CUDA > 時間計測をする実装 > QueryPerformanceCounter()使用

動作環境

GeForce GTX 750 Ti
Windows 8.1 pro (64bit)
Visual Studio Community 2013 (以下、VS)
CUDA 7.5

CPUでの処理とGPUでの処理を比較しようと思っている。
それには時間計測が必要だ。

情報

stackoverflowに例があった。
http://stackoverflow.com/questions/14337278/precise-time-measurement

answered Jan 15 '13 at 12:04
Constantinius

によるコード

準備

http://qiita.com/7of9/items/e9c185351803d9794db5
を参考にCUDAをビルドできるslnを用意する (160910_mreasureTimeとした)。

code > cudaDeviceReset()の時間計測

LARGE_INTEGERを使うにはWindows.hをincludeする必要あり。

kernel.cu

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

# include <Windows.h> // LARGE_INTEGERなどに必要
# include <stdio.h>

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

int main()
{
	cudaError_t cudaStatus;

	LARGE_INTEGER frequency;        // ticks per second
	LARGE_INTEGER t1, t2;           // ticks
	double elapsedTime;

	QueryPerformanceFrequency(&frequency); // get ticks per second
	QueryPerformanceCounter(&t1); // start timer

	// { process to measure ---------
	cudaStatus = cudaDeviceReset();
    if (cudaStatus != cudaSuccess) {
        fprintf(stderr, "cudaDeviceReset failed!");
        return 1;
    }
	// } process to measure ---------

	QueryPerformanceCounter(&t2); // stop timer

	// compute and print the elapsed time in millisec
	elapsedTime = (t2.QuadPart - t1.QuadPart) * 1000.0 / frequency.QuadPart;

	printf("%.1f msec\n", elapsedTime);

    return 0;
}

結果

C:\CudaDev\160910_measureTime\Debug>160910_measureTime.exe
13.5 msec

こちらの環境ではcudaDeviceReset()にはだいたい13msecかかる。

今後はGPUを使う計算を実装し、それをCPU処理と比較する。

code > GPU処理の時間計測

CUDA 7.5 RuntimeのサンプルコードでのGPU処理の時間を計測してみた。

kernel.cu

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

# include <Windows.h> // LARGE_INTEGERなどに必要
# 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 };

	LARGE_INTEGER frequency;        // ticks per second
	LARGE_INTEGER t1, t2;           // ticks
	double elapsedTime;

	QueryPerformanceFrequency(&frequency); // get ticks per second
	QueryPerformanceCounter(&t1); // start timer

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

	QueryPerformanceCounter(&t2); // stop timer
	// compute and print the elapsed time in millisec
	elapsedTime = (t2.QuadPart - t1.QuadPart) * 1000.0 / frequency.QuadPart;

	printf("%.1f msec\n", elapsedTime);

    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;
}

結果について以下の記事とした。
http://qiita.com/7of9/items/70b21e72c3cb9a094f2d

実行の度に処理時間が減っていくのが未消化。

同じ処理をCPUで実行すると0.0msecとなった。処理が速すぎてmsecでは追えない。
一方でGPUの場合はオーバーヘッドが結果として出ているのかもしれない。