diff --git a/compiler/base/design.py b/compiler/base/design.py
index 7a570b28..1fded37e 100644
--- a/compiler/base/design.py
+++ b/compiler/base/design.py
@@ -172,6 +172,20 @@ class design(hierarchy_design):
             self.well_extend_active = max(self.well_extend_active, self.nwell_extend_active)
         if "pwell" in layer:
             self.well_extend_active = max(self.well_extend_active, self.pwell_extend_active)
+
+        # The active offset is due to the well extension
+        if "pwell" in layer:
+            self.pwell_enclose_active = drc("pwell_enclose_active")
+        else:
+            self.pwell_enclose_active = 0
+        if "nwell" in layer:
+            self.nwell_enclose_active = drc("nwell_enclose_active")
+        else:
+            self.nwell_enclose_active = 0
+        # Use the max of either so that the poly gates will align properly
+        self.well_enclose_active = max(self.pwell_enclose_active,
+                                       self.nwell_enclose_active,
+                                       self.active_space)
             
         # These are for debugging previous manual rules
         if False:
diff --git a/compiler/base/hierarchy_layout.py b/compiler/base/hierarchy_layout.py
index d3b84ca5..362c0f9e 100644
--- a/compiler/base/hierarchy_layout.py
+++ b/compiler/base/hierarchy_layout.py
@@ -695,8 +695,8 @@ class layout():
         # we should add a boundary just for DRC in some technologies
         if not self.is_library_cell and not self.bounding_box:
             # If there is a boundary layer, and we didn't create one, add one.
-            if "stdc" in techlayer.keys():
-                boundary_layer = "stdc"
+            if "boundary" in techlayer.keys():
+                boundary_layer = "boundary"
                 boundary = [self.find_lowest_coords(),
                             self.find_highest_coords()]
                 debug.check(boundary[0] and boundary[1], "No shapes to make a boundary.")
@@ -1259,10 +1259,7 @@ class layout():
         if OPTS.netlist_only:
             return
 
-        if "stdc" in techlayer.keys():
-            boundary_layer = "stdc"
-        else:
-            boundary_layer = "boundary"
+        boundary_layer = "boundary"
         if not ur:
             self.bounding_box = self.add_rect(layer=boundary_layer,
                                               offset=ll,
@@ -1307,13 +1304,10 @@ class layout():
                                     pin.ll(),
                                     pin.width(),
                                     pin.height())
-            elif pin.layer == "m1":
-                self.add_power_pin(name, pin.center())
-            else:
-                debug.warning("{0} pins of {1} should be on {2} or metal1 for "\
-                              "supply router."
-                              .format(name, inst.name, self.pwr_grid_layer))
 
+            else:
+                self.add_power_pin(name, pin.center(), start_layer=pin.layer)
+ 
     def add_power_pin(self, name, loc, size=[1, 1], directions=None, start_layer="m1"):
         """
         Add a single power pin from the lowest power_grid layer down to M1 (or li) at
diff --git a/compiler/modules/dff_buf.py b/compiler/modules/dff_buf.py
index 0366b10b..81629771 100644
--- a/compiler/modules/dff_buf.py
+++ b/compiler/modules/dff_buf.py
@@ -7,7 +7,7 @@
 #
 import debug
 import design
-from tech import parameter
+from tech import parameter, layer
 from tech import cell_properties as props
 from vector import vector
 from globals import OPTS
@@ -52,7 +52,6 @@ class dff_buf(design.design):
     def create_layout(self):
         self.place_instances()
         self.width = self.inv2_inst.rx()
-
         self.height = self.dff.height
         self.route_wires()
         self.add_layout_pins()
@@ -120,39 +119,37 @@ class dff_buf(design.design):
         except AttributeError:
             pass
         self.inv1_inst.place(vector(self.dff_inst.rx() + well_spacing + self.well_extend_active, 0))
-        
+       
         # Add INV2 to the right
         self.inv2_inst.place(vector(self.inv1_inst.rx(), 0))
         
     def route_wires(self):
+        if "li" in layer:
+            self.route_layer = "li"
+        else:
+            self.route_layer = "m1"
+        
         # Route dff q to inv1 a
         q_pin = self.dff_inst.get_pin("Q")
         a1_pin = self.inv1_inst.get_pin("A")
-        mid_x_offset = 0.5 * (a1_pin.cx() + q_pin.cx())
-        mid1 = vector(mid_x_offset, q_pin.cy())
-        mid2 = vector(mid_x_offset, a1_pin.cy())
-        self.add_path("m3", [q_pin.center(), mid1, mid2, a1_pin.center()])
-        self.add_via_center(layers=self.m2_stack,
-                            offset=q_pin.center())
-        self.add_via_center(layers=self.m2_stack,
-                            offset=a1_pin.center())
-        self.add_via_center(layers=self.m1_stack,
-                            offset=a1_pin.center())
+        mid1 = vector(a1_pin.cx(), q_pin.cy())
+        self.add_path(q_pin.layer, [q_pin.center(), mid1, a1_pin.center()])
+        self.add_via_stack_center(from_layer=a1_pin.layer,
+                                  to_layer=q_pin.layer,
+                                  offset=a1_pin.center())
 
