diff --git a/test/_inductor/test_expanded_reduction.py b/test/_inductor/test_expanded_reduction.py
new file mode 100644
index 0000000000000000000000000000000000000000..40e0363916c2d5920e7c57b75343421eea46683a
--- /dev/null
+++ b/test/_inductor/test_expanded_reduction.py
@@ -0,0 +1,27 @@
+import torch
+from torch.testing._internal.common_utils import run_tests, parametrize, instantiate_parametrized_tests
+from testutils import TestUtils
+import torch_npu
+
+
+class TestExpandedReduction(TestUtils):
+    def op_calc(self, first_element, second_element):
+        x = first_element * second_element
+        return x.sum((0, 1))
+
+    @parametrize('shape_x', [(2, 197, 256)])
+    @parametrize('shape_y', [(2, 1, 256)])
+    @parametrize('dtype', ['float16'])
+    def test_expanded_reduction_cases(self, shape_x, shape_y, dtype):
+        first_element = self._generate_tensor(shape_x, dtype)
+        second_element = self._generate_tensor(shape_y, dtype)
+        std_result = self.op_calc(first_element, second_element)
+
+        compiled_op_calc = torch.compile(self.op_calc, backend="inductor")
+        inductor_result = compiled_op_calc(first_element, second_element)
+        self.assertEqual(std_result, inductor_result, atol=1e-5, rtol=1e-5)
+
+instantiate_parametrized_tests(TestExpandedReduction)
+
+if __name__ == "__main__":
+    run_tests()
diff --git a/torch_npu/_inductor/codegen/split_tiling.py b/torch_npu/_inductor/codegen/split_tiling.py
index 782cc9f7455cd6d9f4eea63075768fdeb1af0690..9b94ce3c5cce240c3234e1088849ccbcd27b9fb2 100644
--- a/torch_npu/_inductor/codegen/split_tiling.py
+++ b/torch_npu/_inductor/codegen/split_tiling.py
@@ -141,6 +141,8 @@ class SplitTiling:
 
         def select_tiling(low_dim=True, reduction=True):
             for axis in reversed(self.kernel.sorted_axis):
+                if meet_stop_condition():
+                    break
                 if low_dim and axis.sorted_order in self.kernel.low_dims and axis not in self.kernel.tiling_axis:
                     axis.is_tiling_axis = True
                     self.kernel.tiling_axis.append(axis)