diff --git a/modules/core/include/opencv2/core/cuda/detail/reduce.hpp b/modules/core/include/opencv2/core/cuda/detail/reduce.hpp
index 0c35eaba65..44400c8e1d 100644
--- a/modules/core/include/opencv2/core/cuda/detail/reduce.hpp
+++ b/modules/core/include/opencv2/core/cuda/detail/reduce.hpp
@@ -275,7 +275,7 @@ namespace cv { namespace cuda { namespace device
             template <typename Pointer, typename Reference, class Op>
             static __device__ void reduce(Pointer smem, Reference val, unsigned int tid, Op op)
             {
-            #if __CUDA_ARCH__ >= 300
+            #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
                 (void) smem;
                 (void) tid;
 
@@ -298,7 +298,7 @@ namespace cv { namespace cuda { namespace device
             {
                 const unsigned int laneId = Warp::laneId();
 
-            #if __CUDA_ARCH__ >= 300
+            #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
                 Unroll<16, Pointer, Reference, Op>::loopShfl(val, op, warpSize);
 
                 if (laneId == 0)
@@ -321,7 +321,7 @@ namespace cv { namespace cuda { namespace device
 
                 if (tid < 32)
                 {
-                #if __CUDA_ARCH__ >= 300
+                #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
                     Unroll<M / 2, Pointer, Reference, Op>::loopShfl(val, op, M);
                 #else
                     Unroll<M / 2, Pointer, Reference, Op>::loop(smem, val, tid, op);
diff --git a/modules/core/include/opencv2/core/cuda/saturate_cast.hpp b/modules/core/include/opencv2/core/cuda/saturate_cast.hpp
index e7633c76a3..f55ae4f4a7 100644
--- a/modules/core/include/opencv2/core/cuda/saturate_cast.hpp
+++ b/modules/core/include/opencv2/core/cuda/saturate_cast.hpp
@@ -101,7 +101,7 @@ namespace cv { namespace cuda { namespace device
     }
     template<> __device__ __forceinline__ uchar saturate_cast<uchar>(double v)
     {
-    #if __CUDA_ARCH__ >= 130
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 130
         uint res = 0;
         asm("cvt.rni.sat.u8.f64 %0, %1;" : "=r"(res) : "d"(v));
         return res;
@@ -149,7 +149,7 @@ namespace cv { namespace cuda { namespace device
     }
     template<> __device__ __forceinline__ schar saturate_cast<schar>(double v)
     {
-    #if __CUDA_ARCH__ >= 130
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 130
         uint res = 0;
         asm("cvt.rni.sat.s8.f64 %0, %1;" : "=r"(res) : "d"(v));
         return res;
@@ -191,7 +191,7 @@ namespace cv { namespace cuda { namespace device
     }
     template<> __device__ __forceinline__ ushort saturate_cast<ushort>(double v)
     {
-    #if __CUDA_ARCH__ >= 130
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 130
         ushort res = 0;
         asm("cvt.rni.sat.u16.f64 %0, %1;" : "=h"(res) : "d"(v));
         return res;
@@ -226,7 +226,7 @@ namespace cv { namespace cuda { namespace device
     }
     template<> __device__ __forceinline__ short saturate_cast<short>(double v)
     {
-    #if __CUDA_ARCH__ >= 130
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 130
         short res = 0;
         asm("cvt.rni.sat.s16.f64 %0, %1;" : "=h"(res) : "d"(v));
         return res;
diff --git a/modules/core/include/opencv2/core/cuda/warp_shuffle.hpp b/modules/core/include/opencv2/core/cuda/warp_shuffle.hpp
index 5cf42ec41d..256fc2a684 100644
--- a/modules/core/include/opencv2/core/cuda/warp_shuffle.hpp
+++ b/modules/core/include/opencv2/core/cuda/warp_shuffle.hpp
@@ -54,7 +54,7 @@ namespace cv { namespace cuda { namespace device
     template <typename T>
     __device__ __forceinline__ T shfl(T val, int srcLane, int width = warpSize)
     {
-    #if __CUDA_ARCH__ >= 300
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
         return __shfl(val, srcLane, width);
     #else
         return T();
@@ -62,7 +62,7 @@ namespace cv { namespace cuda { namespace device
     }
     __device__ __forceinline__ unsigned int shfl(unsigned int val, int srcLane, int width = warpSize)
     {
-    #if __CUDA_ARCH__ >= 300
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
         return (unsigned int) __shfl((int) val, srcLane, width);
     #else
         return 0;
@@ -70,7 +70,7 @@ namespace cv { namespace cuda { namespace device
     }
     __device__ __forceinline__ double shfl(double val, int srcLane, int width = warpSize)
     {
-    #if __CUDA_ARCH__ >= 300
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
         int lo = __double2loint(val);
         int hi = __double2hiint(val);
 
@@ -86,7 +86,7 @@ namespace cv { namespace cuda { namespace device
     template <typename T>
     __device__ __forceinline__ T shfl_down(T val, unsigned int delta, int width = warpSize)
     {
-    #if __CUDA_ARCH__ >= 300
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
         return __shfl_down(val, delta, width);
     #else
         return T();
@@ -94,7 +94,7 @@ namespace cv { namespace cuda { namespace device
     }
     __device__ __forceinline__ unsigned int shfl_down(unsigned int val, unsigned int delta, int width = warpSize)
     {
-    #if __CUDA_ARCH__ >= 300
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
         return (unsigned int) __shfl_down((int) val, delta, width);
     #else
         return 0;
@@ -102,7 +102,7 @@ namespace cv { namespace cuda { namespace device
     }
     __device__ __forceinline__ double shfl_down(double val, unsigned int delta, int width = warpSize)
     {
-    #if __CUDA_ARCH__ >= 300
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
         int lo = __double2loint(val);
         int hi = __double2hiint(val);
 
@@ -118,7 +118,7 @@ namespace cv { namespace cuda { namespace device
     template <typename T>
     __device__ __forceinline__ T shfl_up(T val, unsigned int delta, int width = warpSize)
     {
-    #if __CUDA_ARCH__ >= 300
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
         return __shfl_up(val, delta, width);
     #else
         return T();
@@ -126,7 +126,7 @@ namespace cv { namespace cuda { namespace device
     }
     __device__ __forceinline__ unsigned int shfl_up(unsigned int val, unsigned int delta, int width = warpSize)
     {
-    #if __CUDA_ARCH__ >= 300
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
         return (unsigned int) __shfl_up((int) val, delta, width);
     #else
         return 0;
@@ -134,7 +134,7 @@ namespace cv { namespace cuda { namespace device
     }
     __device__ __forceinline__ double shfl_up(double val, unsigned int delta, int width = warpSize)
     {
-    #if __CUDA_ARCH__ >= 300
+    #if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
         int lo = __double2loint(val);
         int hi = __double2hiint(val);