diff --git a/flang/module/cudadevice.f90 b/flang/module/cudadevice.f90
index 28577ee495368..a7f7652f5e52c 100644
--- a/flang/module/cudadevice.f90
+++ b/flang/module/cudadevice.f90
@@ -590,8 +590,36 @@ attributes(device) real(8) function sinpi(x) bind(c,name='__nv_sinpi')
     end function
   end interface
 
+  interface __uint2float_rd
+    attributes(device) real function __uint2float_rd(i) bind(c, name='__nv_uint2float_rd')
+      !dir$ ignore_tkr (d) i
+      integer, value :: i
+    end function
+  end interface
+
+  interface __uint2float_rn
+    attributes(device) real function __uint2float_rn(i) bind(c, name='__nv_uint2float_rn')
+      !dir$ ignore_tkr (d) i
+      integer, value :: i
+    end function
+  end interface
+
+  interface __uint2float_ru
+    attributes(device) real function __uint2float_ru(i) bind(c, name='__nv_uint2float_ru')
+      !dir$ ignore_tkr (d) i
+      integer, value :: i
+    end function
+  end interface
+
+  interface __uint2float_rz
+    attributes(device) real function __uint2float_rz(i) bind(c, name='__nv_uint2float_rz')
+      !dir$ ignore_tkr (d) i
+      integer, value :: i
+    end function
+  end interface
+
   interface __uint2double_rn
-    attributes(device) double precision function __uint2double_rn(i) bind(c)
+    attributes(device) double precision function __uint2double_rn(i) bind(c, name='__nv_uint2double_rn')
       !dir$ ignore_tkr (d) i
       integer, value :: i
     end function
diff --git a/flang/test/Lower/CUDA/cuda-libdevice.cuf b/flang/test/Lower/CUDA/cuda-libdevice.cuf
index 06515b1c5f6df..d2cbadf9c05d8 100644
--- a/flang/test/Lower/CUDA/cuda-libdevice.cuf
+++ b/flang/test/Lower/CUDA/cuda-libdevice.cuf
@@ -20,6 +20,30 @@ end subroutine
 ! CHECK-LABEL: _QPtest_usad
 ! CHECK: %{{.*}} = fir.call @__nv_usad(%{{.*}}, %{{.*}}, %{{.*}}) proc_attrs<bind_c> fastmath<contract> : (i32, i32, i32) -> i32
 
+attributes(global) subroutine test_uint2float_rX()
+  real :: res
+  integer :: i
+  res = __uint2float_rd(i)
+  res = __uint2float_rn(i)
+  res = __uint2float_ru(i)
+  res = __uint2float_rz(i)
+end subroutine
+
+! CHECK-LABEL: _QPtest_uint2float_rx
+! CHECK: %{{.*}} = fir.call @__nv_uint2float_rd(%{{.*}}) proc_attrs<bind_c> fastmath<contract> : (i32) -> f32
+! CHECK: %{{.*}} = fir.call @__nv_uint2float_rn(%{{.*}}) proc_attrs<bind_c> fastmath<contract> : (i32) -> f32
+! CHECK: %{{.*}} = fir.call @__nv_uint2float_ru(%{{.*}}) proc_attrs<bind_c> fastmath<contract> : (i32) -> f32
+! CHECK: %{{.*}} = fir.call @__nv_uint2float_rz(%{{.*}}) proc_attrs<bind_c> fastmath<contract> : (i32) -> f32
+
+attributes(global) subroutine test_uint2double_rn()
+  double precision :: res
+  integer :: i
+  res = __uint2double_rn(i)
+end subroutine
+
+! CHECK-LABEL: _QPtest_uint2double_rn
+! CHECK: %{{.*}} = fir.call @__nv_uint2double_rn(%{{.*}}) proc_attrs<bind_c> fastmath<contract> : (i32) -> f64
+
 attributes(global) subroutine test_ull2dloat_rX()
   real :: res
   integer(8) :: i