         # Route inv1 z to inv2 a
         z1_pin = self.inv1_inst.get_pin("Z")
         a2_pin = self.inv2_inst.get_pin("A")
-        mid_x_offset = 0.5 * (z1_pin.cx() + a2_pin.cx())
-        self.mid_qb_pos = vector(mid_x_offset, z1_pin.cy())
-        mid2 = vector(mid_x_offset, a2_pin.cy())
-        self.add_path("m1", [z1_pin.center(), self.mid_qb_pos, mid2, a2_pin.center()])
+        self.mid_qb_pos = vector(0.5 * (z1_pin.cx() + a2_pin.cx()), z1_pin.cy())
+        self.add_zjog(z1_pin.layer, z1_pin.center(), a2_pin.center())
         
     def add_layout_pins(self):
 
         # Continous vdd rail along with label.
         vdd_pin=self.dff_inst.get_pin("vdd")
         self.add_layout_pin(text="vdd",
-                            layer="m1",
+                            layer=vdd_pin.layer,
                             offset=vdd_pin.ll(),
                             width=self.width,
                             height=vdd_pin.height())
@@ -160,7 +157,7 @@ class dff_buf(design.design):
         # Continous gnd rail along with label.
         gnd_pin=self.dff_inst.get_pin("gnd")
         self.add_layout_pin(text="gnd",
-                            layer="m1",
+                            layer=gnd_pin.layer,
                             offset=gnd_pin.ll(),
                             width=self.width,
                             height=vdd_pin.height())
@@ -185,17 +182,20 @@ class dff_buf(design.design):
         self.add_layout_pin_rect_center(text="Q",
                                         layer="m2",
                                         offset=q_pos)
-        self.add_path("m1", [dout_pin.center(), mid_pos, q_pos])
-        self.add_via_center(layers=self.m1_stack,
-                            offset=q_pos)
+        self.add_path(self.route_layer, [dout_pin.center(), mid_pos, q_pos])
+        self.add_via_stack_center(from_layer=dout_pin.layer,
+                                  to_layer="m2",
+                                  offset=q_pos)
 
         qb_pos = self.mid_qb_pos + vector(0, self.m2_pitch)
         self.add_layout_pin_rect_center(text="Qb",
                                         layer="m2",
                                         offset=qb_pos)
-        self.add_path("m1", [self.mid_qb_pos, qb_pos])
-        self.add_via_center(layers=self.m1_stack,
-                            offset=qb_pos)
+        self.add_path(self.route_layer, [self.mid_qb_pos, qb_pos])
+        a2_pin = self.inv2_inst.get_pin("A")
+        self.add_via_stack_center(from_layer=a2_pin.layer,
+                                  to_layer="m2",
+                                  offset=qb_pos)
          
     def get_clk_cin(self):
         """Return the total capacitance (in relative units) that the clock is loaded by in the dff"""
diff --git a/compiler/modules/replica_column.py b/compiler/modules/replica_column.py
index afe87497..603bebfe 100644
--- a/compiler/modules/replica_column.py
+++ b/compiler/modules/replica_column.py
@@ -154,7 +154,7 @@ class replica_column(design.design):
             bl_pin = self.cell_inst[0].get_pin(bl_name)
             self.add_layout_pin(text=bl_name,
                                 layer=bl_pin.layer,
-                                offset=bl_pin.ll(),
+                                offset=bl_pin.ll().scale(1, 0),
                                 width=bl_pin.width(),
                                 height=self.height)
 
diff --git a/compiler/pgates/pgate.py b/compiler/pgates/pgate.py
index 8e18d4d2..67ba4a18 100644
--- a/compiler/pgates/pgate.py
+++ b/compiler/pgates/pgate.py
@@ -46,7 +46,8 @@ class pgate(design.design):
         # This is the space from a S/D contact to the supply rail
         contact_to_vdd_rail_space = 0.5 * self.m1_width + self.m1_space
         # This is a poly-to-poly of a flipped cell
-        poly_to_poly_gate_space = self.poly_extend_active + self.poly_space
+        poly_to_poly_gate_space = self.poly_extend_active + 0.5 * self.poly_space
+        
         self.top_bottom_space = max(contact_to_vdd_rail_space,
                                     poly_to_poly_gate_space)
         
