This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
Try to disable IPP (-DWITH_IPP=OFF) and limit used CUDA architectures (-DCUDA_ARCH_BIN=...?): IPP contains multiple optimized versions of same function for different platforms and, I guess, problem with CUDA can be similar.
Try to enable link time optimization, we've added ENABLE_LTO cmake option and it should be supported by VS (adds /GL compiler option), but we haven't tested it.
We have related advice, but it is mostly for Linux: https://github.com/opencv/opencv/wiki/Compact-build-advice
| | 2 | No.2 Revision |
Try to disable IPP (-DWITH_IPP=OFF) and limit used CUDA architectures (-DCUDA_ARCH_BIN=...?): IPP contains multiple optimized versions of same function for different platforms and, I guess, problem with CUDA can be similar.similar. Probably you should specify CUDA_GENERATION option: https://github.com/opencv/opencv/blob/b143f7100a112a221baf80b7c29a631962ee3a64/cmake/OpenCVDetectCUDA.cmake#L44-L86
Try to enable link time optimization, we've added ENABLE_LTO cmake option and it should be supported by VS (adds /GL compiler option), but we haven't tested it.
We have related advice, but it is mostly for Linux: https://github.com/opencv/opencv/wiki/Compact-build-advice
