I want to use the atomicMin() function in my kernel fuc. In my CUDA Build Rule (2.3.0) i saw the default options are
“echo “C:\CUDA\bin64\nvcc.exe” -arch sm_10 -ccbin …”
So i add -arch sm_11 to the additional options. But when I compile my codes, i get still the error
error: identifier “atomicMin” is undefined.