diff --git a/compiler/pgates/pinvbuf.py b/compiler/pgates/pinvbuf.py
index fe376bc4..5b286e9b 100644
--- a/compiler/pgates/pinvbuf.py
+++ b/compiler/pgates/pinvbuf.py
@@ -9,7 +9,7 @@ import debug
 import pgate
 from vector import vector
 from sram_factory import factory
-
+from tech import layer
 
 class pinvbuf(pgate.pgate):
     """
@@ -111,33 +111,45 @@ class pinvbuf(pgate.pgate):
                              mirror="MX")
         
     def route_wires(self):
+        if "li" in layer:
+            route_stack = self.li_stack
+        else:
+            route_stack = self.m1_stack
+            
         # inv1 Z to inv2 A
         z1_pin = self.inv1_inst.get_pin("Z")
         a2_pin = self.inv2_inst.get_pin("A")
         mid_point = vector(z1_pin.cx(), a2_pin.cy())
-        self.add_path("m1", [z1_pin.center(), mid_point, a2_pin.center()])
+        self.add_path(z1_pin.layer, [z1_pin.center(), mid_point, a2_pin.center()])
+        self.add_via_stack_center(from_layer=z1_pin.layer,
+                                  to_layer=a2_pin.layer,
+                                  offset=a2_pin.center())
         
         # inv2 Z to inv3 A
         z2_pin = self.inv2_inst.get_pin("Z")
         a3_pin = self.inv3_inst.get_pin("A")
         mid_point = vector(z2_pin.cx(), a3_pin.cy())
-        self.add_path("m1", [z2_pin.center(), mid_point, a3_pin.center()])
+        self.add_path(z2_pin.layer, [z2_pin.center(), mid_point, a3_pin.center()])
+        self.add_via_stack_center(from_layer=z2_pin.layer,
+                                  to_layer=a3_pin.layer,
+                                  offset=a3_pin.center())
 
         # inv1 Z to inv4 A (up and over)
         z1_pin = self.inv1_inst.get_pin("Z")
         a4_pin = self.inv4_inst.get_pin("A")
         mid_point = vector(z1_pin.cx(), a4_pin.cy())
-        self.add_wire(self.m1_stack,
+        self.add_wire(route_stack,
                       [z1_pin.center(), mid_point, a4_pin.center()])
-        self.add_via_center(layers=self.m1_stack,
-                            offset=z1_pin.center())
+        self.add_via_stack_center(from_layer=z1_pin.layer,
+                                  to_layer=route_stack[2],
+                                  offset=z1_pin.center())
         
     def add_layout_pins(self):
 
         # Continous vdd rail along with label.
         vdd_pin = self.inv1_inst.get_pin("vdd")
         self.add_layout_pin(text="vdd",
-                            layer="m1",
+                            layer=vdd_pin.layer,
                             offset=vdd_pin.ll().scale(0, 1),
                             width=self.width,
                             height=vdd_pin.height())
@@ -145,7 +157,7 @@ class pinvbuf(pgate.pgate):
         # Continous vdd rail along with label.
         gnd_pin = self.inv4_inst.get_pin("gnd")
         self.add_layout_pin(text="gnd",
-                            layer="m1",
+                            layer=gnd_pin.layer,
                             offset=gnd_pin.ll().scale(0, 1),
                             width=self.width,
                             height=gnd_pin.height())
@@ -153,31 +165,25 @@ class pinvbuf(pgate.pgate):
         # Continous gnd rail along with label.
         gnd_pin = self.inv1_inst.get_pin("gnd")
         self.add_layout_pin(text="gnd",
-                            layer="m1",
+                            layer=gnd_pin.layer,
                             offset=gnd_pin.ll().scale(0, 1),
                             width=self.width,
                             height=vdd_pin.height())
             
         z_pin = self.inv4_inst.get_pin("Z")
         self.add_layout_pin_rect_center(text="Z",
-                                        layer="m2",
+                                        layer=z_pin.layer,
                                         offset=z_pin.center())
-        self.add_via_center(layers=self.m1_stack,
-                            offset=z_pin.center())
 
         zb_pin = self.inv3_inst.get_pin("Z")
         self.add_layout_pin_rect_center(text="Zb",
-                                        layer="m2",
+                                        layer=zb_pin.layer,
                                         offset=zb_pin.center())
-        self.add_via_center(layers=self.m1_stack,
-                            offset=zb_pin.center())
         
         a_pin = self.inv1_inst.get_pin("A")
         self.add_layout_pin_rect_center(text="A",
-                                        layer="m2",
+                                        layer=a_pin.layer,
                                         offset=a_pin.center())
-        self.add_via_center(layers=self.m1_stack,
-                            offset=a_pin.center())
              
     def determine_clk_buf_stage_efforts(self, external_cout, inp_is_rise=False):
         """Get the stage efforts of the clk -> clk_buf path"""
diff --git a/compiler/pgates/pnand2.py b/compiler/pgates/pnand2.py
index 51581e61..5816adcf 100644
--- a/compiler/pgates/pnand2.py
+++ b/compiler/pgates/pnand2.py
@@ -177,9 +177,9 @@ class pnand2(pgate.pgate):
 
