许多 CUDA 示例失败了

Question

安装新的 CUDA 4.0 驱动程序和 SDK 后，许多 SDK 测试失败（例如

fastWalshTransform, matrixMul, reduction

）。这就是

./deviceQuery

:

Device 0: "GeForce GTX 570"
  CUDA Driver Version / Runtime Version          4.0 / 4.0
  CUDA Capability Major/Minor version number:    2.0
  Total amount of global memory:                 1279 MBytes (1341325312 bytes)
  (15) Multiprocessors x (32) CUDA Cores/MP:     480 CUDA Cores
  GPU Clock Speed:                               1.57 GHz
  Memory Clock rate:                             2100.00 Mhz
  Memory Bus Width:                              320-bit
  L2 Cache Size:                                 655360 bytes
  Max Texture Dimension Size (x,y,z)             1D=(65536), 2D=(65536,65535), 3D=(2048,2048,2048)
  Max Layered Texture Size (dim) x layers        1D=(16384) x 2048, 2D=(16384,16384) x 2048
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       49152 bytes
  Total number of registers available per block: 32768
  Warp size:                                     32
  Maximum number of threads per block:           1024
  Maximum sizes of each dimension of a block:    1024 x 1024 x 64
  Maximum sizes of each dimension of a grid:     65535 x 65535 x 65535
  Maximum memory pitch:                          2147483647 bytes
  Texture alignment:                             512 bytes
  Concurrent copy and execution:                 Yes with 1 copy engine(s)
  Run time limit on kernels:                     No
  Integrated GPU sharing Host Memory:            No
  Support host page-locked memory mapping:       Yes
  Concurrent kernel execution:                   Yes
  Alignment requirement for Surfaces:            Yes
  Device has ECC support enabled:                No
  Device is using TCC driver mode:               No
  Device supports Unified Addressing (UVA):      Yes
  Device PCI Bus ID / PCI location ID:           4 / 0

例如

reduction

的输出是：

• GPU 结果 = 2135772699
• CPU 结果 = 2139353471

=>

FAILED

.

解决方案：它曾经是（现在仍然是）硬件问题（驱动程序更新不能解决问题）。也许是一些内存问题，但很常见。我们有几张 NVIDIA 卡显示了该问题（甚至 Tesla！）。到目前为止，我们找到的唯一解决方案是重新启动机器或稍微提高电压。

Answer 1

这曾经是（现在仍然是）硬件问题（驱动程序更新不能解决问题）。也许是一些内存问题，但很常见。我们有几张 NVIDIA 卡显示了该问题（甚至 Tesla！）。到目前为止，我们找到的唯一解决方案是重新启动机器或稍微提高电压。

Answer 2

尝试这个代码，也许它有效。

#include <iostream>
#include <vector>
#include "cuda_runtime.h"
#include "device_launch_parameters.h"

#define threads 10
#define numbers 20

__global__ void calculate(int* rez, int* max) {
    int idx = threadIdx.x + blockIdx.x * blockDim.x;
    int step = blockDim.x * gridDim.x;

    for (int i = idx; i < numbers; i += step) {
        rez[i] = i * i;
        atomicMax(max, rez[i]);
    }
}


int main(){
    int size = numbers * sizeof(int);
    int hostRez[numbers];
    int hostMax = 0;
    int* devRez;
    int* devMax;

    cudaMalloc((void**)&devRez, size);
    cudaMalloc((void**)&devMax, sizeof(int));

    cudaMemcpy(devMax, &hostMax, sizeof(int), cudaMemcpyHostToDevice);

    calculate << < 1, threads >> > (devRez, devMax);

    cudaMemcpy(&hostMax, devMax, sizeof(int), cudaMemcpyDeviceToHost);
    cudaMemcpy(&hostRez, devRez, size, cudaMemcpyDeviceToHost);
    for (int i = 0; i < numbers; i++) {
        std::cout << hostRez[i] << std::endl;
    }

    std::cout << hostMax << std::endl;

    cudaFree(devRez);
    cudaFree(devMax);

    return 0;
}