@@ -146,8 +146,8 @@ def managed_array(shape, dtype=np.float_, strides=None, order='C', stream=0,
     Allocate a np.ndarray with a buffer that is managed.
     Similar to np.empty().
 
-    Managed memory is supported on Linux, and is considered experimental on
-    Windows.
+    Managed memory is supported on Linux / x86 and PowerPC, and is considered
+    experimental on Windows and Linux / AArch64.
 
     :param attach_global: A flag indicating whether to attach globally. Global
                           attachment implies that the memory is accessible from
 
@@ -3,12 +3,13 @@
 from numba.cuda.cudadrv.driver import device_memset, driver
 from numba import cuda
 from numba.cuda.testing import unittest, ContextResettingTestCase
-from numba.cuda.testing import skip_on_cudasim
+from numba.cuda.testing import skip_on_cudasim, skip_on_arm
 from numba.tests.support import linux_only
 
 
 @skip_on_cudasim('CUDA Driver API unsupported in the simulator')
 @linux_only
+@skip_on_arm('Managed Alloc support is experimental/untested on ARM')
 class TestManagedAlloc(ContextResettingTestCase):
 
     def get_total_gpu_memory(self):