diff --git a/src/cambricon/impl/magicmind_runtime_opr.cpp b/src/cambricon/impl/magicmind_runtime_opr.cpp
index c25b0b8d0e71c2e86d1a88d27bc045e88448419a..7bf5c534232279b0cf4d47f119fd5a26ccec42a9 100644
--- a/src/cambricon/impl/magicmind_runtime_opr.cpp
+++ b/src/cambricon/impl/magicmind_runtime_opr.cpp
@@ -168,7 +168,8 @@ MagicMindRuntimeOpr::MagicMindRuntimeOpr(
           m_allocator{std::move(allocator)},
           m_engine{nullptr},
           m_context{nullptr},
-          m_model{std::move(model)} {
+          m_model{std::move(model)}, 
+          m_current_ptr{nullptr} {
     mgb_assert(
             inputs[0]->comp_node().device_type() == CompNode::DeviceType::CAMBRICON,
             "MagicMindRuntimeOpr can only be used on cambricon comp node; "
@@ -230,8 +231,18 @@ void MagicMindRuntimeOpr::scn_do_execute() {
         MM_CHECK(tensor->SetDimensions(mgb_shape_to_mm_dims(output(i)->shape())));
         MM_CHECK(tensor->SetData(output(i)->dev_tensor().raw_ptr()));
     }
-    auto size = output().back()->dev_tensor().layout().span().dist_byte();
-    MM_CHECK(m_context->SetWorkspace(output().back()->dev_tensor().raw_ptr(), size));
+    if (m_current_ptr == nullptr) {
+        auto size = output().back()->dev_tensor().layout().span().dist_byte();
+        m_current_ptr = output().back()->dev_tensor().raw_ptr();
+        MM_CHECK(m_context->SetWorkspace(m_current_ptr, size));
+    } else {
+        auto current_ptr = output().back()->dev_tensor().raw_ptr();
+        mgb_assert(
+                current_ptr == m_current_ptr,
+                "workspace has been changed, the execution context should be "
+                "reconstructed, but now this is not supported (got:%p,prev:%p)",
+                current_ptr, m_current_ptr);
+    }
     MM_CHECK(m_context->Enqueue(inputs, outputs, cnrt_env.queue));
     for (auto&& i : inputs) {
         i->Destroy();
@@ -293,7 +304,6 @@ void MagicMindRuntimeOpr::get_output_var_shape(
                 false, "static shape infer for MagicMindRuntimeOpr(%s) failed",
                 cname());
     }
-    return;
     for (auto&& i : inputs) {
         i->Destroy();
     }
diff --git a/src/cambricon/include/megbrain/cambricon/magicmind_runtime_opr.h b/src/cambricon/include/megbrain/cambricon/magicmind_runtime_opr.h
index b72cf87162679f36dcabe81ef029aa50eb9b14fc..0fd8a318f0f904e4a03cd12e9873d10f8d3483b7 100644
--- a/src/cambricon/include/megbrain/cambricon/magicmind_runtime_opr.h
+++ b/src/cambricon/include/megbrain/cambricon/magicmind_runtime_opr.h
@@ -93,6 +93,7 @@ private:
     IEnginePtr m_engine;
     mutable IContextPtr m_context;
     IModelPtr m_model;
+    dt_byte* m_current_ptr;
 };
 
 }  // namespace opr
diff --git a/src/cambricon/test/magicmind_runtime_opr.cpp b/src/cambricon/test/magicmind_runtime_opr.cpp
index 4a7a446041f39f6e6bfc4b473ceaf1f3246e9f50..7ddeb1117e4340bcd377850db3452d58ee267b07 100644
--- a/src/cambricon/test/magicmind_runtime_opr.cpp
+++ b/src/cambricon/test/magicmind_runtime_opr.cpp
@@ -642,6 +642,7 @@ TEST(TestMagicMindRuntimeOpr, GraphShapeMutable) {
         auto func = graph->compile(
                 {make_callback_copy(out1, o1), make_callback_copy(out2, o2)});
         func->execute();
+        func->execute();
         HostTensorND o1_mm(cn, mkshp(no, co, ho, wo), dtype::Float32()),
                 o2_mm(cn, mkshp(no, co, ho, wo), dtype::Float32());
         std::memcpy(