 
         # Top of NMOS drain
-        nmos_pin = self.nmos2_inst.get_pin("D")
-        bottom_pin_offset = nmos_pin.uy()
-        self.inputA_yoffset = bottom_pin_offset + self.m1_pitch
+        bottom_pin = self.nmos2_inst.get_pin("D")
+        self.inputA_yoffset = max(bottom_pin.uy() + self.m1_pitch,
+                                  self.nmos2_inst.uy() + self.poly_to_active)
 
         self.inputB_yoffset = self.inputA_yoffset + self.m3_pitch
 
diff --git a/compiler/pgates/pnand3.py b/compiler/pgates/pnand3.py
index 6cbd7cca..31024d82 100644
--- a/compiler/pgates/pnand3.py
+++ b/compiler/pgates/pnand3.py
@@ -12,7 +12,7 @@ from vector import vector
 import logical_effort
 from sram_factory import factory
 from globals import OPTS
-
+import contact
 
 class pnand3(pgate.pgate):
     """
@@ -209,13 +209,21 @@ class pnand3(pgate.pgate):
     def route_inputs(self):
         """ Route the A and B and C inputs """
 
+        # We can use this pitch because the contacts and overlap won't be adjacent
+        non_contact_pitch = 0.5 * self.m1_width + self.m1_space + 0.5 * contact.poly_contact.second_layer_height
         pmos_drain_bottom = self.pmos1_inst.get_pin("D").by()
         self.output_yoffset = pmos_drain_bottom - 0.5 * self.route_layer_width - self.route_layer_space
 
-        # This is a more compact offset, but the bottom one works better in the decoders to "center" the pins
-        # in the height of the gates
-        self.inputA_yoffset = self.output_yoffset - 0.5 * self.route_layer_width - self.route_layer_space
-        # self.inputA_yoffset = self.output_yoffset - self.m1_pitch
+        bottom_pin = self.nmos1_inst.get_pin("D")
+        # active contact metal to poly contact metal spacing
+        active_contact_to_poly_contact = bottom_pin.uy() + self.m1_space + 0.5 * contact.poly_contact.second_layer_height
+        # active diffusion to poly contact spacing
+        # doesn't use nmos uy because that is calculated using offset + poly height
+        active_to_poly_contact = self.nmos1_inst.by() + self.nmos1_inst.mod.active_height \
+                                 + self.poly_to_active + 0.5 * contact.poly_contact.first_layer_height
+
+        self.inputA_yoffset = max(active_contact_to_poly_contact,
+                                  active_to_poly_contact)
         self.route_input_gate(self.pmos1_inst,
                               self.nmos1_inst,
                               self.inputA_yoffset,
@@ -223,14 +231,14 @@ class pnand3(pgate.pgate):
                               position="left")
 
         # Put B right on the well line
-        self.inputB_yoffset = self.inputA_yoffset - self.m1_pitch
+        self.inputB_yoffset = self.inputA_yoffset + non_contact_pitch
         self.route_input_gate(self.pmos2_inst,
                               self.nmos2_inst,
                               self.inputB_yoffset,
                               "B",
                               position="center")
 
-        self.inputC_yoffset = self.inputB_yoffset - self.m1_pitch
+        self.inputC_yoffset = self.inputB_yoffset + non_contact_pitch
         self.route_input_gate(self.pmos3_inst,
                               self.nmos3_inst,
                               self.inputC_yoffset,
diff --git a/compiler/pgates/precharge.py b/compiler/pgates/precharge.py
index 76fc8676..07b4f21b 100644
--- a/compiler/pgates/precharge.py
+++ b/compiler/pgates/precharge.py
@@ -149,13 +149,9 @@ class precharge(design.design):
         # Compute the other pmos2 location,
         # but determining offset to overlap the source and drain pins
         overlap_offset = self.pmos.get_pin("D").ll() - self.pmos.get_pin("S").ll()
-        # This is how much the contact is placed inside the ptx active
-        contact_xdiff = self.pmos.get_pin("S").lx()
         
         # adds the lower pmos to layout
-        bl_xoffset = self.bitcell_bl_pin.lx()
-        self.lower_pmos_position = vector(max(bl_xoffset - contact_xdiff,
-                                              self.nwell_enclose_active),
+        self.lower_pmos_position = vector(self.well_enclose_active + 0.5 * self.m1_width,
                                           self.initial_yoffset)
         self.lower_pmos_inst.place(self.lower_pmos_position)
 
