gpt4 book ai didi

cuda - 奇怪的 cuBLAS gemm 批处理性能

转载 作者:行者123 更新时间:2023-12-02 07:32:07 27 4
gpt4 key购买 nike

我注意到 cublasSgemmStridedBatched 的一些奇怪的表现,我正在寻找解释。矩阵大小固定为 20x20。以下是几种不同批量大小的一些时序(仅乘法,无数据传输):

  • 批处理 = 100,时间 = 0.2 毫秒
  • 批处理 = 1,000,时间 = 1.9 毫秒
  • 批处理 = 10,000,时间 = 18.3 毫秒
  • 批处理 = 100,000,时间 = 5.3 毫秒
  • 批处理 = 1,000,000,时间 = 52.8 毫秒

前几个批量大小正如我所期望的那样,随着批量大小增加十倍,时间线性增加。然而,使用 100,000 个矩阵突然会出现 3.4 倍的加速吗?

如果矩阵大小固定为10x10并且再次执行试验我发现:

  • 批处理 = 100,时间 = 0.2 毫秒
  • 批处理 = 1,000,时间 = 2.0 毫秒
  • 批处理 = 10,000,时间 = 20.0 毫秒
  • 批处理 = 100,000,时间 = 0.9 毫秒
  • 批处理 = 1,000,000,时间 = 8.9 毫秒

同样,批量大小为 100,000 时速度竟然意外提高了 22 倍?让我想知道为什么批量大小为 1,000 和 10,000 比批量大小 100,000 慢,因为矩阵大小仍然是 10x10。

不同的批量大小是否使用不同的算法?这个表现我觉得很奇怪。当我使用 cublasSgemmBatched 进行此试验时,会发生类似的结果。这些试验是在 GeForce GTX 1080 Ti 上执行的。授予最少的工作代码:

#include <stdio.h>
#include <stdlib.h>
#include "math.h"
#include "cublas_v2.h"
//nvcc -lcublas cublas.c -o cublas.out

int main(int argc, char* argv[])
{
int i,j,k,index;

// Linear dimension of matrices
int dim = 20;
int batch_count = 10*10*10*10*10*1;
// Allocate host storage for batch_count A,B,C square matrices
float* h_A = malloc(sizeof(float) * dim * dim * batch_count);
float* h_B = malloc(sizeof(float) * dim * dim * batch_count);
float* h_C = malloc(sizeof(float) * dim * dim * batch_count);
for(k=0; k<batch_count; k++) {
for(j=0; j<dim; j++) {
for(i=0; i<dim; i++) {
index = i*dim + j + k*dim*dim;
h_A[index] = index*index + 0.0f;
h_B[index] = index + 1.0f;
h_C[index] = 0.0f;
}
}
}


float *d_A, *d_B, *d_C;
cudaMalloc(&d_A, sizeof(float) * dim * dim * batch_count);
cudaMalloc(&d_B, sizeof(float) * dim * dim * batch_count);
cudaMalloc(&d_C, sizeof(float) * dim * dim * batch_count);
cudaMemcpy(h_A,d_A,sizeof(float) * dim * dim * batch_count,cudaMemcpyDeviceToHost);
cudaMemcpy(h_B,d_B,sizeof(float) * dim * dim * batch_count,cudaMemcpyDeviceToHost);
cudaMemcpy(h_C,d_C,sizeof(float) * dim * dim * batch_count,cudaMemcpyDeviceToHost);

cublasHandle_t handle;
cublasCreate(&handle);

// Do the actual multiplication
float time_cuda_event;
cudaEvent_t start, stop;
cudaEventCreate(&start);
cudaEventCreate(&stop) ;
cudaEventRecord(start, 0);
float alpha = 1.0f; float beta = 1.0f;
cublasSgemmStridedBatched(handle,
CUBLAS_OP_N,
CUBLAS_OP_N,
dim, dim, dim,
&alpha,
(const float*)d_A, dim,
dim*dim,
(const float*)d_B, dim,
dim*dim,
&beta,
d_C, dim,
dim*dim,
batch_count);
( cudaEventRecord(stop, 0) );
( cudaEventSynchronize(stop) );
( cudaEventElapsedTime(&time_cuda_event, start, stop) );
printf("Time : %3.1f ms \n", time_cuda_event);

cudaMemcpy(h_C,d_C,sizeof(float) * dim * dim * batch_count,cudaMemcpyDeviceToHost);
// Destroy the handle
cublasDestroy(handle);


cudaFree(d_A);
cudaFree(d_B);
cudaFree(d_C);
free(h_A);
free(h_B);
free(h_C);
return 0;
}

