diff --git a/openfpga/src/fabric/build_top_module_child_fine_grained_instance.cpp b/openfpga/src/fabric/build_top_module_child_fine_grained_instance.cpp
index 97152ebb4..47067a2cb 100644
--- a/openfpga/src/fabric/build_top_module_child_fine_grained_instance.cpp
+++ b/openfpga/src/fabric/build_top_module_child_fine_grained_instance.cpp
@@ -390,7 +390,10 @@ static void add_top_module_io_children(
   }
   /* If width is odd, add the missing vertical line */
   if ((grids.width() - 2) % 2 == 1) {
-    if (xmin == xmax) {
+    /* Note: Do NOT add a coord two time! So when ymin == ymax, should skip this
+     * point. Think about a fabric of 3x3, where the point (1,1) is added twice
+     */
+    if (xmin == xmax && ymin != ymax) {
       for (size_t iy = ymin; iy < ymax + 1; iy++) {
         coords.push_back(vtr::Point<size_t>(xmin, iy));
       }
@@ -399,6 +402,7 @@ static void add_top_module_io_children(
 
   /* Now walk through the coordinates */
   for (vtr::Point<size_t> coord : coords) {
+    VTR_LOG("Adding coord [%lu][%lu]\n", coord.x(), coord.y());
     t_physical_tile_loc phy_tile_loc(coord.x(), coord.y(), layer);
     t_physical_tile_type_ptr grid_type = grids.get_physical_type(phy_tile_loc);
     /* Bypass EMPTY grid */
diff --git a/openfpga_flow/regression_test_scripts/basic_reg_test.sh b/openfpga_flow/regression_test_scripts/basic_reg_test.sh
index 8803cd5fd..2a3687033 100755
--- a/openfpga_flow/regression_test_scripts/basic_reg_test.sh
+++ b/openfpga_flow/regression_test_scripts/basic_reg_test.sh
@@ -175,6 +175,7 @@ echo -e "Testing tiles with I/O in center grid";
 run-task basic_tests/tile_organization/tileable_io $@
 echo -e "Testing tiles with I/O consisting of subtiles";
 run-task basic_tests/tile_organization/io_subtile $@
+run-task basic_tests/tile_organization/io_subtile_strong $@
 echo -e "Testing tile grouping on a homogeneous FPGA fabric (Full testbench)";
 run-task basic_tests/tile_organization/homo_fabric_tile $@
 echo -e "Testing tile grouping on a homogeneous FPGA fabric (Preconfigured testbench)";
diff --git a/openfpga_flow/tasks/basic_tests/tile_organization/io_subtile_strong/config/task.conf b/openfpga_flow/tasks/basic_tests/tile_organization/io_subtile_strong/config/task.conf
new file mode 100644
index 000000000..8f20754ba
--- /dev/null
+++ b/openfpga_flow/tasks/basic_tests/tile_organization/io_subtile_strong/config/task.conf
@@ -0,0 +1,36 @@
+# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
+# Configuration file for running experiments
+# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
+# timeout_each_job : FPGA Task script splits fpga flow into multiple jobs
+# Each job execute fpga_flow script on combination of architecture & benchmark
+# timeout_each_job is timeout for each job
+# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
+
+[GENERAL]
+run_engine=openfpga_shell
+power_tech_file = ${PATH:OPENFPGA_PATH}/openfpga_flow/tech/PTM_45nm/45nm.xml
+power_analysis = true
+spice_output=false
+verilog_output=true
+timeout_each_job = 20*60
+fpga_flow=yosys_vpr
+
+[OpenFPGA_SHELL]
+openfpga_shell_template=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_shell_scripts/fix_device_example_script.openfpga
+openfpga_arch_file=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_arch/k4_N4_40nm_IoSubtile_cc_openfpga.xml
+openfpga_sim_setting_file=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_simulation_settings/auto_sim_openfpga.xml
+openfpga_vpr_device_layout=3x3
+
+[ARCHITECTURES]
+arch0=${PATH:OPENFPGA_PATH}/openfpga_flow/vpr_arch/k4_N4_tileable_IoSubtile_40nm.xml
+
+[BENCHMARKS]
+bench0=${PATH:OPENFPGA_PATH}/openfpga_flow/benchmarks/micro_benchmark/or2/or2.v
+
+[SYNTHESIS_PARAM]
+bench_read_verilog_options_common = -nolatches
+bench0_top = or2
+
+[SCRIPT_PARAM_MIN_ROUTE_CHAN_WIDTH]
+end_flow_with_test=
+vpr_fpga_verilog_formal_verification_top_netlist=
diff --git a/openfpga_flow/vpr_arch/k4_N4_tileable_IoSubtile_40nm.xml b/openfpga_flow/vpr_arch/k4_N4_tileable_IoSubtile_40nm.xml
index 857064fcd..ad58bdb6b 100644
--- a/openfpga_flow/vpr_arch/k4_N4_tileable_IoSubtile_40nm.xml
+++ b/openfpga_flow/vpr_arch/k4_N4_tileable_IoSubtile_40nm.xml
@@ -29,6 +29,34 @@
     </model>
   </models>
   <tiles>
+    <tile name="hybrid_io_tile_center" area="0">
+      <sub_tile name="fpga_input_center" capacity="6">
+        <equivalent_sites>
+          <site pb_type="fpga_input"/>
+        </equivalent_sites>
+        <output name="inpad" num_pins="1"/>
+        <fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
+        <pinlocations pattern="custom">
+          <loc side="left">fpga_input_center[1:0].inpad</loc>
+          <loc side="top">fpga_input_center[3:2].inpad</loc>
+          <loc side="right">fpga_input_center[4:4].inpad</loc>
+          <loc side="bottom">fpga_input_center[5:5].inpad</loc>
+        </pinlocations>
+      </sub_tile>
+      <sub_tile name="fpga_output_center" capacity="4">
+        <equivalent_sites>
+          <site pb_type="fpga_output"/>
+        </equivalent_sites>
+        <input name="outpad" num_pins="1"/>
+        <fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
+        <pinlocations pattern="custom">
+          <loc side="left">fpga_output_center[1:0].outpad</loc>
+          <loc side="top"/>
+          <loc side="right">fpga_output_center[3:2].outpad</loc>
+          <loc side="bottom"/>
+        </pinlocations>
+      </sub_tile>
+    </tile>
     <tile name="hybrid_io_tile" area="0">
       <sub_tile name="fpga_input" capacity="4">
         <equivalent_sites>
@@ -89,6 +117,14 @@
       <!--Fill with 'clb'-->
       <fill type="clb" priority="10"/>
     </fixed_layout>
+    <fixed_layout name="3x3" width="5" height="5">
+      <!--Perimeter of 'io' blocks with 'EMPTY' blocks at corners-->
+      <perimeter type="hybrid_io_tile" priority="100"/>
+      <single type="hybrid_io_tile_center" x="2" y="2" priority="100"/>
+      <corners type="EMPTY" priority="101"/>
+      <!--Fill with 'clb'-->
+      <fill type="clb" priority="10"/>
+    </fixed_layout>
   </layout>
   <device>
     <sizing R_minW_nmos="8926" R_minW_pmos="16067"/>
diff --git a/vtr-verilog-to-routing b/vtr-verilog-to-routing
index 8f4c5567f..62fa48399 160000
--- a/vtr-verilog-to-routing
+++ b/vtr-verilog-to-routing
@@ -1 +1 @@
-Subproject commit 8f4c5567f4ff92db0659d54dd1eb6435d3b7a9f6
+Subproject commit 62fa48399eabb8788844ad852fca6cdec800accf