diff --git a/examples/post_training_quantization/torch/mobilenet_v2/main.py b/examples/post_training_quantization/torch/mobilenet_v2/main.py
index ad984143eec..898ac3fb34d 100644
--- a/examples/post_training_quantization/torch/mobilenet_v2/main.py
+++ b/examples/post_training_quantization/torch/mobilenet_v2/main.py
@@ -9,7 +9,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import os
 import re
 import subprocess
 from functools import partial
@@ -35,14 +34,6 @@
 DATASET_CLASSES = 10
 
 
-def set_cuda_arch_list():
-    if "TORCH_CUDA_ARCH_LIST" not in os.environ:
-        arch_list = torch.cuda.get_arch_list()
-        formatted_arch_list = [str(int(arch.split("_")[1]) / 10.0) for arch in arch_list]
-        arch_string = ";".join(formatted_arch_list)
-        os.environ["TORCH_CUDA_ARCH_LIST"] = arch_string
-
-
 def download_dataset() -> Path:
     downloader = FastDownload(base=DATASET_PATH.resolve(), archive="downloaded", data="extracted")
     return downloader.get(DATASET_URL)
@@ -103,8 +94,6 @@ def get_model_size(ir_path: Path, m_type: str = "Mb") -> float:
 
 ###############################################################################
 # Create a PyTorch model and dataset
-set_cuda_arch_list()
-
 dataset_path = download_dataset()
 
 normalize = transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
diff --git a/nncf/torch/extensions/__init__.py b/nncf/torch/extensions/__init__.py
index 59edea1ca24..81cd90af8a3 100644
--- a/nncf/torch/extensions/__init__.py
+++ b/nncf/torch/extensions/__init__.py
@@ -12,6 +12,7 @@
 import enum
 import os
 import textwrap
+import warnings
 from abc import ABC
 from abc import abstractmethod
 from multiprocessing.context import TimeoutError as MPTimeoutError
@@ -47,6 +48,14 @@ def get_build_directory_for_extension(name: str) -> Path:
     return build_dir
 
 
+def set_cuda_arch_list():
+    if "TORCH_CUDA_ARCH_LIST" not in os.environ:
+        arch_list = torch.cuda.get_arch_list()
+        formatted_arch_list = [str(int(arch.split("_")[1]) / 10.0) for arch in arch_list]
+        arch_string = ";".join(formatted_arch_list)
+        os.environ["TORCH_CUDA_ARCH_LIST"] = arch_string
+
+
 class ExtensionLoader(ABC):
     @classmethod
     @abstractmethod
@@ -98,9 +107,13 @@ def get(self, fn_name: str) -> Callable:
 
             with extension_is_loading_info_log(self._loader.name()):
                 try:
-                    pool = ThreadPool(processes=1)
-                    async_result = pool.apply_async(self._loader.load)
-                    self._loaded_namespace = async_result.get(timeout=timeout)
+                    with warnings.catch_warnings():
+                        warnings.filterwarnings("ignore", message="TORCH_CUDA_ARCH_LIST is not set")
+                        set_cuda_arch_list()
+
+                        pool = ThreadPool(processes=1)
+                        async_result = pool.apply_async(self._loader.load)
+                        self._loaded_namespace = async_result.get(timeout=timeout)
                 except MPTimeoutError as error:
                     msg = textwrap.dedent(
                         f"""\