diff --git a/compiler/base/graph_util.py b/compiler/base/graph_util.py
index 7cf8ee6f..5d1ee692 100644
--- a/compiler/base/graph_util.py
+++ b/compiler/base/graph_util.py
@@ -18,13 +18,16 @@ class timing_graph():
     def __init__(self):
         self.graph = defaultdict(set)
         self.all_paths = []
+        self.edge_mods = {}
 
-    def add_edge(self, src_node, dest_node):
-        """Adds edge to graph. Nodes added as well if they do not exist."""
+    def add_edge(self, src_node, dest_node, edge_mod):
+        """Adds edge to graph. Nodes added as well if they do not exist.
+           Module which defines the edge must be provided for timing information."""
         
         src_node = src_node.lower()
         dest_node = dest_node.lower()
         self.graph[src_node].add(dest_node)
+        self.edge_mods[(src_node, dest_node)] = edge_mod
 
     def add_node(self, node):
         """Add node to graph with no edges"""
@@ -34,8 +37,8 @@ class timing_graph():
             self.graph[node] = set()
             
     def remove_edges(self, node):
-        
         """Helper function to remove edges, useful for removing vdd/gnd"""
+        
         node = node.lower()
         self.graph[node] = set()    
             
@@ -93,7 +96,33 @@ class timing_graph():
                       
         # Remove current vertex from path[] and mark it as unvisited 
         path.pop() 
-        visited.remove(cur_node)        
+        visited.remove(cur_node)    
+
+    def get_timing(self, path, corner, slew, load):
+        """Returns the analytical delays in the input path"""
+        
+        if len(path) == 0:
+            return []
+            
+        delays = []
+        cur_slew = slew
+        for i in range(len(path)-1):
+            
+            path_edge_mod = self.edge_mods[(path[i], path[i+1])]
+            
+            # On the output of the current stage, get COUT from all other mods connected
+            cout = 0
+            for node in self.graph[path[i+1]]:
+                output_edge_mod = self.edge_mods[(path[i+1], node)]
+                cout+=output_edge_mod.get_cin()
+            # If at the last output, include the final output load    
+            if i == len(path)-2:
+               cout+=load
+            
+            delays.append(path_edge_mod.analytical_delay(corner, slew, cout))    
+            cur_slew = delays[-1].slew
+            
+        return delays
             
     def __str__(self):
         """ override print function output """
diff --git a/compiler/base/hierarchy_design.py b/compiler/base/hierarchy_design.py
index 05b2182f..055bf412 100644
--- a/compiler/base/hierarchy_design.py
+++ b/compiler/base/hierarchy_design.py
@@ -110,7 +110,7 @@ class hierarchy_design(hierarchy_spice.spice, hierarchy_layout.layout):
         """Initializes parameters relevant to the graph creation"""
         #Only initializes a set for checking instances which should not be added
         self.graph_inst_exclude = set()
-            
+    
     def build_graph(self, graph, inst_name, port_nets):        
         """Recursively create graph from instances in module."""
         
@@ -219,7 +219,7 @@ class hierarchy_design(hierarchy_spice.spice, hierarchy_layout.layout):
         for inp in input_pins+inout_pins:
             for out in output_pins+inout_pins:
                 if inp != out: #do not add self loops
-                    graph.add_edge(pin_dict[inp], pin_dict[out])        
+                    graph.add_edge(pin_dict[inp], pin_dict[out], self)        
  
     def __str__(self):
         """ override print function output """
diff --git a/compiler/base/hierarchy_spice.py b/compiler/base/hierarchy_spice.py
index 2dcc3d73..5166ac6b 100644
--- a/compiler/base/hierarchy_spice.py
+++ b/compiler/base/hierarchy_spice.py
@@ -13,7 +13,7 @@ import tech
 from delay_data import *
 from wire_spice_model import *
 from power_data import *
-
+import logical_effort
 
 class spice():
     """
@@ -304,14 +304,48 @@ class spice():
 
     def analytical_delay(self, corner, slew, load=0.0):
         """Inform users undefined delay module while building new modules"""
-        debug.warning("Design Class {0} delay function needs to be defined"
+        
+        # FIXME: Slew is not used in the model right now. Can be added heuristically as linear factor 
+        relative_cap = logical_effort.convert_farad_to_relative_c(load)
+        stage_effort = self.get_stage_effort(relative_cap)
+        
+        # If it fails, then keep running with a valid object.
+        if stage_effort == None:
+            return delay_data(0.0, 0.0)
+            
+        abs_delay = stage_effort.get_absolute_delay()
+        corner_delay = self.apply_corners_analytically(abs_delay, corner)
+        SLEW_APPROXIMATION = 0.1
+        corner_slew = SLEW_APPROXIMATION*corner_delay
+        return delay_data(corner_delay, corner_slew)
+
+    def get_stage_effort(self, corner, slew, load=0.0):
+        """Inform users undefined delay module while building new modules"""
+        debug.warning("Design Class {0} logical effort function needs to be defined"
                       .format(self.__class__.__name__))
         debug.warning("Class {0} name {1}"
                       .format(self.__class__.__name__, 
                               self.name))         
-        # return 0 to keep code running while building
-        return delay_data(0.0, 0.0)
-
+        return None   
+        
+    def get_cin(self):
+        """Returns input load in Femto-Farads. All values generated using
+           relative capacitance function then converted based on tech file parameter."""
+        
+        # Override this function within a module if a more accurate input capacitance is needed.
+        # Input/outputs with differing capacitances is not implemented.
+        relative_cap = self.input_load()
+        return logical_effort.convert_relative_c_to_farad(relative_cap)
+        
+    def input_load(self):
+        """Inform users undefined relative capacitance functions used for analytical delays."""
+        debug.warning("Design Class {0} input capacitance function needs to be defined"
+                      .format(self.__class__.__name__))
+        debug.warning("Class {0} name {1}"
+                      .format(self.__class__.__name__, 
+                              self.name))         
+        return 0   
+        
     def cal_delay_with_rc(self, corner, r, c ,slew, swing = 0.5):
         """ 
         Calculate the delay of a mosfet by 
diff --git a/compiler/bitcells/bitcell.py b/compiler/bitcells/bitcell.py
index 96432cea..cad069a5 100644
--- a/compiler/bitcells/bitcell.py
+++ b/compiler/bitcells/bitcell.py
@@ -36,7 +36,7 @@ class bitcell(design.design):
         self.add_pin_types(self.type_list)
         self.nets_match = self.do_nets_exist(self.storage_nets)
         
-    def analytical_delay(self, corner, slew, load=0, swing = 0.5):
+    def get_stage_effort(self, load):
         parasitic_delay = 1
         size = 0.5 #This accounts for bitline being drained thought the access TX and internal node
         cin = 3 #Assumes always a minimum sizes inverter. Could be specified in the tech.py file.
@@ -94,10 +94,10 @@ class bitcell(design.design):
             debug.info(1,"Storage nodes={} not found in spice file.".format(self.storage_nets))
             return None
     
-    def get_wl_cin(self):
+    def input_load(self):
         """Return the relative capacitance of the access transistor gates"""
-        #This is a handmade cell so the value must be entered in the tech.py file or estimated.
-        #Calculated in the tech file by summing the widths of all the related gates and dividing by the minimum width.
+        
+        # FIXME: This applies to bitline capacitances as well.
         access_tx_cin = parameter["6T_access_size"]/drc["minwidth_tx"]
         return 2*access_tx_cin
 
diff --git a/compiler/bitcells/bitcell_1rw_1r.py b/compiler/bitcells/bitcell_1rw_1r.py
index f627a4ac..0d536b38 100644
--- a/compiler/bitcells/bitcell_1rw_1r.py
+++ b/compiler/bitcells/bitcell_1rw_1r.py
@@ -36,7 +36,7 @@ class bitcell_1rw_1r(design.design):
         self.add_pin_types(self.type_list)
         self.nets_match = self.do_nets_exist(self.storage_nets)
         
-    def analytical_delay(self, corner, slew, load=0, swing = 0.5):
+    def get_stage_effort(self, load):
         parasitic_delay = 1
         size = 0.5 #This accounts for bitline being drained thought the access TX and internal node
         cin = 3 #Assumes always a minimum sizes inverter. Could be specified in the tech.py file.
@@ -118,11 +118,11 @@ class bitcell_1rw_1r(design.design):
         total_power = self.return_power(dynamic, leakage)
         return total_power
 
-    def get_wl_cin(self):
+    def input_load(self):
         """Return the relative capacitance of the access transistor gates"""
-        #This is a handmade cell so the value must be entered in the tech.py file or estimated.
-        #Calculated in the tech file by summing the widths of all the related gates and dividing by the minimum width.
-        #FIXME: sizing is not accurate with the handmade cell. Change once cell widths are fixed.
+        
+        # FIXME: This applies to bitline capacitances as well.
+        # FIXME: sizing is not accurate with the handmade cell. Change once cell widths are fixed.
         access_tx_cin = parameter["6T_access_size"]/drc["minwidth_tx"]
         return 2*access_tx_cin
         
@@ -141,8 +141,8 @@ class bitcell_1rw_1r(design.design):
         pin_dict = {pin:port for pin,port in zip(self.pins, port_nets)} 
         #Edges hardcoded here. Essentially wl->bl/br for both ports.
         # Port 0 edges
-        graph.add_edge(pin_dict["wl0"], pin_dict["bl0"])   
-        graph.add_edge(pin_dict["wl0"], pin_dict["br0"])   
+        graph.add_edge(pin_dict["wl0"], pin_dict["bl0"], self)   
+        graph.add_edge(pin_dict["wl0"], pin_dict["br0"], self)   
         # Port 1 edges
-        graph.add_edge(pin_dict["wl1"], pin_dict["bl1"])   
-        graph.add_edge(pin_dict["wl1"], pin_dict["br1"])   
+        graph.add_edge(pin_dict["wl1"], pin_dict["bl1"], self)   
+        graph.add_edge(pin_dict["wl1"], pin_dict["br1"], self)   
diff --git a/compiler/bitcells/bitcell_1w_1r.py b/compiler/bitcells/bitcell_1w_1r.py
index 6063cf86..7e8c9e75 100644
--- a/compiler/bitcells/bitcell_1w_1r.py
+++ b/compiler/bitcells/bitcell_1w_1r.py
@@ -36,7 +36,7 @@ class bitcell_1w_1r(design.design):
         self.add_pin_types(self.type_list)
         self.nets_match = self.do_nets_exist(self.storage_nets)
 
-    def analytical_delay(self, corner, slew, load=0, swing = 0.5):
+    def get_stage_effort(self, load):
         parasitic_delay = 1
         size = 0.5 #This accounts for bitline being drained thought the access TX and internal node
         cin = 3 #Assumes always a minimum sizes inverter. Could be specified in the tech.py file.
@@ -116,11 +116,11 @@ class bitcell_1w_1r(design.design):
         total_power = self.return_power(dynamic, leakage)
         return total_power
 
-    def get_wl_cin(self):
+    def input_load(self):
         """Return the relative capacitance of the access transistor gates"""
-        #This is a handmade cell so the value must be entered in the tech.py file or estimated.
-        #Calculated in the tech file by summing the widths of all the related gates and dividing by the minimum width.
-        #FIXME: sizing is not accurate with the handmade cell. Change once cell widths are fixed.
+        
+        # FIXME: This applies to bitline capacitances as well.
+        # FIXME: sizing is not accurate with the handmade cell. Change once cell widths are fixed.
         access_tx_cin = parameter["6T_access_size"]/drc["minwidth_tx"]
         return 2*access_tx_cin
         
@@ -139,6 +139,6 @@ class bitcell_1w_1r(design.design):
         pin_dict = {pin:port for pin,port in zip(self.pins, port_nets)} 
         #Edges hardcoded here. Essentially wl->bl/br for both ports.
         # Port 0 edges
-        graph.add_edge(pin_dict["wl1"], pin_dict["bl1"])   
-        graph.add_edge(pin_dict["wl1"], pin_dict["br1"])     
+        graph.add_edge(pin_dict["wl1"], pin_dict["bl1"], self)   
+        graph.add_edge(pin_dict["wl1"], pin_dict["br1"], self)     
         # Port 1 is a write port, so its timing is not considered here.      
diff --git a/compiler/bitcells/pbitcell.py b/compiler/bitcells/pbitcell.py
index 3c59a9e3..919a0316 100644
--- a/compiler/bitcells/pbitcell.py
+++ b/compiler/bitcells/pbitcell.py
@@ -934,7 +934,7 @@ class pbitcell(design.design):
         return "wl{}".format(port)
     
 
-    def analytical_delay(self, corner, slew, load=0, swing = 0.5):
+    def get_stage_effort(self, load):
         parasitic_delay = 1
         size = 0.5 #This accounts for bitline being drained thought the access TX and internal node
         cin = 3 #Assumes always a minimum sizes inverter. Could be specified in the tech.py file.
@@ -954,9 +954,11 @@ class pbitcell(design.design):
         total_power = self.return_power(dynamic, leakage)
         return total_power
         
