diff --git a/mindspore/ccsrc/backend/kernel_compiler/gpu/math/broadcast_gpu_kernel.cc b/mindspore/ccsrc/backend/kernel_compiler/gpu/math/broadcast_gpu_kernel.cc
index 6109f72e8e6d132f1f26d5efeec0486bcae600ab..850a1c7da43175cc608b89d00826cfbb30d7e2fa 100644
--- a/mindspore/ccsrc/backend/kernel_compiler/gpu/math/broadcast_gpu_kernel.cc
+++ b/mindspore/ccsrc/backend/kernel_compiler/gpu/math/broadcast_gpu_kernel.cc
@@ -109,6 +109,9 @@ MS_REG_GPU_KERNEL_ONE(
   BroadcastOpGpuKernel, half)
 
 // int32
+MS_REG_GPU_KERNEL_ONE(
+  Greater, KernelAttr().AddInputAttr(kNumberTypeInt32).AddInputAttr(kNumberTypeInt32).AddOutputAttr(kNumberTypeBool),
+  BroadcastOpGpuKernel, int)
 MS_REG_GPU_KERNEL_ONE(
   Less, KernelAttr().AddInputAttr(kNumberTypeInt32).AddInputAttr(kNumberTypeInt32).AddOutputAttr(kNumberTypeBool),
   BroadcastOpGpuKernel, int)
diff --git a/tests/st/ops/gpu/test_broadcast_op.py b/tests/st/ops/gpu/test_broadcast_op.py
index 4681b608b50f0ceb42b478cbf15b4978b09eeda5..6de4c2b1e23c33d68a3033dce2af9da1f33dd2da 100644
--- a/tests/st/ops/gpu/test_broadcast_op.py
+++ b/tests/st/ops/gpu/test_broadcast_op.py
@@ -43,6 +43,9 @@ def test_nobroadcast():
     output_ms = P.Greater()(Tensor(x1_np), Tensor(x2_np))
     output_np = x1_np > x2_np
     assert np.allclose(output_ms.asnumpy(), output_np)
+    output_ms = P.Greater()(Tensor(x1_np_int32), Tensor(x2_np_int32))
+    output_np = x1_np_int32 > x2_np_int32
+    assert np.allclose(output_ms.asnumpy(), output_np)
 
     output_ms = P.Less()(Tensor(x1_np), Tensor(x2_np))
     output_np = x1_np < x2_np
@@ -132,6 +135,9 @@ def test_broadcast():
     output_ms = P.Greater()(Tensor(x1_np), Tensor(x2_np))
     output_np = x1_np > x2_np
     assert np.allclose(output_ms.asnumpy(), output_np)
+    output_ms = P.Greater()(Tensor(x1_np_int32), Tensor(x2_np_int32))
+    output_np = x1_np_int32 > x2_np_int32
+    assert np.allclose(output_ms.asnumpy(), output_np)
 
     output_ms = P.Less()(Tensor(x1_np), Tensor(x2_np))
     output_np = x1_np < x2_np
@@ -175,6 +181,9 @@ def test_broadcast_diff_dims():
     output_ms = P.Maximum()(Tensor(x1_np), Tensor(x2_np))
     output_np = np.maximum(x1_np, x2_np)
     assert np.allclose(output_ms.asnumpy(), output_np)
+    output_ms = P.Greater()(Tensor(x1_np_int32), Tensor(x2_np_int32))
+    output_np = x1_np_int32 > x2_np_int32
+    assert np.allclose(output_ms.asnumpy(), output_np)
 
     output_ms = P.Greater()(Tensor(x1_np), Tensor(x2_np))
     output_np = x1_np > x2_np