最佳答案

这似乎只是 CUBLAS 内部启发式的结果。如果我运行代码的修改(和工作)版本,我会得到 5x5 情况下的这些计时:

Batch size :           10   Time :  0.019104 ms 
Batch size : 100 Time : 0.038304 ms
Batch size : 1000 Time : 0.163520 ms
Batch size : 10000 Time : 1.410944 ms
Batch size : 100000 Time : 1.614144 ms
Batch size : 1000000 Time : 16.057407 ms

分析显示,在多达 10000 个条目的批处理的情况下,该库运行一个内核:

1.10759s  16.831us             (1 1 10)       (128 1 1)       120  12.250KB        0B         -           -           -           -  GeForce GTX 970         1         7  maxwell_sgemm_128x64_nn [3939]
1.10766s 19.168us (1 1 100) (128 1 1) 120 12.250KB 0B - - - - GeForce GTX 970 1 7 maxwell_sgemm_128x64_nn [3971]
1.10773s 147.71us (1 1 1000) (128 1 1) 120 12.250KB 0B - - - - GeForce GTX 970 1 7 maxwell_sgemm_128x64_nn [4003]
1.10791s 1.4064ms (1 1 10000) (128 1 1) 120 12.250KB 0B - - - - GeForce GTX 970 1 7 maxwell_sgemm_128x64_nn [4035]

当尺寸较大时,它会运行对另一个内核的多次调用来为调用提供服务:

1.10935s  1.1518ms          (1 1 65535)       (16 16 1)        31  2.1250KB        0B         -           -           -           -  GeForce GTX 970         1         7  void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4063]
1.11050s 606.54us (1 1 34465) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4087]
1.11113s 1.1498ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4115]
1.11228s 1.1501ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4139]
1.11344s 1.1511ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4163]
1.11459s 1.1494ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4187]
1.11574s 1.1507ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4211]
1.11689s 1.1503ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4235]
1.11804s 1.1499ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4259]
1.11919s 1.1507ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4283]
1.12035s 1.1507ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4307]
1.12150s 1.1509ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4331]
1.12265s 1.1489ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4355]
1.12380s 1.1496ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4379]
1.12495s 1.1500ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4403]
1.12610s 1.1494ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4427]
1.12726s 1.1503ms (1 1 65535) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4451]
1.12841s 299.35us (1 1 16975) (16 16 1) 31 2.1250KB 0B - - - - GeForce GTX 970 1 7 void batch_gemm_kernel1x1_core<float, float, float, bool=0, bool=0, bool=0, bool=0, bool=0, bool=1, bool=1>(float* const *, float const * const *, float const * const *, float*, float const *, float const *, int, int, int, int, int, int, __int64, __int64, __int64, float const *, float const *, float, float, int, int) [4475]

您观察到的不一致似乎是由库内从一个内核到另一个内核的更改引起的,这可能是由某些批量大小标准造成的。您可以看到,两个内核似乎每个批处理项都使用一个 block ,较大尺寸的内核使用具有 256 个线程的 2D block ,而较小尺寸的内核使用具有 128 个线程的 1D block 。除此之外,性能差异取决于内部实现细节。尽管这样做可能违反最终用户许可,但如果您想了解更多信息,您将需要反汇编内核并查看它们是如何工作的。该工具包包含执行此操作所需的所有工具,但我并不建议您这样做。

关于cuda - 奇怪的 cuBLAS gemm 批处理性能,我们在Stack Overflow上找到一个类似的问题: https://stackoverflow.com/questions/48519861/

27 4 0
Copyright 2021 - 2024 cfsdn All Rights Reserved 蜀ICP备2022000587号
广告合作:1813099741@qq.com 6ren.com