-    def get_wl_cin(self):
+    def input_load(self):
         """Return the relative capacitance of the access transistor gates"""
-        #pbitcell uses the different sizing for the port access tx's. Not accounted for in this model.
+        
+        # FIXME: This applies to bitline capacitances as well.
+        # pbitcell uses the different sizing for the port access tx's. Not accounted for in this model.
         access_tx_cin = self.readwrite_nmos.get_cin()
         return 2*access_tx_cin
 
@@ -973,6 +975,6 @@ class pbitcell(design.design):
 
         for pin_zip in [rw_pin_names, r_pin_names]: 
             for wl,bl,br in pin_zip:
-                graph.add_edge(pin_dict[wl],pin_dict[bl])
-                graph.add_edge(pin_dict[wl],pin_dict[br])
+                graph.add_edge(pin_dict[wl],pin_dict[bl], self)
+                graph.add_edge(pin_dict[wl],pin_dict[br], self)
 
diff --git a/compiler/bitcells/replica_bitcell.py b/compiler/bitcells/replica_bitcell.py
index f8ded969..2f804bf0 100644
--- a/compiler/bitcells/replica_bitcell.py
+++ b/compiler/bitcells/replica_bitcell.py
@@ -31,6 +31,20 @@ class replica_bitcell(design.design):
         self.height = replica_bitcell.height
         self.pin_map = replica_bitcell.pin_map
         self.add_pin_types(self.type_list)
+        
+    def get_stage_effort(self, load):
+        parasitic_delay = 1
+        size = 0.5 #This accounts for bitline being drained thought the access TX and internal node
+        cin = 3 #Assumes always a minimum sizes inverter. Could be specified in the tech.py file.
+        read_port_load = 0.5 #min size NMOS gate load
+        return logical_effort.logical_effort('bitline', size, cin, load+read_port_load, parasitic_delay, False)
+        
+    def input_load(self):
+        """Return the relative capacitance of the access transistor gates"""
+        
+        # FIXME: This applies to bitline capacitances as well.
+        access_tx_cin = parameter["6T_access_size"]/drc["minwidth_tx"]
+        return 2*access_tx_cin    
     
     def analytical_power(self, corner, load):
         """Bitcell power in nW. Only characterizes leakage."""
@@ -40,13 +54,6 @@ class replica_bitcell(design.design):
         total_power = self.return_power(dynamic, leakage)
         return total_power
 
-    def get_wl_cin(self):
-        """Return the relative capacitance of the access transistor gates"""
-        #This is a handmade cell so the value must be entered in the tech.py file or estimated.
-        #Calculated in the tech file by summing the widths of all the related gates and dividing by the minimum width.
-        access_tx_cin = parameter["6T_access_size"]/drc["minwidth_tx"]
-        return 2*access_tx_cin
-
     def build_graph(self, graph, inst_name, port_nets):        
         """Adds edges based on inputs/outputs. Overrides base class function."""
         self.add_graph_edges(graph, port_nets) 
\ No newline at end of file
diff --git a/compiler/bitcells/replica_bitcell_1rw_1r.py b/compiler/bitcells/replica_bitcell_1rw_1r.py
index e5c33317..0f56319e 100644
--- a/compiler/bitcells/replica_bitcell_1rw_1r.py
+++ b/compiler/bitcells/replica_bitcell_1rw_1r.py
@@ -32,11 +32,18 @@ class replica_bitcell_1rw_1r(design.design):
         self.pin_map = replica_bitcell_1rw_1r.pin_map
         self.add_pin_types(self.type_list)
 
-    def get_wl_cin(self):
+    def get_stage_effort(self, load):
+        parasitic_delay = 1
+        size = 0.5 #This accounts for bitline being drained thought the access TX and internal node
+        cin = 3 #Assumes always a minimum sizes inverter. Could be specified in the tech.py file.
+        read_port_load = 0.5 #min size NMOS gate load
+        return logical_effort.logical_effort('bitline', size, cin, load+read_port_load, parasitic_delay, False)
+        
+    def input_load(self):
         """Return the relative capacitance of the access transistor gates"""
-        #This is a handmade cell so the value must be entered in the tech.py file or estimated.
-        #Calculated in the tech file by summing the widths of all the related gates and dividing by the minimum width.
-        #FIXME: sizing is not accurate with the handmade cell. Change once cell widths are fixed.
+        
+        # FIXME: This applies to bitline capacitances as well.
+        # FIXME: sizing is not accurate with the handmade cell. Change once cell widths are fixed.
         access_tx_cin = parameter["6T_access_size"]/drc["minwidth_tx"]
         return 2*access_tx_cin
 
@@ -46,8 +53,8 @@ class replica_bitcell_1rw_1r(design.design):
         pin_dict = {pin:port for pin,port in zip(self.pins, port_nets)} 
         #Edges hardcoded here. Essentially wl->bl/br for both ports.
         # Port 0 edges
-        graph.add_edge(pin_dict["wl0"], pin_dict["bl0"])   
-        graph.add_edge(pin_dict["wl0"], pin_dict["br0"])   
+        graph.add_edge(pin_dict["wl0"], pin_dict["bl0"], self)   
+        graph.add_edge(pin_dict["wl0"], pin_dict["br0"], self)   
         # Port 1 edges
-        graph.add_edge(pin_dict["wl1"], pin_dict["bl1"])   
-        graph.add_edge(pin_dict["wl1"], pin_dict["br1"])  
\ No newline at end of file
+        graph.add_edge(pin_dict["wl1"], pin_dict["bl1"], self)   
+        graph.add_edge(pin_dict["wl1"], pin_dict["br1"], self)  
\ No newline at end of file
diff --git a/compiler/bitcells/replica_bitcell_1w_1r.py b/compiler/bitcells/replica_bitcell_1w_1r.py
index 79171bf5..b903e0ad 100644
--- a/compiler/bitcells/replica_bitcell_1w_1r.py
+++ b/compiler/bitcells/replica_bitcell_1w_1r.py
@@ -32,11 +32,18 @@ class replica_bitcell_1w_1r(design.design):
         self.pin_map = replica_bitcell_1w_1r.pin_map
         self.add_pin_types(self.type_list)
 
-    def get_wl_cin(self):
+    def get_stage_effort(self, load):
+        parasitic_delay = 1
+        size = 0.5 #This accounts for bitline being drained thought the access TX and internal node
+        cin = 3 #Assumes always a minimum sizes inverter. Could be specified in the tech.py file.
+        read_port_load = 0.5 #min size NMOS gate load
+        return logical_effort.logical_effort('bitline', size, cin, load+read_port_load, parasitic_delay, False)
+        
+    def input_load(self):
         """Return the relative capacitance of the access transistor gates"""
-        #This is a handmade cell so the value must be entered in the tech.py file or estimated.
-        #Calculated in the tech file by summing the widths of all the related gates and dividing by the minimum width.
-        #FIXME: sizing is not accurate with the handmade cell. Change once cell widths are fixed.
+        
+        # FIXME: This applies to bitline capacitances as well.
+        # FIXME: sizing is not accurate with the handmade cell. Change once cell widths are fixed.
         access_tx_cin = parameter["6T_access_size"]/drc["minwidth_tx"]
         return 2*access_tx_cin
 
@@ -47,6 +54,6 @@ class replica_bitcell_1w_1r(design.design):
         pin_dict = {pin:port for pin,port in zip(self.pins, port_nets)} 
         #Edges hardcoded here. Essentially wl->bl/br for the read port.
         # Port 1 edges
-        graph.add_edge(pin_dict["wl1"], pin_dict["bl1"])   
-        graph.add_edge(pin_dict["wl1"], pin_dict["br1"])   
+        graph.add_edge(pin_dict["wl1"], pin_dict["bl1"], self)   
+        graph.add_edge(pin_dict["wl1"], pin_dict["br1"], self)   
         # Port 0 is a write port, so its timing is not considered here.
\ No newline at end of file
diff --git a/compiler/bitcells/replica_pbitcell.py b/compiler/bitcells/replica_pbitcell.py
index 8453411f..4fcfb4c5 100644
--- a/compiler/bitcells/replica_pbitcell.py
+++ b/compiler/bitcells/replica_pbitcell.py
@@ -84,8 +84,4 @@ class replica_pbitcell(design.design):
             self.copy_layout_pin(self.prbc_inst, "wl{}".format(port))    
         self.copy_layout_pin(self.prbc_inst, "vdd")
         self.copy_layout_pin(self.prbc_inst, "gnd")
-    
-    def get_wl_cin(self):
-        """Return the relative capacitance of the access transistor gates"""
-        #This module is made using a pbitcell. Get the cin from that module
-        return self.prbc.get_wl_cin()
+    
\ No newline at end of file
diff --git a/compiler/characterizer/delay.py b/compiler/characterizer/delay.py
index bb48564e..65bc0b78 100644
--- a/compiler/characterizer/delay.py
+++ b/compiler/characterizer/delay.py
@@ -247,7 +247,7 @@ class delay(simulation):
     def create_graph(self):
         """Creates timing graph to generate the timing paths for the SRAM output."""
         
-        self.sram.bank.bitcell_array.init_graph_params() # Removes previous bit exclusions
+        self.sram.bank.bitcell_array.bitcell_array.init_graph_params() # Removes previous bit exclusions
         self.sram.bank.bitcell_array.graph_exclude_bits(self.wordline_row, self.bitline_column)
         
         # Generate new graph every analysis as edges might change depending on test bit
@@ -800,6 +800,8 @@ class delay(simulation):
 
             debug.info(2,"{}={}".format(meas.name,val))
 
+        dout_success = True
+        bl_success = False
         for meas in self.dout_volt_meas:
             val = meas.retrieve_measure(port=port)
             debug.info(2,"{}={}".format(meas.name, val))
@@ -813,15 +815,7 @@ class delay(simulation):
                 dout_success = False
                 debug.info(1, "Debug measurement failed. Value {}V was read on read 0 cycle.".format(val))
                 bl_success = self.check_bitline_meas(br_vals[sram_op.READ_ONE], bl_vals[sram_op.READ_ONE])
-            elif meas.meta_str == sram_op.READ_ONE and val > self.vdd_voltage*0.9:
-                dout_success = True
-                bl_success = True
-            elif meas.meta_str == sram_op.READ_ZERO and val < self.vdd_voltage*0.1:
-                dout_success = True
-                bl_success = True
-            else:
-                dout_success = False
-                bl_success = False
+                
             # If the bitlines have a correct value while the output does not then that is a 
             # sen error. FIXME: there are other checks that can be done to solidfy this conclusion.
             if not dout_success and bl_success:
@@ -1275,39 +1269,71 @@ class delay(simulation):
             # Add test cycle of read/write port pair. One port could have been used already, but the other has not.
             self.gen_test_cycles_one_port(cur_read_port, cur_write_port)
 
+    def sum_delays(self, delays):
+        """Adds the delays (delay_data objects) so the correct slew is maintained"""
+        
+        delay = delays[0]
+        for i in range(1, len(delays)):
+            delay+=delays[i]
+        return delay 
+        
     def analytical_delay(self, slews, loads):
         """ 
         Return the analytical model results for the SRAM. 
         """
         if OPTS.num_rw_ports > 1 or OPTS.num_w_ports > 0 and OPTS.num_r_ports > 0:
             debug.warning("Analytical characterization results are not supported for multiport.")
+        
+        # Probe set to 0th bit, does not matter for analytical delay.
+        self.set_probe('0', 0)
+        self.create_graph()
+        self.set_internal_spice_names()
         self.create_measurement_names()
-        power = self.analytical_power(slews, loads)
+        
+        port = self.read_ports[0]
+        self.graph.get_all_paths('{}{}'.format(tech.spice["clk"], port), 
+                                 '{}{}_{}'.format(self.dout_name, port, self.probe_data))
+        
+        # Select the path with the bitline (bl)
+        bl_name,br_name = self.get_bl_name(self.graph.all_paths, port)
+        bl_path = [path for path in self.graph.all_paths if bl_name in path][0]
+        
+        # Set delay/power for slews and loads
         port_data = self.get_empty_measure_data_dict()
-        relative_loads = [logical_effort.convert_farad_to_relative_c(c_farad) for c_farad in loads]
+        power = self.analytical_power(slews, loads)
+        debug.info(1,'Slew, Load, Delay(ns), Slew(ns)')
+        max_delay = 0.0
         for slew in slews:
-            for load in relative_loads:
-                self.set_load_slew(load,slew)
-                bank_delay = self.sram.analytical_delay(self.corner, self.slew,self.load)
+            for load in loads:
+                # Calculate delay based on slew and load
+                path_delays = self.graph.get_timing(bl_path, self.corner, slew, load)
+                
+                total_delay = self.sum_delays(path_delays)
+                max_delay = max(max_delay, total_delay.delay)
+                debug.info(1,'{}, {}, {}, {}'.format(slew,load,total_delay.delay/1e3, total_delay.slew/1e3))
+                
+                # Delay is only calculated on a single port and replicated for now.
                 for port in self.all_ports:
                     for mname in self.delay_meas_names+self.power_meas_names:
                         if "power" in mname:
                             port_data[port][mname].append(power.dynamic)
-                        elif "delay" in mname:
-                            port_data[port][mname].append(bank_delay[port].delay/1e3)
-                        elif "slew" in mname:
-                            port_data[port][mname].append(bank_delay[port].slew/1e3)
+                        elif "delay" in mname and port in self.read_ports:
+                            port_data[port][mname].append(total_delay.delay/1e3)
+                        elif "slew" in mname and port in self.read_ports:
+                            port_data[port][mname].append(total_delay.slew/1e3)
                         else:
                             debug.error("Measurement name not recognized: {}".format(mname),1)
+        
+        # Estimate the period as double the delay with margin
         period_margin = 0.1
-        risefall_delay = bank_delay[self.read_ports[0]].delay/1e3
-        sram_data = { "min_period":risefall_delay*2*period_margin, 
+        sram_data = { "min_period":(max_delay/1e3)*2*period_margin, 
                       "leakage_power": power.leakage}
+                      
         debug.info(2,"SRAM Data:\n{}".format(sram_data))                 
         debug.info(2,"Port Data:\n{}".format(port_data)) 
-        return (sram_data,port_data)
         
