diff --git a/ggml/src/ggml-openvino.cpp b/ggml/src/ggml-openvino.cpp
index c25a927c30..c33e3f2be0 100644
--- a/ggml/src/ggml-openvino.cpp
+++ b/ggml/src/ggml-openvino.cpp
@@ -487,7 +487,7 @@ static bool ggml_backend_openvino_device_supports_op(ggml_backend_dev_t dev, con
         case GGML_OP_MUL_MAT:
             return false;
         default:
-            return false;
+            return true;
     }
 }
 
diff --git a/ggml/src/ggml-openvino/ggml-ov-frontend-utils.cpp b/ggml/src/ggml-openvino/ggml-ov-frontend-utils.cpp
new file mode 100644
index 0000000000..f1b865aacf
--- /dev/null
+++ b/ggml/src/ggml-openvino/ggml-ov-frontend-utils.cpp
@@ -0,0 +1,54 @@
+#include "ggml-ov-frontend-utils.h"
+#include "ggml-backend-impl.h"
+#include <openvino/frontend/manager.hpp>
+
+using ov::frontend::tensorflow::ggml::GgmlOvGraphIterator;
+
+std::shared_ptr<GgmlOvGraphIterator> get_ggml_graph_iterator(struct ggml_cgraph * cgraph) {
+    return std::make_shared<GgmlOvGraphIterator>(cgraph);
+}
+
+static ov::frontend::FrontEnd::Ptr get_ggml_frontend() {
+    ov::frontend::FrontEnd::Ptr front_end = nullptr;
+    auto fem = ov::frontend::FrontEndManager();
+    std::string fe_so_path = "/home/yumeng/Code/ov-ggml-frontend/openvino/bin/intel64/Release/libopenvino_ggml_frontend.so";
+    fem.register_front_end("ggml", fe_so_path);
+    front_end = fem.load_by_framework("ggml");
+    return front_end;
+}
+
+enum ggml_status openvino_frontend_compute (ggml_backend_t backend, struct ggml_cgraph * cgraph) {
+    // Get GGML Frontend 
+    auto front_end = get_ggml_frontend();
+    if (!front_end) {
+        GGML_LOG_ERROR("GGML FrontEnd is not initialized \n");
+        return GGML_STATUS_FAILED;
+    } else {
+        GGML_LOG_ERROR("GGML FrontEnd is initialized \n");
+    }
+
+    auto ggml_graph_iterator = get_ggml_graph_iterator(cgraph);
+    std::shared_ptr<ov::frontend::tensorflow::GraphIterator> graph_iterator = ggml_graph_iterator;
+    GGML_LOG_ERROR("Decoder count in current GraphIterator: ");
+    GGML_LOG_ERROR(std::to_string(graph_iterator->size()).c_str());
+    
+    // Load GraphIterator -> InputModel
+    ov::frontend::InputModel::Ptr input_model = front_end->load(graph_iterator);
+    if (!input_model) {
+        GGML_LOG_ERROR("\nInput Model is not loaded \n");
+        return GGML_STATUS_FAILED;
+    } else {
+        GGML_LOG_ERROR("\nInput Model loaded \n");
+    }
+
+    // TODO: Convert InputModel -> ov::Model 
+    // std::shared_ptr<ov::Model> model = front_end->convert(input_model);
+    // if (!model) {
+    //     GGML_LOG_ERROR("Model is not converted");
+    // }
+
+    // TODO: Compute
+
+    return GGML_STATUS_SUCCESS;
+    GGML_UNUSED(backend);
+}
diff --git a/ggml/src/ggml-openvino/ggml-ov-frontend-utils.h b/ggml/src/ggml-openvino/ggml-ov-frontend-utils.h
new file mode 100644
index 0000000000..15dd46ed4e
--- /dev/null
+++ b/ggml/src/ggml-openvino/ggml-ov-frontend-utils.h
@@ -0,0 +1,6 @@
+#include "ggml-graph-iterator.h"
+#include "ggml-backend-impl.h"
+
+std::shared_ptr<ov::frontend::tensorflow::ggml::GgmlOvGraphIterator> get_ggml_graph_iterator(struct ggml_cgraph * cgraph);
+
+enum ggml_status openvino_frontend_compute (ggml_backend_t backend, struct ggml_cgraph * cgraph);