@@ -218,7 +214,7 @@ class precharge(design.design):
         
         # adds the contact from active to metal1
         offset_height = self.upper_pmos1_inst.uy() + \
-                        0.5 * contact.active_contact.height + \
+                        contact.active_contact.height + \
                         self.nwell_extend_active
         self.well_contact_pos = self.upper_pmos1_inst.get_pin("D").center().scale(1, 0) + \
                                 vector(0, offset_height)
diff --git a/compiler/pgates/ptx.py b/compiler/pgates/ptx.py
index f6716a1f..6fba247e 100644
--- a/compiler/pgates/ptx.py
+++ b/compiler/pgates/ptx.py
@@ -205,28 +205,15 @@ class ptx(design.design):
         # Poly height must include poly extension over active
         self.poly_height = self.tx_width + 2 * self.poly_extend_active
         
-        # The active offset is due to the well extension
-        if "pwell" in layer:
-            pwell_enclose_active = drc("pwell_enclose_active")
-        else:
-            pwell_enclose_active = 0
-        if "nwell" in layer:
-            nwell_enclose_active = drc("nwell_enclose_active")
-        else:
-            nwell_enclose_active = 0
-        # Use the max of either so that the poly gates will align properly
-        well_enclose_active = max(pwell_enclose_active,
-                                  nwell_enclose_active)
-        self.active_offset = vector([well_enclose_active] * 2)
+        self.active_offset = vector([self.well_enclose_active] * 2)
 
         # Well enclosure of active, ensure minwidth as well
         well_name = "{}well".format(self.well_type)
         if well_name in layer:
             well_width_rule = drc("minwidth_" + well_name)
-            well_enclose_active = drc(well_name + "_enclose_active")
-            self.well_width = max(self.active_width + 2 * well_enclose_active,
+            self.well_width = max(self.active_width + 2 * self.well_enclose_active,
                                   well_width_rule)
-            self.well_height = max(self.active_height + 2 * well_enclose_active,
+            self.well_height = max(self.active_height + 2 * self.well_enclose_active,
                                    well_width_rule)
             # We are going to shift the 0,0, so include that in the width and height
             self.height = self.well_height - self.active_offset.y
diff --git a/compiler/tests/11_dff_buf_test.py b/compiler/tests/04_dff_buf_test.py
similarity index 100%
rename from compiler/tests/11_dff_buf_test.py
rename to compiler/tests/04_dff_buf_test.py
diff --git a/compiler/tests/04_pand2_test.py b/compiler/tests/04_pand2_test.py
index 21a0a38e..f7e5f304 100755
--- a/compiler/tests/04_pand2_test.py
+++ b/compiler/tests/04_pand2_test.py
@@ -29,10 +29,6 @@ class pand2_test(openram_test):
         a = pand2.pand2(name="pand2x4", size=4)
         self.local_check(a)
 
-        debug.info(2, "Testing vertical pand2 gate 4x")
-        a = pand2.pand2(name="pand2x4", size=4, vertical=True)
-        self.local_check(a)
-        
         globals.end_openram()
 
 # instantiate a copdsay of the class to actually run the test
diff --git a/compiler/tests/04_pand3_test.py b/compiler/tests/04_pand3_test.py
index f851077b..e58f1ee9 100755
--- a/compiler/tests/04_pand3_test.py
+++ b/compiler/tests/04_pand3_test.py
@@ -29,10 +29,6 @@ class pand3_test(openram_test):
         a = pand3.pand3(name="pand3x4", size=4)
         self.local_check(a)
 
-        debug.info(2, "Testing vertical pand3 gate 4x")
-        a = pand3.pand3(name="pand3x4", size=4, vertical=True)
-        self.local_check(a)
-        
         globals.end_openram()
 
