diff --git a/test/decoders/test_video_decoder.py b/test/decoders/test_video_decoder.py
index 2fcd324d..4632e5ff 100644
--- a/test/decoders/test_video_decoder.py
+++ b/test/decoders/test_video_decoder.py
@@ -443,14 +443,14 @@ def test_get_frame_played_at(self, device):
         assert isinstance(decoder.get_frame_played_at(6.02).pts_seconds, float)
         assert isinstance(decoder.get_frame_played_at(6.02).duration_seconds, float)
 
-    def test_get_frame_played_at_h265(self):
+    @pytest.mark.parametrize("device", cpu_and_cuda())
+    def test_get_frame_played_at_h265(self, device):
         # Non-regression test for https://github.com/pytorch/torchcodec/issues/179
-        # We don't parametrize with CUDA because the current GPUs on CI do not
-        # support x265:
-        # https://github.com/pytorch/torchcodec/pull/350#issuecomment-2465011730
-        decoder = VideoDecoder(H265_VIDEO.path)
+        decoder = VideoDecoder(H265_VIDEO.path, device=device)
         ref_frame6 = H265_VIDEO.get_frame_data_by_index(5)
-        assert_frames_equal(ref_frame6, decoder.get_frame_played_at(0.5).data)
+        assert_frames_equal(
+            ref_frame6.to(device=device), decoder.get_frame_played_at(0.5).data
+        )
 
     @pytest.mark.parametrize("device", cpu_and_cuda())
     def test_get_frame_played_at_fails(self, device):