-    
+        return (sram_data,port_data)        
+
     def analytical_power(self, slews, loads):
         """Get the dynamic and leakage power from the SRAM"""
         
diff --git a/compiler/characterizer/logical_effort.py b/compiler/characterizer/logical_effort.py
index 024cd4f1..42ac8861 100644
--- a/compiler/characterizer/logical_effort.py
+++ b/compiler/characterizer/logical_effort.py
@@ -23,23 +23,23 @@ class logical_effort():
         self.cin = cin
         self.cout = cout
         self.logical_effort = (self.cin/size)/logical_effort.min_inv_cin
-        self.eletrical_effort = self.cout/self.cin
+        self.electrical_effort = self.cout/self.cin
         self.parasitic_scale = parasitic
         self.is_rise = out_is_rise
      
     def __str__(self):
         return  "Name={}, g={}, h={}, p={}*pinv, rise_delay={}".format(self.name,
                                                                        self.logical_effort,
-                                                                       self.eletrical_effort,
+                                                                       self.electrical_effort,
                                                                        self.parasitic_scale,
                                                                        self.is_rise
                                                                        ) 
 
     def get_stage_effort(self):
-        return  self.logical_effort*self.eletrical_effort
+        return  self.logical_effort*self.electrical_effort
         
     def get_parasitic_delay(self):
-        return logical_effort.pinv * self.parasitic_scale
+        return logical_effort.pinv*self.parasitic_scale
     
     def get_stage_delay(self):
         return self.get_stage_effort()+self.get_parasitic_delay()
@@ -78,4 +78,7 @@ def calculate_relative_rise_fall_delays(stage_effort_list):
 def convert_farad_to_relative_c(c_farad):
     """Converts capacitance in Femto-Farads to relative capacitance."""
     return c_farad*parameter['cap_relative_per_ff']
-   
\ No newline at end of file
+   
+def convert_relative_c_to_farad(c_relative):
+    """Converts capacitance in logical effort relative units to Femto-Farads."""
+    return c_relative/parameter['cap_relative_per_ff']   
\ No newline at end of file
diff --git a/compiler/modules/bank.py b/compiler/modules/bank.py
index 4fa95c44..a5ab4909 100644
--- a/compiler/modules/bank.py
+++ b/compiler/modules/bank.py
@@ -82,7 +82,7 @@ class bank(design.design):
         for port in self.read_ports:
             for bit in range(self.word_size):
                 self.add_pin("dout{0}_{1}".format(port,bit),"OUTPUT")
-        for port in self.read_ports:
+        for port in self.all_ports:
             self.add_pin(self.bitcell_array.get_rbl_bl_name(self.port_rbl_map[port]),"OUTPUT")
         for port in self.write_ports:
             for bit in range(self.word_size):
@@ -128,14 +128,13 @@ class bank(design.design):
     def route_rbl(self,port):
         """ Route the rbl_bl and rbl_wl """
         
-        if self.port_data[port].has_rbl():
-            bl_pin_name = self.bitcell_array.get_rbl_bl_name(self.port_rbl_map[port])
-            bl_pin = self.bitcell_array_inst.get_pin(bl_pin_name)
-            self.add_layout_pin(text="rbl_bl{0}".format(port),
-                                layer=bl_pin.layer,
-                                offset=bl_pin.ll(),
-                                height=bl_pin.height(),
-                                width=bl_pin.width())
+        bl_pin_name = self.bitcell_array.get_rbl_bl_name(self.port_rbl_map[port])
+        bl_pin = self.bitcell_array_inst.get_pin(bl_pin_name)
+        self.add_layout_pin(text="rbl_bl{0}".format(port),
+                            layer=bl_pin.layer,
+                            offset=bl_pin.ll(),
+                            height=bl_pin.height(),
+                            width=bl_pin.width())
             
             
             
@@ -202,10 +201,7 @@ class bank(design.design):
 
         # LOWER RIGHT QUADRANT
         # Below the bitcell array
-        if self.port_data[port].has_rbl():
-            self.port_data_offsets[port] = vector(self.main_bitcell_array_left - self.bitcell_array.cell.width,0)
-        else:
-            self.port_data_offsets[port] = vector(self.main_bitcell_array_left,0)
+        self.port_data_offsets[port] = vector(self.main_bitcell_array_left - self.bitcell_array.cell.width,0)
 
         # UPPER LEFT QUADRANT
         # To the left of the bitcell array
@@ -364,36 +360,15 @@ class bank(design.design):
         self.add_mod(self.port_address)
 
 
-        # The number of replica lines depends on the port configuration
-        rbl_counts = [self.read_ports.count(p) for p in self.all_ports]
-        self.num_rbl = sum(rbl_counts)
-            
-        # The replica array indices always start at 0, so this will map them to
-        # the correct SRAM port
-        # (e.g. if port 0 is w, then port 1 will use RBL 0 in replica bitcell array
-        # because write ports don't use an RBL)
-        self.port_rbl_map = {}
-        index = 0
-        for (i,num) in enumerate(rbl_counts):
-            if num>0:
-                self.port_rbl_map[i]=index
-                index += 1
-        if len(rbl_counts)<2:
-            rbl_counts.append(0)
-
-
-        # Which bitcell port should be used in the RBL
-        # For now (since only 2 ports), if port 0 is not a read port, skip it in the RBLs
-        bitcell_ports=list(range(len(self.read_ports)))
-        if 0 not in self.read_ports:
-            bitcell_ports = [x+1 for x in bitcell_ports]
+        self.port_rbl_map = self.all_ports
+        self.num_rbl = len(self.all_ports)
                 
         self.bitcell_array = factory.create(module_type="replica_bitcell_array",
                                             cols=self.num_cols,
                                             rows=self.num_rows,
-                                            left_rbl=rbl_counts[0],
-                                            right_rbl=rbl_counts[1],
-                                            bitcell_ports=bitcell_ports)
+                                            left_rbl=1,
+                                            right_rbl=1 if len(self.all_ports)>1 else 0,
+                                            bitcell_ports=self.all_ports)
         self.add_mod(self.bitcell_array)
 
         
@@ -421,8 +396,7 @@ class bank(design.design):
             for wordline in self.wl_names:
                     temp.append("{0}_{1}".format(wordline,row))
         for port in self.all_ports:
-            if self.port_data[port].has_rbl():
-                temp.append("wl_en{0}".format(port))
+            temp.append("wl_en{0}".format(port))
         temp.append("vdd")
         temp.append("gnd")
         self.connect_inst(temp)
@@ -442,11 +416,10 @@ class bank(design.design):
                                                     mod=self.port_data[port])
 
             temp = []
-            if self.port_data[port].has_rbl():
-                rbl_bl_name=self.bitcell_array.get_rbl_bl_name(self.port_rbl_map[port])
-                rbl_br_name=self.bitcell_array.get_rbl_br_name(self.port_rbl_map[port])
-                temp.append(rbl_bl_name)
-                temp.append(rbl_br_name)
+            rbl_bl_name=self.bitcell_array.get_rbl_bl_name(self.port_rbl_map[port])
+            rbl_br_name=self.bitcell_array.get_rbl_br_name(self.port_rbl_map[port])
+            temp.append(rbl_bl_name)
+            temp.append(rbl_br_name)
             for col in range(self.num_cols):            
                 temp.append("{0}_{1}".format(self.bl_names[port],col))
                 temp.append("{0}_{1}".format(self.br_names[port],col))
@@ -710,11 +683,10 @@ class bank(design.design):
                               inst1_bl_name=inst1_bl_name, inst1_br_name=inst1_br_name)
 
         # Connect the replica bitlines
-        if self.port_data[port].has_rbl():
-            rbl_bl_name=self.bitcell_array.get_rbl_bl_name(self.port_rbl_map[port])
-            rbl_br_name=self.bitcell_array.get_rbl_br_name(self.port_rbl_map[port])
-            self.connect_bitline(inst1, inst2, rbl_bl_name, "rbl_bl")
-            self.connect_bitline(inst1, inst2, rbl_br_name, "rbl_br")
+        rbl_bl_name=self.bitcell_array.get_rbl_bl_name(self.port_rbl_map[port])
+        rbl_br_name=self.bitcell_array.get_rbl_br_name(self.port_rbl_map[port])
+        self.connect_bitline(inst1, inst2, rbl_bl_name, "rbl_bl")
+        self.connect_bitline(inst1, inst2, rbl_br_name, "rbl_br")
         
         
         
@@ -943,9 +915,8 @@ class bank(design.design):
         connection = []
         connection.append((self.prefix+"p_en_bar{}".format(port), self.port_data_inst[port].get_pin("p_en_bar").lc()))
 
-        if port in self.read_ports:
-            rbl_wl_name = self.bitcell_array.get_rbl_wl_name(self.port_rbl_map[port])
-            connection.append((self.prefix+"wl_en{}".format(port), self.bitcell_array_inst.get_pin(rbl_wl_name).lc()))
+        rbl_wl_name = self.bitcell_array.get_rbl_wl_name(self.port_rbl_map[port])
+        connection.append((self.prefix+"wl_en{}".format(port), self.bitcell_array_inst.get_pin(rbl_wl_name).lc()))
             
         if port in self.write_ports:
             connection.append((self.prefix+"w_en{}".format(port), self.port_data_inst[port].get_pin("w_en").lc()))
@@ -974,40 +945,7 @@ class bank(design.design):
         self.add_wire(("metal1","via1","metal2"),[pin_pos, mid_pos, control_pos])
         self.add_via_center(layers=("metal1", "via1", "metal2"),
                             offset=control_pos)
-
-    def analytical_delay(self, corner, slew, load, port):
-        """ return  analytical delay of the bank. This will track the clock to output path"""
-        #FIXME: This delay is determined in the control logic. Should be moved here.
-        # word_driver_delay = self.wordline_driver.analytical_delay(corner, 
-                                                                  # slew, 
-                                                                  # self.bitcell_array.input_load())
-
-        #FIXME: Array delay is the same for every port.
-        word_driver_slew =  0
-        if self.words_per_row > 1:
-            bitline_ext_load = self.port_data[port].column_mux_array.get_drain_cin()
-        else:
-            bitline_ext_load = self.port_data[port].sense_amp_array.get_drain_cin()
-            
-        bitcell_array_delay = self.bitcell_array.analytical_delay(corner, word_driver_slew, bitline_ext_load)
-
-        bitcell_array_slew = 0
-        #This also essentially creates the same delay for each port. Good structure, no substance
-        if self.words_per_row > 1:
-            sa_load = self.port_data[port].sense_amp_array.get_drain_cin()
-            column_mux_delay = self.port_data[port].column_mux_array.analytical_delay(corner, 
-                                                                                            bitcell_array_slew,
-                                                                                            sa_load)
-        else:
-            column_mux_delay = []  
-            
-        column_mux_slew =  0    
-        sense_amp_delay = self.port_data[port].sense_amp_array.analytical_delay(corner, 
-                                                                                      column_mux_slew,
-                                                                                      load)
-        # output load of bitcell_array is set to be only small part of bl for sense amp.
-        return  bitcell_array_delay + column_mux_delay + sense_amp_delay
-    
+ 
     def determine_wordline_stage_efforts(self, external_cout, inp_is_rise=True):    
         """Get the all the stage efforts for each stage in the path within the bank clk_buf to a wordline"""
         #Decoder is assumed to have settled before the negative edge of the clock. Delay model relies on this assumption
diff --git a/compiler/modules/bitcell_array.py b/compiler/modules/bitcell_array.py
index 999382ec..bf4932a3 100644
--- a/compiler/modules/bitcell_array.py
+++ b/compiler/modules/bitcell_array.py
@@ -148,17 +148,7 @@ class bitcell_array(design.design):
                 for pin_name in ["vdd", "gnd"]:
                     for pin in inst.get_pins(pin_name):
                         self.add_power_pin(name=pin_name, loc=pin.center(), vertical=True, start_layer=pin.layer)
-    
-    def analytical_delay(self, corner, slew, load):
-        """Returns relative delay of the bitline in the bitcell array"""
-        from tech import parameter
-        #The load being driven/drained is mostly the bitline but could include the sense amp or the column mux.
-        #The load from the bitlines is due to the drain capacitances from all the other bitlines and wire parasitics.
-        drain_load = logical_effort.convert_farad_to_relative_c(parameter['bitcell_drain_cap'])
-        wire_unit_load = .05 * drain_load #Wires add 5% to this.
-        bitline_load = (drain_load+wire_unit_load)*self.row_size
-        return [self.cell.analytical_delay(corner, slew, load+bitline_load)]
-    
+
     def analytical_power(self, corner, load):
         """Power of Bitcell array and bitline in nW."""
         from tech import drc, parameter
@@ -197,15 +187,6 @@ class bitcell_array(design.design):
         bl_wire.wire_c =spice["min_tx_drain_c"] + bl_wire.wire_c # 1 access tx d/s per cell
         return bl_wire
 
