Device Management
[CUDA Runtime API]


 Thread Management [DEPRECATED]


cudaError_t cudaChooseDevice (int *device, const struct cudaDeviceProp *prop)
 Select compute-device which best matches criteria.
cudaError_t cudaDeviceGetCacheConfig (enum cudaFuncCache *pCacheConfig)
 Returns the preferred cache configuration for the current device.
cudaError_t cudaDeviceGetLimit (size_t *pValue, enum cudaLimit limit)
 Returns resource limits.
cudaError_t cudaDeviceReset (void)
 Destroy all allocations and reset all state on the current device in the current process.
cudaError_t cudaDeviceSetCacheConfig (enum cudaFuncCache cacheConfig)
 Sets the preferred cache configuration for the current device.
cudaError_t cudaDeviceSetLimit (enum cudaLimit limit, size_t value)
 Set resource limits.
cudaError_t cudaDeviceSynchronize (void)
 Wait for compute device to finish.
cudaError_t cudaGetDevice (int *device)
 Returns which device is currently being used.
cudaError_t cudaGetDeviceCount (int *count)
 Returns the number of compute-capable devices.
cudaError_t cudaGetDeviceProperties (struct cudaDeviceProp *prop, int device)
 Returns information about the compute-device.
cudaError_t cudaSetDevice (int device)
 Set device to be used for GPU executions.
cudaError_t cudaSetDeviceFlags (unsigned int flags)
 Sets flags to be used for device executions.
cudaError_t cudaSetValidDevices (int *device_arr, int len)
 Set a list of devices that can be used for CUDA.

Detailed Description

This section describes the device management functions of the CUDA runtime application programming interface.

Generated by Doxygen for NVIDIA CUDA Library  NVIDIA