 # instantiate a copdsay of the class to actually run the test
diff --git a/compiler/tests/04_precharge_1rw_1r_test.py b/compiler/tests/04_precharge_1rw_1r_test.py
new file mode 100755
index 00000000..4765c77c
--- /dev/null
+++ b/compiler/tests/04_precharge_1rw_1r_test.py
@@ -0,0 +1,47 @@
+#!/usr/bin/env python3
+# See LICENSE for licensing information.
+#
+# Copyright (c) 2016-2019 Regents of the University of California and The Board
+# of Regents for the Oklahoma Agricultural and Mechanical College
+# (acting for and on behalf of Oklahoma State University)
+# All rights reserved.
+#
+import unittest
+from testutils import *
+import sys,os
+sys.path.append(os.getenv("OPENRAM_HOME"))
+import globals
+from globals import OPTS
+from sram_factory import factory
+import debug
+
+
+class precharge_test(openram_test):
+
+    def runTest(self):
+        config_file = "{}/tests/configs/config".format(os.getenv("OPENRAM_HOME"))
+        globals.init_openram(config_file)
+        
+        # check precharge array in multi-port
+        OPTS.num_rw_ports = 1
+        OPTS.num_r_ports = 1
+        OPTS.num_w_ports = 0
+        globals.setup_bitcell()
+
+        debug.info(2, "Checking precharge for 1rw1r port 0")
+        tx = factory.create(module_type="precharge", size=1, bitcell_bl="bl0", bitcell_br="br0")
+        self.local_check(tx)
+
+        factory.reset()
+        debug.info(2, "Checking precharge for 1rw1r port 1")
+        tx = factory.create(module_type="precharge", size=1, bitcell_bl="bl1", bitcell_br="br1")
+        self.local_check(tx)
+
+        globals.end_openram()
+        
+# run the test from the command line
+if __name__ == "__main__":
+    (OPTS, args) = globals.parse_args()
+    del sys.argv[1:]
+    header(__file__, OPTS.tech_name)
+    unittest.main(testRunner=debugTestRunner())
diff --git a/compiler/tests/05_bitcell_1rw_1r_array_test.py b/compiler/tests/05_bitcell_array_1rw_1r_test.py
similarity index 96%
rename from compiler/tests/05_bitcell_1rw_1r_array_test.py
rename to compiler/tests/05_bitcell_array_1rw_1r_test.py
index 0683d127..0e7d5665 100755
--- a/compiler/tests/05_bitcell_1rw_1r_array_test.py
+++ b/compiler/tests/05_bitcell_array_1rw_1r_test.py
@@ -17,7 +17,7 @@ import debug
 
 #@unittest.skip("SKIPPING 05_bitcell_1rw_1r_array_test")
 
-class bitcell_1rw_1r_array_test(openram_test):
+class bitcell_array_1rw_1r_test(openram_test):
 
     def runTest(self):
         config_file = "{}/tests/configs/config".format(os.getenv("OPENRAM_HOME"))
diff --git a/compiler/tests/09_sense_amp_array_1rw_1r_test.py b/compiler/tests/09_sense_amp_array_1rw_1r_test.py
new file mode 100755
index 00000000..a8ed4d2f
--- /dev/null
+++ b/compiler/tests/09_sense_amp_array_1rw_1r_test.py
@@ -0,0 +1,48 @@
+#!/usr/bin/env python3
+# See LICENSE for licensing information.
+#
+# Copyright (c) 2016-2019 Regents of the University of California and The Board
+# of Regents for the Oklahoma Agricultural and Mechanical College
+# (acting for and on behalf of Oklahoma State University)
+# All rights reserved.
+#
+import unittest
+from testutils import *
+import sys,os
+sys.path.append(os.getenv("OPENRAM_HOME"))
+import globals
+from globals import OPTS
+from sram_factory import factory
+import debug
+
+class sense_amp_test(openram_test):
+
+    def runTest(self):
+        config_file = "{}/tests/configs/config".format(os.getenv("OPENRAM_HOME"))
+        globals.init_openram(config_file)
+
+        OPTS.num_rw_ports = 1
+        OPTS.num_r_ports = 1
+        OPTS.num_w_ports = 0
+        globals.setup_bitcell()
+
+        debug.info(2, "Testing sense_amp_array for word_size=4, words_per_row=1")
+        a = factory.create(module_type="sense_amp_array", word_size=4, words_per_row=1)
+        self.local_check(a)
+
+        debug.info(2, "Testing sense_amp_array for word_size=4, words_per_row=2")
+        a = factory.create(module_type="sense_amp_array", word_size=4, words_per_row=2)
+        self.local_check(a)
+
+        debug.info(2, "Testing sense_amp_array for word_size=4, words_per_row=4")
+        a = factory.create(module_type="sense_amp_array", word_size=4, words_per_row=4)
+        self.local_check(a)
+
+        globals.end_openram()
+
+# run the test from the command line
+if __name__ == "__main__":
+    (OPTS, args) = globals.parse_args()
+    del sys.argv[1:]
+    header(__file__, OPTS.tech_name)
+    unittest.main(testRunner=debugTestRunner())
diff --git a/compiler/tests/14_replica_bitcell_1rw_1r_array_test.py b/compiler/tests/14_replica_bitcell_array_1rw_1r_test.py
similarity index 95%
rename from compiler/tests/14_replica_bitcell_1rw_1r_array_test.py
rename to compiler/tests/14_replica_bitcell_array_1rw_1r_test.py
index 3f869255..a36fc80f 100755
--- a/compiler/tests/14_replica_bitcell_1rw_1r_array_test.py
+++ b/compiler/tests/14_replica_bitcell_array_1rw_1r_test.py
@@ -13,7 +13,7 @@ from globals import OPTS
 from sram_factory import factory
 import debug
 