-    def output_load(self, bl_pos=0):
-        bl_wire = self.gen_bl_wire()
-        return bl_wire.wire_c # sense amp only need to charge small portion of the bl
-                              # set as one segment for now
-
-    def input_load(self):
-        wl_wire = self.gen_wl_wire()
-        return wl_wire.return_input_cap()
-
     def get_wordline_cin(self):
         """Get the relative input capacitance from the wordline connections in all the bitcell"""
         #A single wordline is connected to all the bitcells in a single row meaning the capacitance depends on the # of columns
diff --git a/compiler/modules/control_logic.py b/compiler/modules/control_logic.py
index a5c5dc12..06883125 100644
--- a/compiler/modules/control_logic.py
+++ b/compiler/modules/control_logic.py
@@ -125,33 +125,35 @@ class control_logic(design.design):
         self.add_mod(self.wl_en_driver)
 
         # w_en drives every write driver
-        self.wen_and2 = factory.create(module_type="pand2",
+        self.wen_and = factory.create(module_type="pand3",
                                        size=self.word_size+8,
                                        height=dff_height)
-        self.add_mod(self.wen_and2)
+        self.add_mod(self.wen_and)
 
         # s_en drives every sense amp
-        self.sen_and2 = factory.create(module_type="pand2",
+        self.sen_and3 = factory.create(module_type="pand3",
                                        size=self.word_size,
                                        height=dff_height)
-        self.add_mod(self.sen_and2)
+        self.add_mod(self.sen_and3)
 
         # used to generate inverted signals with low fanout 
         self.inv = factory.create(module_type="pinv",
                                   size=1,
                                   height=dff_height)
         self.add_mod(self.inv)
-
+        
         # p_en_bar drives every column in the bitcell array
         # but it is sized the same as the wl_en driver with
         # prepended 3 inverter stages to guarantee it is slower and odd polarity
         self.p_en_bar_driver = factory.create(module_type="pdriver",
                                               fanout=self.num_cols,
-                                              neg_polarity=True,
                                               height=dff_height)
         self.add_mod(self.p_en_bar_driver)
 
-        
+
+        self.nand2 = factory.create(module_type="pnand2",
+                                    height=dff_height)
+        self.add_mod(self.nand2)
         
         # if (self.port_type == "rw") or (self.port_type == "r"):
         #     from importlib import reload
@@ -330,12 +332,9 @@ class control_logic(design.design):
         if self.port_type == "rw":
             self.input_list = ["csb", "web"]
             self.rbl_list = ["rbl_bl"]
-        elif self.port_type == "r":
-            self.input_list = ["csb"]
-            self.rbl_list = ["rbl_bl"]
         else:
             self.input_list = ["csb"]
-            self.rbl_list = []
+            self.rbl_list = ["rbl_bl"]
 
         if self.port_type == "rw":
             self.dff_output_list = ["cs_bar", "cs", "we_bar", "we"]
@@ -344,11 +343,11 @@ class control_logic(design.design):
         
         # list of output control signals (for making a vertical bus)
         if self.port_type == "rw":
-            self.internal_bus_list = ["gated_clk_bar", "gated_clk_buf", "we", "clk_buf", "we_bar", "cs"]
+            self.internal_bus_list = ["rbl_bl_delay_bar", "rbl_bl_delay", "gated_clk_bar", "gated_clk_buf", "we", "clk_buf", "we_bar", "cs"]
         elif self.port_type == "r":
-            self.internal_bus_list = ["gated_clk_bar", "gated_clk_buf", "clk_buf", "cs_bar", "cs"]
+            self.internal_bus_list = ["rbl_bl_delay_bar", "rbl_bl_delay", "gated_clk_bar", "gated_clk_buf", "clk_buf", "cs_bar", "cs"]
         else:
-            self.internal_bus_list = ["gated_clk_bar", "gated_clk_buf", "clk_buf", "cs"]
+            self.internal_bus_list = ["rbl_bl_delay_bar", "rbl_bl_delay", "gated_clk_bar", "gated_clk_buf", "clk_buf", "cs"]
         # leave space for the bus plus one extra space
         self.internal_bus_width = (len(self.internal_bus_list)+1)*self.m2_pitch 
         
@@ -382,10 +381,11 @@ class control_logic(design.design):
         self.create_gated_clk_buf_row()
         self.create_wlen_row()
         if (self.port_type == "rw") or (self.port_type == "w"):
+            self.create_rbl_delay_row()
             self.create_wen_row()
         if (self.port_type == "rw") or (self.port_type == "r"):
             self.create_sen_row()
-            self.create_delay()
+        self.create_delay()
         self.create_pen_row()            
 
 
@@ -419,12 +419,15 @@ class control_logic(design.design):
             row += 1
         self.place_pen_row(row)
         row += 1
+        if (self.port_type == "rw") or (self.port_type == "w"):        
+            self.place_rbl_delay_row(row)
+            row += 1
         if (self.port_type == "rw") or (self.port_type == "r"):            
             self.place_sen_row(row)
             row += 1
-            self.place_delay(row)
-            height = self.delay_inst.uy()
-            control_center_y = self.delay_inst.by()
+        self.place_delay(row)
+        height = self.delay_inst.uy()
+        control_center_y = self.delay_inst.by()
 
         # This offset is used for placement of the control logic in the SRAM level.
         self.control_logic_center = vector(self.ctrl_dff_inst.rx(), control_center_y)
@@ -443,9 +446,11 @@ class control_logic(design.design):
         self.route_dffs()
         self.route_wlen()
         if (self.port_type == "rw") or (self.port_type == "w"):
+            self.route_rbl_delay()
             self.route_wen()
         if (self.port_type == "rw") or (self.port_type == "r"):
             self.route_sen()
+        self.route_delay()
         self.route_pen()
         self.route_clk_buf()
         self.route_gated_clk_bar()
@@ -457,7 +462,7 @@ class control_logic(design.design):
         """ Create the replica bitline """
         self.delay_inst=self.add_inst(name="delay_chain",
                                       mod=self.delay_chain)
-        self.connect_inst(["rbl_bl", "pre_s_en", "vdd", "gnd"])
+        self.connect_inst(["rbl_bl", "rbl_bl_delay", "vdd", "gnd"])
 
     def place_delay(self,row):
         """ Place the replica bitline """
@@ -468,6 +473,22 @@ class control_logic(design.design):
         offset = vector(self.delay_chain.width, y_off)
         self.delay_inst.place(offset, mirror="MY")
         
+    def route_delay(self):
+
+        out_pos = self.delay_inst.get_pin("out").bc()
+        # Connect to the rail level with the vdd rail
+        # Use pen since it is in every type of control logic
+        vdd_ypos = self.p_en_bar_nand_inst.get_pin("vdd").by()
+        in_pos = vector(self.rail_offsets["rbl_bl_delay"].x,vdd_ypos)
+        mid1 = vector(out_pos.x,in_pos.y)
+        self.add_wire(("metal1","via1","metal2"),[out_pos, mid1, in_pos])
+        self.add_via_center(layers=("metal1","via1","metal2"),
+                            offset=in_pos)
+        
+
+        # Input from RBL goes to the delay line for futher delay
+        self.copy_layout_pin(self.delay_inst, "in", "rbl_bl")
+        
         
     def create_clk_buf_row(self):
         """ Create the multistage and gated clock buffer  """
@@ -476,12 +497,9 @@ class control_logic(design.design):
         self.connect_inst(["clk","clk_buf","vdd","gnd"])
         
     def place_clk_buf_row(self,row):
-        """ Place the multistage clock buffer below the control flops """
-        x_off = self.control_x_offset
-        (y_off,mirror)=self.get_offset(row)
+        x_offset = self.control_x_offset
         
-        offset = vector(x_off,y_off)
-        self.clk_buf_inst.place(offset, mirror)
+        x_offset = self.place_util(self.clk_buf_inst, x_offset, row)
         
         self.row_end_inst.append(self.clk_buf_inst)
 
@@ -518,17 +536,10 @@ class control_logic(design.design):
         self.connect_inst(["cs","clk_bar","gated_clk_bar","vdd","gnd"])
 
     def place_gated_clk_bar_row(self,row):
-        """ Place the gated clk logic below the control flops """
-        x_off = self.control_x_offset
-        (y_off,mirror)=self.get_offset(row)
+        x_offset = self.control_x_offset
         
-        offset = vector(x_off,y_off)
-        self.clk_bar_inst.place(offset, mirror)
-        
-        x_off += self.inv.width
-        
-        offset = vector(x_off,y_off)
-        self.gated_clk_bar_inst.place(offset, mirror)
+        x_offset = self.place_util(self.clk_bar_inst, x_offset, row)
+        x_offset = self.place_util(self.gated_clk_bar_inst, x_offset, row)        
         
         self.row_end_inst.append(self.gated_clk_bar_inst)
 
@@ -562,12 +573,9 @@ class control_logic(design.design):
         self.connect_inst(["clk_buf", "cs","gated_clk_buf","vdd","gnd"])
 
     def place_gated_clk_buf_row(self,row):
-        """ Place the gated clk logic below the control flops """
-        x_off = self.control_x_offset
-        (y_off,mirror)=self.get_offset(row)
-        
-        offset = vector(x_off,y_off)
-        self.gated_clk_buf_inst.place(offset, mirror)
+        x_offset = self.control_x_offset
+
+        x_offset = self.place_util(self.gated_clk_buf_inst, x_offset, row)
         
         self.row_end_inst.append(self.gated_clk_buf_inst)
         
@@ -589,11 +597,9 @@ class control_logic(design.design):
         self.connect_inst(["gated_clk_bar", "wl_en", "vdd", "gnd"])
 
     def place_wlen_row(self, row):
-        x_off = self.control_x_offset
-        (y_off,mirror)=self.get_offset(row)
-        
-        offset = vector(x_off, y_off)
-        self.wl_en_inst.place(offset, mirror)
+        x_offset = self.control_x_offset
+
+        x_offset = self.place_util(self.wl_en_inst, x_offset, row)        
 
         self.row_end_inst.append(self.wl_en_inst)
 
@@ -604,25 +610,32 @@ class control_logic(design.design):
         self.connect_output(self.wl_en_inst, "Z", "wl_en")
 
     def create_pen_row(self):
-        # input: gated_clk_bar, output: p_en_bar
-        self.p_en_bar_inst=self.add_inst(name="inv_p_en_bar",
-                                         mod=self.p_en_bar_driver)
-        self.connect_inst(["gated_clk_buf", "p_en_bar", "vdd", "gnd"])
-        
+        self.p_en_bar_nand_inst=self.add_inst(name="nand_p_en_bar",
+                                              mod=self.nand2)
+        self.connect_inst(["gated_clk_buf", "rbl_bl_delay", "p_en_bar_unbuf", "vdd", "gnd"])
+
+        self.p_en_bar_driver_inst=self.add_inst(name="buf_p_en_bar",
+                                                mod=self.p_en_bar_driver)
+        self.connect_inst(["p_en_bar_unbuf", "p_en_bar", "vdd", "gnd"])
         
     def place_pen_row(self,row):
-        x_off = self.control_x_offset
-        (y_off,mirror)=self.get_offset(row)
-        offset = vector(x_off,y_off)
-        self.p_en_bar_inst.place(offset, mirror)
+        x_offset = self.control_x_offset
 
