreduce: single_thread_allreduce.cpp
	g++ -g $< -o $@ $(LD_FLAGS)

oneServer_multiDevice_multiThread: oneServer_multiDevice_multiThread.cpp
	g++ -g $< -o $@ $(LD_FLAGS) $(MPI_FLAGS)

mpi_test: mpi_test.cpp
	g++ -g $< -o $@ $(LD_FLAGS) $(MPI_FLAGS)

.PHONY: clean
clean:
	-rm $(EXE)

    5.3，运行

 这里没有使用 mpi，故可以直接编译运行

make && ./single_thread_allreduce

    5.4，效果

 稍微注释一下上图：

实现了数学目标：

R0 = S0 + S1；R[0] = S0[0] + S1[0];

R1 = S0 + S1；R[0] = S0[0] + S1[0];

6，另一个示例

另一效果类似，但是是多进程的示例，其中的多进程部分是使用mpi实现的，太长了不容易翻，见下一篇