No kernel image available for execution on the device

Recieved this error trying to run ( ) this code on the Jetson Nano. After googling, found out that the error is meant to occur on devices which are not supported.

The code given in the repository is written in CUDA 8, and Jetson Nano works with CUDA 10 (with no possibility of a backport I assume).

How do I go about this error? The line causing the error is this ( and the exact error says :

CUDA Error at RGBDImageCudaKernel.cuh:31 code=48(cudaErrorNoKernelImageForDevice) "cudaGetLastError()"

I think this has something to do with the fact that I am running CUDA 10 and the project has been implemented in CUDA 8.0. Kindly guide me towards the correct direction. If at all the difference of CUDA is the problem here, then kindly provide with any resources which might help me port the code to CUDA 10. I am fairly new to CUDA.

Hi, what the error likely means is that the code wasn’t compiled with sm_53 gencode flag using NVCC. So you may need to add that to the project’s Makefile/CMakeLists/ect.

Thank you for the quick and precise response!