-        self.row_end_inst.append(self.p_en_bar_inst)
+        x_offset = self.place_util(self.p_en_bar_nand_inst, x_offset, row)        
+        x_offset = self.place_util(self.p_en_bar_driver_inst, x_offset, row)        
+
+        self.row_end_inst.append(self.p_en_bar_driver_inst)
 
     def route_pen(self):
-        in_map = zip(["A"], ["gated_clk_buf"])
-        self.connect_vertical_bus(in_map, self.p_en_bar_inst, self.rail_offsets)  
+        in_map = zip(["A", "B"], ["gated_clk_buf", "rbl_bl_delay"])
+        self.connect_vertical_bus(in_map, self.p_en_bar_nand_inst, self.rail_offsets)
+
+        out_pos = self.p_en_bar_nand_inst.get_pin("Z").rc()
+        in_pos = self.p_en_bar_driver_inst.get_pin("A").lc()
+        mid1 = vector(out_pos.x,in_pos.y)
+        self.add_wire(("metal1","via1","metal2"),[out_pos, mid1,in_pos])                
         
-        self.connect_output(self.p_en_bar_inst, "Z", "p_en_bar")
+        self.connect_output(self.p_en_bar_driver_inst, "Z", "p_en_bar")
         
     def create_sen_row(self):
         """ Create the sense enable buffer. """
@@ -632,20 +645,14 @@ class control_logic(design.design):
             input_name = "cs_bar"
         # GATE FOR S_EN
         self.s_en_gate_inst = self.add_inst(name="buf_s_en_and",
-                                            mod=self.sen_and2)
-        self.connect_inst(["pre_s_en", input_name, "s_en", "vdd", "gnd"])
+                                            mod=self.sen_and3)
+        self.connect_inst(["rbl_bl_delay", "gated_clk_bar", input_name, "s_en", "vdd", "gnd"])
         
         
     def place_sen_row(self,row):
-        """ 
-        The sense enable buffer gets placed to the far right of the 
-        row. 
-        """
-        x_off = self.control_x_offset
-        (y_off,mirror)=self.get_offset(row)
+        x_offset = self.control_x_offset
 
-        offset = vector(x_off, y_off)
-        self.s_en_gate_inst.place(offset, mirror)
+        x_offset = self.place_util(self.s_en_gate_inst, x_offset, row)
         
         self.row_end_inst.append(self.s_en_gate_inst)
 
@@ -657,22 +664,41 @@ class control_logic(design.design):
         else:
             input_name = "cs_bar"
             
-        sen_map = zip(["B"], [input_name])
+        sen_map = zip(["A", "B", "C"], ["rbl_bl_delay", "gated_clk_bar", input_name])
         self.connect_vertical_bus(sen_map, self.s_en_gate_inst, self.rail_offsets)
         
-        out_pos = self.delay_inst.get_pin("out").bc()
-        in_pos = self.s_en_gate_inst.get_pin("A").lc()
-        mid1 = vector(out_pos.x,in_pos.y)
-        self.add_wire(("metal1","via1","metal2"),[out_pos, mid1,in_pos])                
-
         self.connect_output(self.s_en_gate_inst, "Z", "s_en")
 
-        # Input from RBL goes to the delay line for futher delay
-        self.copy_layout_pin(self.delay_inst, "in", "rbl_bl")
-
         
+    def create_rbl_delay_row(self):
+
+        self.rbl_bl_delay_inv_inst = self.add_inst(name="rbl_bl_delay_inv",
+                                                   mod=self.inv)
+        self.connect_inst(["rbl_bl_delay", "rbl_bl_delay_bar", "vdd", "gnd"])
+
+    def place_rbl_delay_row(self,row):
+        x_offset = self.control_x_offset
+
+        x_offset = self.place_util(self.rbl_bl_delay_inv_inst, x_offset, row)
+        
+        self.row_end_inst.append(self.rbl_bl_delay_inv_inst)
+        
+    def route_rbl_delay(self):
+        # Connect from delay line
+        # Connect to rail
+
+        rbl_map = zip(["Z"], ["rbl_bl_delay_bar"])
+        self.connect_vertical_bus(rbl_map, self.rbl_bl_delay_inv_inst, self.rail_offsets, ("metal3", "via2", "metal2"))
+        # The pin is on M1, so we need another via as well
+        self.add_via_center(layers=("metal1","via1","metal2"),
+                            offset=self.rbl_bl_delay_inv_inst.get_pin("Z").center())
+        
+        
+        rbl_map = zip(["A"], ["rbl_bl_delay"])
+        self.connect_vertical_bus(rbl_map, self.rbl_bl_delay_inv_inst, self.rail_offsets)
         
     def create_wen_row(self):
+
         # input: we (or cs) output: w_en
         if self.port_type == "rw":
             input_name = "we"
@@ -682,16 +708,14 @@ class control_logic(design.design):
 
         # GATE THE W_EN
         self.w_en_gate_inst = self.add_inst(name="w_en_and",
-                                            mod=self.wen_and2)
-        self.connect_inst([input_name, "gated_clk_bar", "w_en", "vdd", "gnd"])
+                                            mod=self.wen_and)
+        self.connect_inst([input_name, "rbl_bl_delay_bar", "gated_clk_bar", "w_en", "vdd", "gnd"])
         
 
     def place_wen_row(self,row):
-        x_off = self.ctrl_dff_inst.width + self.internal_bus_width
-        (y_off,mirror)=self.get_offset(row)
-            
-        offset = vector(x_off, y_off)
-        self.w_en_gate_inst.place(offset, mirror)
+        x_offset = self.control_x_offset
+
+        x_offset = self.place_util(self.w_en_gate_inst, x_offset, row)
         
         self.row_end_inst.append(self.w_en_gate_inst)
         
@@ -702,7 +726,7 @@ class control_logic(design.design):
             # No we for write-only reports, so use cs
             input_name = "cs"
             
-        wen_map = zip(["A", "B"], [input_name, "gated_clk_bar"])
+        wen_map = zip(["A", "B", "C"], [input_name, "rbl_bl_delay_bar", "gated_clk_bar"])
         self.connect_vertical_bus(wen_map, self.w_en_gate_inst, self.rail_offsets)
 
         self.connect_output(self.w_en_gate_inst, "Z", "w_en")
@@ -781,9 +805,8 @@ class control_logic(design.design):
                     self.add_power_pin("gnd", pin_loc)
                     self.add_path("metal1", [row_loc, pin_loc])
             
-        if (self.port_type == "rw") or (self.port_type == "r"):
-            self.copy_layout_pin(self.delay_inst,"gnd")
-            self.copy_layout_pin(self.delay_inst,"vdd")        
+        self.copy_layout_pin(self.delay_inst,"gnd")
+        self.copy_layout_pin(self.delay_inst,"vdd")        
 
         self.copy_layout_pin(self.ctrl_dff_inst,"gnd")
         self.copy_layout_pin(self.ctrl_dff_inst,"vdd")        
@@ -901,7 +924,8 @@ class control_logic(design.design):
         last_stage_rise = False 
         
         #First stage(s), clk -(pdriver)-> clk_buf. 
-        clk_buf_cout = self.replica_bitline.get_en_cin()
+        #clk_buf_cout = self.replica_bitline.get_en_cin()
+        clk_buf_cout = 0
         stage_effort_list += self.clk_buf_driver.get_stage_efforts(clk_buf_cout, last_stage_rise)
         last_stage_rise = stage_effort_list[-1].is_rise
         
@@ -948,3 +972,11 @@ class control_logic(design.design):
         self.graph_inst_exclude.add(self.ctrl_dff_inst)
         if self.port_type=="rw" or self.port_type=="w":
             self.graph_inst_exclude.add(self.w_en_gate_inst)
+
+    def place_util(self, inst, x_offset, row):
+        """ Utility to place a row and compute the next offset """
+
+        (y_offset,mirror)=self.get_offset(row)
+        offset = vector(x_offset, y_offset)
+        inst.place(offset, mirror)
+        return x_offset+inst.width
diff --git a/compiler/modules/dff.py b/compiler/modules/dff.py
index c8d7414a..083d92b7 100644
--- a/compiler/modules/dff.py
+++ b/compiler/modules/dff.py
@@ -48,11 +48,6 @@ class dff(design.design):
         transition_prob = spice["flop_transition_prob"]
         return transition_prob*(c_load + c_para) 
 
-    def analytical_delay(self, corner, slew, load = 0.0):
-        # dont know how to calculate this now, use constant in tech file
-        result = self.return_delay(spice["dff_delay"], spice["dff_slew"])
-        return result
-        
     def get_clk_cin(self):
         """Return the total capacitance (in relative units) that the clock is loaded by in the dff"""
         #This is a handmade cell so the value must be entered in the tech.py file or estimated.
diff --git a/compiler/modules/dff_array.py b/compiler/modules/dff_array.py
index 32b36765..89b29476 100644
--- a/compiler/modules/dff_array.py
+++ b/compiler/modules/dff_array.py
@@ -160,11 +160,6 @@ class dff_array(design.design):
             self.add_via_center(layers=("metal2","via2","metal3"),
                                 offset=vector(clk_pin.cx(),clk_ypos))
             
-        
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        return self.dff.analytical_delay(corner, slew=slew, load=load)
-
     def get_clk_cin(self):
         """Return the total capacitance (in relative units) that the clock is loaded by in the dff array"""
         dff_clk_cin = self.dff.get_clk_cin()
diff --git a/compiler/modules/dff_buf.py b/compiler/modules/dff_buf.py
index 9359329a..f6fc1cf2 100644
--- a/compiler/modules/dff_buf.py
+++ b/compiler/modules/dff_buf.py
@@ -177,16 +177,7 @@ class dff_buf(design.design):
         self.add_path("metal1", [self.mid_qb_pos, qb_pos])
         self.add_via_center(layers=("metal1","via1","metal2"),
                             offset=qb_pos)
-        
-        
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        """ Calculate the analytical delay of DFF-> INV -> INV """
-        dff_delay=self.dff.analytical_delay(corner, slew=slew, load=self.inv1.input_load())
-        inv1_delay = self.inv1.analytical_delay(corner, slew=dff_delay.slew, load=self.inv2.input_load()) 
-        inv2_delay = self.inv2.analytical_delay(corner, slew=inv1_delay.slew, load=load)
-        return dff_delay + inv1_delay + inv2_delay
-            
+         
     def get_clk_cin(self):
         """Return the total capacitance (in relative units) that the clock is loaded by in the dff"""
         #This is a handmade cell so the value must be entered in the tech.py file or estimated.
diff --git a/compiler/modules/dff_buf_array.py b/compiler/modules/dff_buf_array.py
index df36b2aa..2fafee76 100644
--- a/compiler/modules/dff_buf_array.py
+++ b/compiler/modules/dff_buf_array.py
@@ -193,11 +193,6 @@ class dff_buf_array(design.design):
                 self.add_via_center(layers=("metal2","via2","metal3"),
                                     offset=vector(clk_pin.cx(),clk_ypos))
                 
-        
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        return self.dff.analytical_delay(slew=slew, load=load)
-
     def get_clk_cin(self):
         """Return the total capacitance (in relative units) that the clock is loaded by in the dff array"""
         dff_clk_cin = self.dff.get_clk_cin()
diff --git a/compiler/modules/dff_inv.py b/compiler/modules/dff_inv.py
index 226db290..207a5ad0 100644
--- a/compiler/modules/dff_inv.py
+++ b/compiler/modules/dff_inv.py
@@ -150,15 +150,7 @@ class dff_inv(design.design):
                                         offset=dout_pin.center())
         self.add_via_center(layers=("metal1","via1","metal2"),
                             offset=dout_pin.center())
-        
-        
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        """ Calculate the analytical delay of DFF-> INV -> INV """
-        dff_delay=self.dff.analytical_delay(corner, slew=slew, load=self.inv1.input_load())
-        inv1_delay = self.inv1.analytical_delay(corner, slew=dff_delay.slew, load=load) 
-        return dff_delay + inv1_delay 
-            
+    
     def get_clk_cin(self):
         """Return the total capacitance (in relative units) that the clock is loaded by in the dff"""
         return self.dff.get_clk_cin()
diff --git a/compiler/modules/dff_inv_array.py b/compiler/modules/dff_inv_array.py
index 051dd237..760a2337 100644
--- a/compiler/modules/dff_inv_array.py
+++ b/compiler/modules/dff_inv_array.py
@@ -190,12 +190,6 @@ class dff_inv_array(design.design):
                 self.add_via_center(layers=("metal2","via2","metal3"),
                                     offset=vector(clk_pin.cx(),clk_ypos))
                 
-                
-        
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        return self.dff.analytical_delay(corner, slew=slew, load=load)
-
     def get_clk_cin(self):
         """Return the total capacitance (in relative units) that the clock is loaded by in the dff array"""
         dff_clk_cin = self.dff.get_clk_cin()
diff --git a/compiler/modules/hierarchical_decoder.py b/compiler/modules/hierarchical_decoder.py
index 75eb3345..0e70db55 100644
--- a/compiler/modules/hierarchical_decoder.py
+++ b/compiler/modules/hierarchical_decoder.py
@@ -596,28 +596,6 @@ class hierarchical_decoder(design.design):
         self.add_via_center(layers=("metal2", "via2", "metal3"),
                             offset=rail_pos)
 
