Reputation: 1914
Many framework require that your nvidia graphic card has a specific compute capability version.
I am developing a C++ application that uses Cuda. I should get this information by code. so that I can assign the needed framework for each graphic compute capability. How to know the compute capability of my nvidia graphic in C/C++?
Upvotes: 3
Views: 1698
Reputation:
From the CUDA Runtime API
__host__ cudaError_t cudaGetDeviceProperties ( cudaDeviceProp* prop, int device )
Returns information about the compute-device.
Alternatively, you could use cudaDeviceGetAttribute
to get the specific properties you want.
precisely: Returns in *prop the properties of device dev. The cudaDeviceProp structure is defined as:
struct cudaDeviceProp {
....
int major;
int minor;
.....
}
major, minor are the major and minor revision numbers defining the device's compute capability.
Upvotes: 6
Reputation: 1672
If you right click the nvidia icon on the bottom right of your screen then go to Nvidia control panel, a window will open then you can check all of the specs under "System Information" it's in the bottom left of the window .
Upvotes: 0