-class replica_bitcell_array_test(openram_test):
+class replica_bitcell_array_1rw_1r_test(openram_test):
 
     def runTest(self):
         config_file = "{}/tests/configs/config".format(os.getenv("OPENRAM_HOME"))
diff --git a/compiler/tests/14_replica_column_1rw_1r_test.py b/compiler/tests/14_replica_column_1rw_1r_test.py
new file mode 100755
index 00000000..3c1a0e1e
--- /dev/null
+++ b/compiler/tests/14_replica_column_1rw_1r_test.py
@@ -0,0 +1,46 @@
+#!/usr/bin/env python3
+# See LICENSE for licensing information.
+#
+# Copyright (c) 2016-2019 Regents of the University of California 
+# All rights reserved.
+#
+import unittest
+from testutils import *
+import sys,os
+sys.path.append(os.getenv("OPENRAM_HOME"))
+import globals
+from globals import OPTS
+from sram_factory import factory
+import debug
+
+class replica_column_test(openram_test):
+
+    def runTest(self):
+        config_file = "{}/tests/configs/config".format(os.getenv("OPENRAM_HOME"))
+        globals.init_openram(config_file)
+
+        OPTS.num_rw_ports = 1
+        OPTS.num_r_ports = 1
+        OPTS.num_w_ports = 0
+        globals.setup_bitcell()
+
+        debug.info(2, "Testing replica column for 6t_cell")
+        a = factory.create(module_type="replica_column", rows=4, left_rbl=1, right_rbl=0, replica_bit=1)
+        self.local_check(a)
+
+        debug.info(2, "Testing replica column for 6t_cell")
+        a = factory.create(module_type="replica_column", rows=4, left_rbl=1, right_rbl=1, replica_bit=6)
+        self.local_check(a)
+        
+        debug.info(2, "Testing replica column for 6t_cell")
+        a = factory.create(module_type="replica_column", rows=4, left_rbl=2, right_rbl=0, replica_bit=2)
+        self.local_check(a)
+        
+        globals.end_openram()
+
+# run the test from the command line
+if __name__ == "__main__":
+    (OPTS, args) = globals.parse_args()
+    del sys.argv[1:]
+    header(__file__, OPTS.tech_name)
+    unittest.main(testRunner=debugTestRunner())
diff --git a/compiler/tests/20_psram_1bank_4mux_1rw_1r_test.py b/compiler/tests/20_psram_1bank_4mux_1rw_1r_test.py
index 145d1723..ecbd0863 100755
--- a/compiler/tests/20_psram_1bank_4mux_1rw_1r_test.py
+++ b/compiler/tests/20_psram_1bank_4mux_1rw_1r_test.py
@@ -15,7 +15,6 @@ from globals import OPTS
 from sram_factory import factory
 import debug
 
-@unittest.skip("SKIPPING 20_psram_1bank_4mux_1rw_1r_test - Matt sucks, don't do this")
 class psram_1bank_4mux_1rw_1r_test(openram_test):
 
     def runTest(self):
diff --git a/compiler/tests/regress.py b/compiler/tests/regress.py
index e60b010d..3fd5d6eb 100755
--- a/compiler/tests/regress.py
+++ b/compiler/tests/regress.py
@@ -22,14 +22,26 @@ header(__file__, OPTS.tech_name)
 # get a list of all files in the tests directory
 files = os.listdir(sys.path[0])
 
+# load a file with all tests to skip in a given technology
+# since tech_name is dynamically loaded, we can't use @skip directives
+try:
+    skip_file_name = "{0}/tests/skip_tests_{1}.txt".format(os.getenv("OPENRAM_HOME"), OPTS.tech_name)
+    skip_file = open(skip_file_name, "r")
+    skip_tests = skip_file.read().splitlines()
+    for st in skip_tests:
+        debug.warning("Skipping: " + st)
+except FileNotFoundError:
+    skip_tests = []
+
 # assume any file that ends in "test.py" in it is a regression test
 nametest = re.compile("test\.py$", re.IGNORECASE)
-tests = list(filter(nametest.search, files))
-tests.sort()
+all_tests = list(filter(nametest.search, files))
+filtered_tests = list(filter(lambda i: i not in skip_tests, all_tests))
+filtered_tests.sort()
 