-        
-    def analytical_delay(self, corner, slew, load = 0.0):
-        # A -> out
-        if self.determine_predecodes(self.num_inputs)[1]==0:
-            pre = self.pre2_4
-            nand = self.nand2
-        else:
-            pre = self.pre3_8
-            nand = self.nand3
-        a_t_out_delay = pre.analytical_delay(corner, slew=slew,load = nand.input_load())
-
-        # out -> z
-        out_t_z_delay = nand.analytical_delay(corner, slew= a_t_out_delay.slew,
-                                  load = self.inv.input_load())
-        result = a_t_out_delay + out_t_z_delay
-
-        # Z -> decode_out
-        z_t_decodeout_delay = self.inv.analytical_delay(corner, slew = out_t_z_delay.slew , load = load)
-        result = result + z_t_decodeout_delay
-        return result
-
-        
     def input_load(self):
         if self.determine_predecodes(self.num_inputs)[1]==0:
             pre = self.pre2_4
diff --git a/compiler/modules/hierarchical_predecode2x4.py b/compiler/modules/hierarchical_predecode2x4.py
index 56046c74..f05f54b0 100644
--- a/compiler/modules/hierarchical_predecode2x4.py
+++ b/compiler/modules/hierarchical_predecode2x4.py
@@ -56,21 +56,4 @@ class hierarchical_predecode2x4(hierarchical_predecode):
                        ["A_0",    "Abar_1"],
                        ["Abar_0", "A_1"],
                        ["A_0",    "A_1"]]
-        return combination 
-
-
-    def analytical_delay(self, corner, slew, load = 0.0 ):
-        # in -> inbar
-        a_t_b_delay = self.inv.analytical_delay(corner, slew=slew, load=self.nand.input_load())
-
-        # inbar -> z
-        b_t_z_delay = self.nand.analytical_delay(corner, slew=a_t_b_delay.slew, load=self.inv.input_load())
-
-        # Z -> out
-        a_t_out_delay = self.inv.analytical_delay(corner, slew=b_t_z_delay.slew, load=load)
-
-        return a_t_b_delay + b_t_z_delay + a_t_out_delay
-
-        
-    def input_load(self):
-        return self.nand.input_load()
+        return combination 
\ No newline at end of file
diff --git a/compiler/modules/hierarchical_predecode3x8.py b/compiler/modules/hierarchical_predecode3x8.py
index a7794395..20b629bb 100644
--- a/compiler/modules/hierarchical_predecode3x8.py
+++ b/compiler/modules/hierarchical_predecode3x8.py
@@ -65,21 +65,4 @@ class hierarchical_predecode3x8(hierarchical_predecode):
                        ["A_0",    "Abar_1", "A_2"], 
                        ["Abar_0", "A_1",    "A_2"], 
                        ["A_0",    "A_1",    "A_2"]]
-        return combination
-
-
-    def analytical_delay(self, corner, slew, load = 0.0 ):
-        # A -> Abar
-        a_t_b_delay = self.inv.analytical_delay(corner, slew=slew, load=self.nand.input_load())
-
-        # Abar -> z
-        b_t_z_delay = self.nand.analytical_delay(corner, slew=a_t_b_delay.slew, load=self.inv.input_load())
-
-        # Z -> out
-        a_t_out_delay = self.inv.analytical_delay(corner, slew=b_t_z_delay.slew, load=load)
-
-        return a_t_b_delay + b_t_z_delay + a_t_out_delay
-
-
-    def input_load(self):
-        return self.nand.input_load()
+        return combination
\ No newline at end of file
diff --git a/compiler/modules/multibank.py b/compiler/modules/multibank.py
index fd061273..28f689f7 100644
--- a/compiler/modules/multibank.py
+++ b/compiler/modules/multibank.py
@@ -827,22 +827,4 @@ class multibank(design.design):
                      rotate=90)
         self.add_via(layers=("metal2","via2","metal3"),
                      offset=in_pin + self.m2m3_via_offset,
-                     rotate=90)
-        
-    def analytical_delay(self, corner, slew, load):
-        """ return  analytical delay of the bank"""
-        decoder_delay = self.row_decoder.analytical_delay(corner, slew, self.wordline_driver.input_load())
-
-        word_driver_delay = self.wordline_driver.analytical_delay(corner, decoder_delay.slew, self.bitcell_array.input_load())
-
-        bitcell_array_delay = self.bitcell_array.analytical_delay(corner, word_driver_delay.slew)
-
-        bl_t_data_out_delay = self.sense_amp_array.analytical_delay(corner, bitcell_array_delay.slew,
-                                                                    self.bitcell_array.output_load())
-        # output load of bitcell_array is set to be only small part of bl for sense amp.
-
-        data_t_DATA_delay = self.tri_gate_array.analytical_delay(corner, bl_t_data_out_delay.slew, load)
-
-        result = decoder_delay + word_driver_delay + bitcell_array_delay + bl_t_data_out_delay + data_t_DATA_delay
-        return result
-        
+                     rotate=90)
\ No newline at end of file
diff --git a/compiler/modules/port_data.py b/compiler/modules/port_data.py
index 26104593..2eed4bb9 100644
--- a/compiler/modules/port_data.py
+++ b/compiler/modules/port_data.py
@@ -82,9 +82,8 @@ class port_data(design.design):
     def add_pins(self):
         """ Adding pins for port address module"""
         
-        if self.has_rbl():
-            self.add_pin("rbl_bl","INOUT")
-            self.add_pin("rbl_br","INOUT")
+        self.add_pin("rbl_bl","INOUT")
+        self.add_pin("rbl_br","INOUT")
         for bit in range(self.num_cols):
             self.add_pin("{0}_{1}".format(self.bl_names[self.port], bit),"INOUT")
             self.add_pin("{0}_{1}".format(self.br_names[self.port], bit),"INOUT")
@@ -155,29 +154,20 @@ class port_data(design.design):
         
     def add_modules(self):
 
-        if self.port in self.read_ports:
-            # Extra column +1 is for RBL
-            # Precharge will be shifted left if needed
-            self.precharge_array = factory.create(module_type="precharge_array",
-                                                  columns=self.num_cols + 1,
-                                                  bitcell_bl=self.bl_names[self.port],
-                                                  bitcell_br=self.br_names[self.port])
-            self.add_mod(self.precharge_array)
+        # Extra column +1 is for RBL
+        # Precharge will be shifted left if needed
+        self.precharge_array = factory.create(module_type="precharge_array",
+                                              columns=self.num_cols + 1,
+                                              bitcell_bl=self.bl_names[self.port],
+                                              bitcell_br=self.br_names[self.port])
+        self.add_mod(self.precharge_array)
 
+        if self.port in self.read_ports:
             self.sense_amp_array = factory.create(module_type="sense_amp_array",
                                                   word_size=self.word_size, 
                                                   words_per_row=self.words_per_row)
             self.add_mod(self.sense_amp_array)
         else:
-            # Precharge is needed when we have a column mux or for byte writes
-            # to prevent corruption of half-selected cells, so just always add it
-            # This is a little power inefficient for write ports without a column mux,
-            # but it is simpler.
-            self.precharge_array = factory.create(module_type="precharge_array",
-                                                  columns=self.num_cols,
-                                                  bitcell_bl=self.bl_names[self.port],
-                                                  bitcell_br=self.br_names[self.port])
-            self.add_mod(self.precharge_array)
             self.sense_amp_array = None
             
 
@@ -205,7 +195,7 @@ class port_data(design.design):
                                                            write_size=self.write_size)
                 self.add_mod(self.write_mask_and_array)
             else:
-                self.write_mask_and_array_inst = None
+                self.write_mask_and_array = None
 
         else:
             self.write_driver_array = None
@@ -243,14 +233,14 @@ class port_data(design.design):
 
         temp = []
         # Use left BLs for RBL
-        if self.has_rbl() and self.port==0:
+        if self.port==0:
             temp.append("rbl_bl")
             temp.append("rbl_br")
         for bit in range(self.num_cols):
             temp.append(self.bl_names[self.port]+"_{0}".format(bit))
             temp.append(self.br_names[self.port]+"_{0}".format(bit))
         # Use right BLs for RBL
-        if self.has_rbl() and self.port==1:
+        if self.port==1:
             temp.append("rbl_bl")
             temp.append("rbl_br")
         temp.extend(["p_en_bar", "vdd"])
@@ -379,7 +369,7 @@ class port_data(design.design):
         vertical_port_order.append(self.write_mask_and_array_inst)
 
         # Add one column for the the RBL
-        if self.has_rbl() and self.port == 0:
+        if self.port==0:
             x_offset = self.bitcell.width
         else:
             x_offset = 0
@@ -478,7 +468,7 @@ class port_data(design.design):
         
         inst1 = self.column_mux_array_inst
         inst2 = self.precharge_array_inst
-        if self.has_rbl() and self.port==0:
+        if self.port==0:
             self.connect_bitlines(inst1, inst2, self.num_cols, inst2_start_bit=1)
         else:
             self.connect_bitlines(inst1, inst2, self.num_cols)
@@ -499,7 +489,7 @@ class port_data(design.design):
             inst1 = self.precharge_array_inst
             inst1_bl_name = "bl_{}"
             inst1_br_name = "br_{}"
-            if self.has_rbl() and self.port==0:
+            if self.port==0:
                 start_bit=1
             else:
                 start_bit=0
@@ -524,7 +514,7 @@ class port_data(design.design):
             inst1 = self.precharge_array_inst
             inst1_bl_name = "bl_{}"
             inst1_br_name = "br_{}"
-            if self.has_rbl() and self.port==0:
+            if self.port==0:
                 start_bit=1
             else:
                 start_bit=0
@@ -560,11 +550,11 @@ class port_data(design.design):
         """ Add the bitline pins for the given port """
 
         # Connect one bitline to the RBL and offset the indices for the other BLs
-        if self.has_rbl() and self.port==0:
+        if self.port==0:
             self.copy_layout_pin(self.precharge_array_inst, "bl_0", "rbl_bl")
             self.copy_layout_pin(self.precharge_array_inst, "br_0", "rbl_br")
             bit_offset=1
-        elif self.has_rbl() and self.port==1:
+        elif self.port==1:
             self.copy_layout_pin(self.precharge_array_inst, "bl_{}".format(self.num_cols), "rbl_bl")
             self.copy_layout_pin(self.precharge_array_inst, "br_{}".format(self.num_cols), "rbl_br")
             bit_offset=0
@@ -669,5 +659,3 @@ class port_data(design.design):
         if self.precharge_array_inst:
             self.graph_inst_exclude.add(self.precharge_array_inst)
 
-    def has_rbl(self):
-        return self.port in self.read_ports
diff --git a/compiler/modules/replica_bitcell_array.py b/compiler/modules/replica_bitcell_array.py
index 6f4e96c5..b70a7bce 100644
--- a/compiler/modules/replica_bitcell_array.py
+++ b/compiler/modules/replica_bitcell_array.py
@@ -35,7 +35,7 @@ class replica_bitcell_array(design.design):
         self.right_rbl = right_rbl
         self.bitcell_ports = bitcell_ports
 
-        debug.check(left_rbl+right_rbl==len(self.read_ports),"Invalid number of RBLs for port configuration.")
+        debug.check(left_rbl+right_rbl==len(self.all_ports),"Invalid number of RBLs for port configuration.")
         debug.check(left_rbl+right_rbl==len(self.bitcell_ports),"Bitcell ports must match total RBLs.")
         
         # Two dummy rows/cols plus replica for each port
@@ -373,17 +373,7 @@ class replica_bitcell_array(design.design):
     def get_rbl_br_name(self, port):
         """ Return the BR for the given RBL port """
         return self.rbl_br_names[port]
-    
-    def analytical_delay(self, corner, slew, load):
-        """Returns relative delay of the bitline in the bitcell array"""
-        from tech import parameter
-        #The load being driven/drained is mostly the bitline but could include the sense amp or the column mux.
-        #The load from the bitlines is due to the drain capacitances from all the other bitlines and wire parasitics.
-        drain_load = logical_effort.convert_farad_to_relative_c(parameter['bitcell_drain_cap'])
-        wire_unit_load = 0.05 * drain_load #Wires add 5% to this.
-        bitline_load = (drain_load+wire_unit_load)*self.row_size
-        return [self.cell.analytical_delay(corner, slew, load+bitline_load)]
-    
+
     def analytical_power(self, corner, load):
         """Power of Bitcell array and bitline in nW."""
         from tech import drc, parameter
@@ -403,15 +393,6 @@ class replica_bitcell_array(design.design):
                                         cell_power.leakage * self.column_size * self.row_size)
         return total_power
 
-    def gen_wl_wire(self):
-        if OPTS.netlist_only:
-            width = 0
-        else:
-            width = self.width
-        wl_wire = self.generate_rc_net(int(self.column_size), width, drc("minwidth_metal1"))
-        wl_wire.wire_c = 2*spice["min_tx_gate_c"] + wl_wire.wire_c # 2 access tx gate per cell
-        return wl_wire
-
     def gen_bl_wire(self):
         if OPTS.netlist_only:
             height = 0
@@ -422,15 +403,6 @@ class replica_bitcell_array(design.design):
         bl_wire.wire_c =spice["min_tx_drain_c"] + bl_wire.wire_c # 1 access tx d/s per cell
         return bl_wire
 
-    def output_load(self, bl_pos=0):
-        bl_wire = self.gen_bl_wire()
-        return bl_wire.wire_c # sense amp only need to charge small portion of the bl
-                              # set as one segment for now
-
-    def input_load(self):
-        wl_wire = self.gen_wl_wire()
-        return wl_wire.return_input_cap()
-
     def get_wordline_cin(self):
         """Get the relative input capacitance from the wordline connections in all the bitcell"""
         #A single wordline is connected to all the bitcells in a single row meaning the capacitance depends on the # of columns
diff --git a/compiler/modules/sense_amp.py b/compiler/modules/sense_amp.py
index e104356f..e77d577f 100644
--- a/compiler/modules/sense_amp.py
+++ b/compiler/modules/sense_amp.py
@@ -33,15 +33,18 @@ class sense_amp(design.design):
         self.pin_map = sense_amp.pin_map
         self.add_pin_types(self.type_list)
         
-    def input_load(self):
+    def get_cin(self):
+    
+        # FIXME: This input load will be applied to both the s_en timing and bitline timing.
+        
         #Input load for the bitlines which are connected to the source/drain of a TX. Not the selects.
         from tech import spice, parameter
         # Default is 8x. Per Samira and Hodges-Jackson book:
         # "Column-mux transistors driven by the decoder must be sized for optimal speed"
         bitline_pmos_size = 8 #FIXME: This should be set somewhere and referenced. Probably in tech file.
-        return spice["min_tx_drain_c"]*(bitline_pmos_size/parameter["min_tx_size"])#ff   
+        return spice["min_tx_drain_c"]*(bitline_pmos_size)#ff   
         
-    def analytical_delay(self, corner, slew, load):
+    def get_stage_effort(self, load):
         #Delay of the sense amp will depend on the size of the amp and the output load.
         parasitic_delay = 1
         cin = (parameter["sa_inv_pmos_size"] + parameter["sa_inv_nmos_size"])/drc("minwidth_tx")
diff --git a/compiler/modules/sense_amp_array.py b/compiler/modules/sense_amp_array.py
index 40c3adeb..598886c0 100644
--- a/compiler/modules/sense_amp_array.py
+++ b/compiler/modules/sense_amp_array.py
@@ -144,10 +144,7 @@ class sense_amp_array(design.design):
 
     def input_load(self):
         return self.amp.input_load()
-    
-    def analytical_delay(self, corner, slew, load):
-        return [self.amp.analytical_delay(corner, slew=slew, load=load)]
-        
+      
     def get_en_cin(self):
         """Get the relative capacitance of all the sense amp enable connections in the array"""
         sense_amp_en_cin = self.amp.get_en_cin()
diff --git a/compiler/modules/single_level_column_mux_array.py b/compiler/modules/single_level_column_mux_array.py
index e270f2f6..7e3beaad 100644
--- a/compiler/modules/single_level_column_mux_array.py
+++ b/compiler/modules/single_level_column_mux_array.py
@@ -210,15 +210,6 @@ class single_level_column_mux_array(design.design):
                 self.add_via_center(layers=("metal1", "via1", "metal2"),
                                     offset=br_out_offset)
 
-       
-    def analytical_delay(self, corner, slew, load):
-        from tech import parameter
-        """Returns relative delay that the column mux adds"""
-        #Single level column mux will add parasitic loads from other mux pass transistors and the sense amp.
-        drain_load = logical_effort.convert_farad_to_relative_c(parameter['bitcell_drain_cap'])
-        array_load = drain_load*self.words_per_row
-        return [self.mux.analytical_delay(corner, slew, load+array_load)]
-        
     def get_drain_cin(self):
         """Get the relative capacitance of the drain of the NMOS pass TX"""
         from tech import parameter
diff --git a/compiler/modules/tri_gate.py b/compiler/modules/tri_gate.py
index 3907a03b..998562b1 100644
--- a/compiler/modules/tri_gate.py
+++ b/compiler/modules/tri_gate.py
@@ -12,7 +12,7 @@ from tech import GDS,layer
 
 class tri_gate(design.design):
     """
-    This module implements the tri gate cell used in the design for
+    This module implements the tri gate cell used in the design forS
     bit-line isolation. It is a hand-made cell, so the layout and
     netlist should be available in the technology library.  
     """
@@ -36,19 +36,13 @@ class tri_gate(design.design):
         self.pin_map = tri_gate.pin_map
         self.add_pin_types(self.type_list)
 
-    def analytical_delay(self, corner, slew, load=0.0):
-        from tech import spice
-        r = spice["min_tx_r"]
-        c_para = spice["min_tx_drain_c"]
-        return self.cal_delay_with_rc(corner, r = r, c =  c_para+load, slew = slew)
-        
     def analytical_power(self, corner, load):
         """Returns dynamic and leakage power. Results in nW"""
         #Power in this module currently not defined. Returns 0 nW (leakage and dynamic).
         total_power = self.return_power() 
         return total_power
 
-    def input_load(self):
+    def get_cin(self):
         return 9*spice["min_tx_gate_c"]
 
     def build_graph(self, graph, inst_name, port_nets):        
diff --git a/compiler/modules/tri_gate_array.py b/compiler/modules/tri_gate_array.py
index df39ad96..7d1c21d0 100644
--- a/compiler/modules/tri_gate_array.py
+++ b/compiler/modules/tri_gate_array.py
@@ -120,10 +120,4 @@ class tri_gate_array(design.design):
                             layer="metal1",
                             offset=enbar_pin.ll().scale(0, 1),
                             width=width,
-                            height=drc("minwidth_metal1"))
-        
-
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        return self.tri.analytical_delay(corner, slew = slew, load = load)
-        
+                            height=drc("minwidth_metal1"))
\ No newline at end of file
diff --git a/compiler/modules/wordline_driver.py b/compiler/modules/wordline_driver.py
index 39700799..b1292560 100644
--- a/compiler/modules/wordline_driver.py
+++ b/compiler/modules/wordline_driver.py
@@ -210,21 +210,6 @@ class wordline_driver(design.design):
                                                start=wl_offset,
                                                end=wl_offset-vector(self.m1_width,0))
 
-
-    def analytical_delay(self, corner, slew, load=0):
-        # decode -> net
-        decode_t_net = self.nand2.analytical_delay(corner, slew, self.inv.input_load())
-
-        # net -> wl
-        net_t_wl = self.inv.analytical_delay(corner, decode_t_net.slew, load)
-
-        return decode_t_net + net_t_wl
-
-        
-    def input_load(self):
-        """Gets the capacitance of the wordline driver in absolute units (fF)"""
-        return self.nand2.input_load()
-
     def determine_wordline_stage_efforts(self, external_cout, inp_is_rise=True):
         """Follows the clk_buf to a wordline signal adding each stages stage effort to a list"""
         stage_effort_list = []
diff --git a/compiler/pgates/pand2.py b/compiler/pgates/pand2.py
index d299cc56..48f111c7 100644
--- a/compiler/pgates/pand2.py
+++ b/compiler/pgates/pand2.py
@@ -110,14 +110,6 @@ class pand2(pgate.pgate):
                                             width=pin.width(),
                                             height=pin.height())
         
-        
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        """ Calculate the analytical delay of DFF-> INV -> INV """
-        nand_delay = self.nand.analytical_delay(corner, slew=slew, load=self.inv.input_load()) 
-        inv_delay = self.inv.analytical_delay(corner, slew=nand_delay.slew, load=load)
-        return nand_delay + inv_delay
-    
     def get_stage_efforts(self, external_cout, inp_is_rise=False):
         """Get the stage efforts of the A or B -> Z path"""
         stage_effort_list = []
diff --git a/compiler/pgates/pand3.py b/compiler/pgates/pand3.py
new file mode 100644
index 00000000..b221ad84
--- /dev/null
+++ b/compiler/pgates/pand3.py
@@ -0,0 +1,138 @@
+# 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 debug
+from tech import drc
+from math import log
+from vector import vector
+from globals import OPTS
+import pgate
+from sram_factory import factory
+
+class pand3(pgate.pgate):
+    """
+    This is a simple buffer used for driving loads. 
+    """
+    def __init__(self, name, size=1, height=None):
+        debug.info(1, "Creating pand3 {}".format(name))
+        self.add_comment("size: {}".format(size))
+        
+        self.size = size
+        
+        # Creates the netlist and layout        
+        pgate.pgate.__init__(self, name, height)
+
+    def create_netlist(self):
+        self.add_pins()
+        self.create_modules()
+        self.create_insts()
+
+    def create_modules(self):
+        # Shield the cap, but have at least a stage effort of 4
+        self.nand = factory.create(module_type="pnand3",height=self.height) 
+        self.add_mod(self.nand)
+        
+        self.inv = factory.create(module_type="pinv", size=self.size, height=self.height)
+        self.add_mod(self.inv)
+
+    def create_layout(self):
+        self.width = self.nand.width + self.inv.width
+        self.place_insts()
+        self.add_wires()
+        self.add_layout_pins()
+        self.DRC_LVS()
+        
+    def add_pins(self):
+        self.add_pin("A", "INPUT")
+        self.add_pin("B", "INPUT")
+        self.add_pin("C", "INPUT")
+        self.add_pin("Z", "OUTPUT")
+        self.add_pin("vdd", "POWER")
+        self.add_pin("gnd", "GROUND")
+
+    def create_insts(self):
+        self.nand_inst=self.add_inst(name="pand3_nand",
+                                     mod=self.nand)
+        self.connect_inst(["A", "B", "C", "zb_int",  "vdd", "gnd"])
+        
+        self.inv_inst=self.add_inst(name="pand3_inv",
+                                    mod=self.inv)
+        self.connect_inst(["zb_int", "Z",  "vdd", "gnd"])
+
+    def place_insts(self):
+        # Add NAND to the right 
+        self.nand_inst.place(offset=vector(0,0))
+
+        # Add INV to the right
+        self.inv_inst.place(offset=vector(self.nand_inst.rx(),0))
+        
+    def add_wires(self):
+        # nand Z to inv A
+        z1_pin = self.nand_inst.get_pin("Z")
+        a2_pin = self.inv_inst.get_pin("A")
+        mid1_point = vector(0.5*(z1_pin.cx()+a2_pin.cx()), z1_pin.cy())
+        mid2_point = vector(mid1_point, a2_pin.cy())
+        self.add_path("metal1", [z1_pin.center(), mid1_point, mid2_point, a2_pin.center()])
+        
+        
+    def add_layout_pins(self):
+        # Continous vdd rail along with label.
+        vdd_pin=self.inv_inst.get_pin("vdd")
+        self.add_layout_pin(text="vdd",
+                            layer="metal1",
+                            offset=vdd_pin.ll().scale(0,1),
+                            width=self.width,
+                            height=vdd_pin.height())
+        
+        # Continous gnd rail along with label.
+        gnd_pin=self.inv_inst.get_pin("gnd")
+        self.add_layout_pin(text="gnd",
+                            layer="metal1",
+                            offset=gnd_pin.ll().scale(0,1),
+                            width=self.width,
+                            height=vdd_pin.height())
+            
+        pin = self.inv_inst.get_pin("Z")
+        self.add_layout_pin_rect_center(text="Z",
+                                        layer=pin.layer,
+                                        offset=pin.center(),
+                                        width=pin.width(),
+                                        height=pin.height())
+
+        for pin_name in ["A","B", "C"]:
+            pin = self.nand_inst.get_pin(pin_name)
+            self.add_layout_pin_rect_center(text=pin_name,
+                                            layer=pin.layer,
+                                            offset=pin.center(),
+                                            width=pin.width(),
+                                            height=pin.height())
+        
+        
+
+    def analytical_delay(self, corner, slew, load=0.0):
+        """ Calculate the analytical delay of DFF-> INV -> INV """
+        nand_delay = self.nand.analytical_delay(corner, slew=slew, load=self.inv.input_load()) 
+        inv_delay = self.inv.analytical_delay(corner, slew=nand_delay.slew, load=load)
+        return nand_delay + inv_delay
+    
+    def get_stage_efforts(self, external_cout, inp_is_rise=False):
+        """Get the stage efforts of the A or B -> Z path"""
+        stage_effort_list = []
+        stage1_cout = self.inv.get_cin()
+        stage1 = self.nand.get_stage_effort(stage1_cout, inp_is_rise)
+        stage_effort_list.append(stage1)
+        last_stage_is_rise = stage1.is_rise
+        
+        stage2 = self.inv.get_stage_effort(external_cout, last_stage_is_rise)
+        stage_effort_list.append(stage2)
+        
+        return stage_effort_list
+
+    def get_cin(self):
+        """Return the relative input capacitance of a single input"""
+        return self.nand.get_cin()
+        
diff --git a/compiler/pgates/pbuf.py b/compiler/pgates/pbuf.py
index 125a1190..fcfc6586 100644
--- a/compiler/pgates/pbuf.py
+++ b/compiler/pgates/pbuf.py
@@ -113,14 +113,6 @@ class pbuf(pgate.pgate):
                                         width=a_pin.width(),
                                         height=a_pin.height())
         