 # import all of the modules
 filenameToModuleName = lambda f: os.path.splitext(f)[0]
-moduleNames = map(filenameToModuleName, tests)
+moduleNames = map(filenameToModuleName, filtered_tests)
 modules = map(__import__, moduleNames)
 suite = unittest.TestSuite()
 load = unittest.defaultTestLoader.loadTestsFromModule
diff --git a/compiler/tests/skip_tests_s8.txt b/compiler/tests/skip_tests_s8.txt
new file mode 100644
index 00000000..68a6549b
--- /dev/null
+++ b/compiler/tests/skip_tests_s8.txt
@@ -0,0 +1,87 @@
+04_dummy_pbitcell_test.py
+04_pbitcell_test.py
+04_precharge_pbitcell_test.py
+04_replica_pbitcell_test.py
+04_single_level_column_mux_pbitcell_test.py
+05_bitcell_1rw_1r_array_test.py
+05_bitcell_array_test.py
+05_dummy_array_test.py
+05_pbitcell_array_test.py
+06_hierarchical_decoder_pbitcell_test.py
+06_hierarchical_decoder_test.py
+06_hierarchical_predecode2x4_pbitcell_test.py
+06_hierarchical_predecode2x4_test.py
+06_hierarchical_predecode3x8_pbitcell_test.py
+06_hierarchical_predecode3x8_test.py
+06_hierarchical_predecode4x16_test.py
+07_single_level_column_mux_array_pbitcell_test.py
+08_wordline_driver_array_pbitcell_test.py
+08_wordline_driver_array_test.py
+09_sense_amp_array_test_pbitcell.py
+09_sense_amp_array_test.py
+10_write_driver_array_pbitcell_test.py
+10_write_driver_array_test.py
+10_write_driver_array_wmask_pbitcell_test.py
+10_write_driver_array_wmask_test.py
+10_write_mask_and_array_pbitcell_test.py
+10_write_mask_and_array_test.py
+12_tri_gate_array_test.py
+14_replica_pbitcell_array_test.py
+14_replica_bitcell_array_test.py
+14_replica_column_test.py
+14_replica_column_1rw_1r_test.py
+18_port_address_test.py
+18_port_data_test.py
+18_port_data_wmask_test.py
+19_bank_select_pbitcell_test.py
+19_bank_select_test.py
+19_psingle_bank_test.py
+19_bank_select_pbitcell_test.py
+19_pmulti_bank_test.py
+19_multi_bank_test.py
+19_psingle_bank_test.py
+19_single_bank_1w_1r_test.py
+19_single_bank_wmask_1rw_1r_test.py
+19_single_bank_1rw_1r_test.py
+19_single_bank_test.py
+19_single_bank_wmask_test.py
+20_psram_1bank_2mux_1rw_1w_test.py
+20_psram_1bank_2mux_1rw_1w_wmask_test.py
+20_psram_1bank_2mux_1w_1r_test.py
+20_psram_1bank_2mux_test.py
+20_psram_1bank_4mux_1rw_1r_test.py
+20_sram_1bank_2mux_1w_1r_test.py
+20_sram_1bank_2mux_test.py
+20_sram_1bank_2mux_wmask_test.py
+20_sram_1bank_32b_1024_wmask_test.py
+20_sram_1bank_4mux_test.py
+20_sram_1bank_8mux_test.py
+20_sram_1bank_nomux_test.py
+20_sram_1bank_nomux_wmask_test.py
+20_sram_2bank_test.py
+21_hspice_delay_test.py
+21_hspice_setuphold_test.py
+21_model_delay_test.py
+21_ngspice_delay_test.py
+21_ngspice_setuphold_test.py
+22_psram_1bank_2mux_func_test.py
+22_psram_1bank_4mux_func_test.py
+22_psram_1bank_8mux_func_test.py
+22_psram_1bank_nomux_func_test.py
+22_sram_1bank_2mux_func_test.py
+22_sram_1bank_4mux_func_test.py
+22_sram_1bank_8mux_func_test.py
+22_sram_1bank_nomux_func_test.py
+22_sram_1rw_1r_1bank_nomux_func_test.py
+22_sram_wmask_func_test.py
+23_lib_sram_model_corners_test.py
+23_lib_sram_model_test.py
+23_lib_sram_prune_test.py
+23_lib_sram_test.py
+24_lef_sram_test.py
+25_verilog_sram_test.py
+26_hspice_pex_pinv_test.py
+26_ngspice_pex_pinv_test.py
+26_pex_test.py
+30_openram_back_end_test.py
+30_openram_front_end_test.py
diff --git a/technology/freepdk45/gds_lib/dff.gds b/technology/freepdk45/gds_lib/dff.gds
index 526a1861..2a6004d1 100644
Binary files a/technology/freepdk45/gds_lib/dff.gds and b/technology/freepdk45/gds_lib/dff.gds differ