-        
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        """ Calculate the analytical delay of DFF-> INV -> INV """
-        inv1_delay = self.inv1.analytical_delay(corner, slew=slew, load=self.inv2.input_load()) 
-        inv2_delay = self.inv2.analytical_delay(corner, slew=inv1_delay.slew, load=load)
-        return inv1_delay + inv2_delay
-    
     def get_stage_efforts(self, external_cout, inp_is_rise=False):
         """Get the stage efforts of the A -> Z path"""
         stage_effort_list = []
diff --git a/compiler/pgates/pdriver.py b/compiler/pgates/pdriver.py
index a3180006..36811a5e 100644
--- a/compiler/pgates/pdriver.py
+++ b/compiler/pgates/pdriver.py
@@ -173,30 +173,6 @@ class pdriver(pgate.pgate):
                                         offset=a_pin.center(),
                                         width = a_pin.width(),
                                         height = a_pin.height())
-        
-    def input_load(self):
-        return self.inv_list[0].input_load()
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        """ Calculate the analytical delay of INV1 -> ... -> INVn """
-
-        cout_list = []
-        for prev_inv,inv in zip(self.inv_list, self.inv_list[1:]):
-            cout_list.append(inv.input_load())
-        cout_list.append(load)
-        
-        input_slew = slew
-        
-        delays = []
-        for inv,cout in zip(self.inv_list,cout_list):
-            delays.append(inv.analytical_delay(corner, slew=input_slew, load=cout))
-            input_slew = delays[-1].slew
-
-        delay = delays[0]
-        for i in range(len(delays)-1):
-            delay += delays[i]
-            
-        return delay
 
     def get_sizes(self):
         """ Return the relative sizes of the buffers """
diff --git a/compiler/pgates/pinv.py b/compiler/pgates/pinv.py
index 773a0452..0844c762 100644
--- a/compiler/pgates/pinv.py
+++ b/compiler/pgates/pinv.py
@@ -255,15 +255,6 @@ class pinv(pgate.pgate):
 
         self.connect_pin_to_rail(self.pmos_inst,"S","vdd")
         
-
-    def input_load(self):
-        return ((self.nmos_size+self.pmos_size)/parameter["min_tx_size"])*spice["min_tx_gate_c"]
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        r = spice["min_tx_r"]/(self.nmos_size/parameter["min_tx_size"])
-        c_para = spice["min_tx_drain_c"]*(self.nmos_size/parameter["min_tx_size"])#ff
-        return self.cal_delay_with_rc(corner, r = r, c =  c_para+load, slew = slew)
-        
     def analytical_power(self, corner, load):
         """Returns dynamic and leakage power. Results in nW"""
         c_eff = self.calculate_effective_capacitance(load)
@@ -281,11 +272,11 @@ class pinv(pgate.pgate):
         transition_prob = spice["inv_transition_prob"]
         return transition_prob*(c_load + c_para) 
 
-    def get_cin(self):
+    def input_load(self):
         """Return the capacitance of the gate connection in generic capacitive
            units relative to the minimum width of a transistor"""
         return self.nmos_size + self.pmos_size
-        
+      
     def get_stage_effort(self, cout, inp_is_rise=True):
         """Returns an object representing the parameters for delay in tau units.
            Optional is_rise refers to the input direction rise/fall. Input inverted by this stage.
@@ -293,7 +284,7 @@ class pinv(pgate.pgate):
         parasitic_delay = 1 
         return logical_effort.logical_effort(self.name, 
                                              self.size, 
-                                             self.get_cin(), 
+                                             self.input_load(), 
                                              cout, 
                                              parasitic_delay, 
                                              not inp_is_rise)
diff --git a/compiler/pgates/pinvbuf.py b/compiler/pgates/pinvbuf.py
index 72ab6a8a..a8d2c8b8 100644
--- a/compiler/pgates/pinvbuf.py
+++ b/compiler/pgates/pinvbuf.py
@@ -178,15 +178,7 @@ class pinvbuf(pgate.pgate):
                                         offset=a_pin.center())
         self.add_via_center(layers=("metal1","via1","metal2"),
                             offset=a_pin.center())
-        
-        
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        """ Calculate the analytical delay of DFF-> INV -> INV """
-        inv1_delay = self.inv1.analytical_delay(corner, slew=slew, load=self.inv2.input_load()) 
-        inv2_delay = self.inv2.analytical_delay(corner, slew=inv1_delay.slew, load=load)
-        return inv1_delay + inv2_delay
-            
+             
     def determine_clk_buf_stage_efforts(self, external_cout, inp_is_rise=False):
         """Get the stage efforts of the clk -> clk_buf path"""
         stage_effort_list = []
diff --git a/compiler/pgates/pnand2.py b/compiler/pgates/pnand2.py
index 956f0a30..21ca4184 100644
--- a/compiler/pgates/pnand2.py
+++ b/compiler/pgates/pnand2.py
@@ -37,10 +37,7 @@ class pnand2(pgate.pgate):
 
         # Creates the netlist and layout
         pgate.pgate.__init__(self, name, height)
-
-        #For characterization purposes only
-        #self.exclude_nmos_from_graph()
-        
+   
     def create_netlist(self):
         self.add_pins()
         self.add_ptx()
@@ -233,17 +230,6 @@ class pnand2(pgate.pgate):
                                         width=contact.m1m2.first_layer_height,
                                         height=contact.m1m2.first_layer_width)
 
-
-
-
-    def input_load(self):
-        return ((self.nmos_size+self.pmos_size)/parameter["min_tx_size"])*spice["min_tx_gate_c"]
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        r = spice["min_tx_r"]/(self.nmos_size/parameter["min_tx_size"])
-        c_para = spice["min_tx_drain_c"]*(self.nmos_size/parameter["min_tx_size"])#ff
-        return self.cal_delay_with_rc(corner, r = r, c =  c_para+load, slew = slew)
-        
     def analytical_power(self, corner, load):
         """Returns dynamic and leakage power. Results in nW"""
         c_eff = self.calculate_effective_capacitance(load)
@@ -261,23 +247,16 @@ class pnand2(pgate.pgate):
         transition_prob = spice["nand2_transition_prob"]
         return transition_prob*(c_load + c_para) 
 
-    def get_cin(self):
+    def input_load(self):
         """Return the relative input capacitance of a single input"""
         return self.nmos_size+self.pmos_size
-        
+    
     def get_stage_effort(self, cout, inp_is_rise=True):
         """Returns an object representing the parameters for delay in tau units.
            Optional is_rise refers to the input direction rise/fall. Input inverted by this stage.
         """
         parasitic_delay = 2 
-        return logical_effort.logical_effort(self.name, self.size, self.get_cin(), cout, parasitic_delay, not inp_is_rise)
-        
-    def exclude_nmos_from_graph(self):
-        """Exclude the nmos TXs from the graph to reduce complexity"""
-        #The pull-down network has an internal net which causes 2 different in->out paths
-        #Removing them simplifies generic path searching.
-        self.graph_inst_exclude.add(self.nmos1_inst)
-        self.graph_inst_exclude.add(self.nmos2_inst)
+        return logical_effort.logical_effort(self.name, self.size, self.input_load(), cout, parasitic_delay, not inp_is_rise)
 
     def build_graph(self, graph, inst_name, port_nets):        
         """Adds edges based on inputs/outputs. Overrides base class function."""
diff --git a/compiler/pgates/pnand3.py b/compiler/pgates/pnand3.py
index 30da9165..89628267 100644
--- a/compiler/pgates/pnand3.py
+++ b/compiler/pgates/pnand3.py
@@ -243,16 +243,6 @@ class pnand3(pgate.pgate):
                                         width=contact.m1m2.first_layer_width,
                                         height=contact.m1m2.first_layer_height)
 
-
-
-    def input_load(self):
-        return ((self.nmos_size+self.pmos_size)/parameter["min_tx_size"])*spice["min_tx_gate_c"]
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        r = spice["min_tx_r"]/(self.nmos_size/parameter["min_tx_size"])
-        c_para = spice["min_tx_drain_c"]*(self.nmos_size/parameter["min_tx_size"])#ff
-        return self.cal_delay_with_rc(corner, r = r, c =  c_para+load, slew = slew)
-        
     def analytical_power(self, corner, load):
         """Returns dynamic and leakage power. Results in nW"""
         c_eff = self.calculate_effective_capacitance(load)
@@ -270,7 +260,7 @@ class pnand3(pgate.pgate):
         transition_prob = spice["nand3_transition_prob"]
         return transition_prob*(c_load + c_para) 
 
-    def get_cin(self):
+    def input_load(self):
         """Return the relative input capacitance of a single input"""
         return self.nmos_size+self.pmos_size
         
@@ -279,7 +269,7 @@ class pnand3(pgate.pgate):
            Optional is_rise refers to the input direction rise/fall. Input inverted by this stage.
         """
         parasitic_delay = 3 
-        return logical_effort.logical_effort(self.name, self.size, self.get_cin(), cout, parasitic_delay, not inp_is_rise)
+        return logical_effort.logical_effort(self.name, self.size, self.input_load(), cout, parasitic_delay, not inp_is_rise)
 
     def build_graph(self, graph, inst_name, port_nets):        
         """Adds edges based on inputs/outputs. Overrides base class function."""
diff --git a/compiler/pgates/pnor2.py b/compiler/pgates/pnor2.py
index fa52c528..472c75a8 100644
--- a/compiler/pgates/pnor2.py
+++ b/compiler/pgates/pnor2.py
@@ -5,12 +5,13 @@
 # (acting for and on behalf of Oklahoma State University)
 # All rights reserved.
 #
+import contact
 import pgate
 import debug
 from tech import drc, parameter, spice
 from vector import vector
 from globals import OPTS
-import contact
+import logical_effort
 from sram_factory import factory
 
 class pnor2(pgate.pgate):
@@ -38,28 +39,30 @@ class pnor2(pgate.pgate):
         pgate.pgate.__init__(self, name, height)
 
         
-    def add_pins(self):
-        """ Adds pins for spice netlist """
-        pin_list = ["A", "B", "Z", "vdd", "gnd"]
-        dir_list = ["INPUT", "INPUT", "OUTPUT", "INOUT", "INOUT"]
-        self.add_pin_list(pin_list, dir_list)
-
     def create_netlist(self):
         self.add_pins()
+        self.add_ptx()
         self.create_ptx()
-        self.setup_layout_constants()
         
     def create_layout(self):
         """ Calls all functions related to the generation of the layout """
-        self.add_supply_rails()
-        self.add_ptx()
+        
+        self.setup_layout_constants()
+        self.route_supply_rails()
+        self.place_ptx()
         self.connect_rails()
         self.add_well_contacts()
         self.extend_wells(self.well_pos)
         self.route_inputs()
         self.route_output()
 
-    def create_ptx(self):
+    def add_pins(self):
+        """ Adds pins for spice netlist """
+        pin_list = ["A", "B", "Z", "vdd", "gnd"]
+        dir_list = ["INPUT", "INPUT", "OUTPUT", "INOUT", "INOUT"]
+        self.add_pin_list(pin_list, dir_list)
+
+    def add_ptx(self):
         """ Create the PMOS and NMOS transistors. """
         self.nmos = factory.create(module_type="ptx",
                                    width=self.nmos_width,
@@ -104,7 +107,7 @@ class pnor2(pgate.pgate):
         self.top_bottom_space = max(0.5*self.m1_width + self.m1_space + extra_contact_space, 
                                     drc("poly_extend_active"), self.poly_space)
         
-    def add_supply_rails(self):
+    def route_supply_rails(self):
         """ Add vdd/gnd rails to the top and bottom. """
         self.add_layout_pin_rect_center(text="gnd",
                                         layer="metal1",
@@ -116,7 +119,31 @@ class pnor2(pgate.pgate):
                                         offset=vector(0.5*self.width,self.height),
                                         width=self.width)
 
-    def add_ptx(self):
+    def create_ptx(self):
+        """ 
+        Add PMOS and NMOS to the layout at the upper-most and lowest position
+        to provide maximum routing in channel
+        """
+
+        self.pmos1_inst=self.add_inst(name="pnor2_pmos1",
+                                      mod=self.pmos)
+        self.connect_inst(["vdd", "A", "net1", "vdd"])
+
+        self.pmos2_inst = self.add_inst(name="pnor2_pmos2",
+                                        mod=self.pmos)
+        self.connect_inst(["net1", "B", "Z", "vdd"])
+
+        
+        self.nmos1_inst=self.add_inst(name="pnor2_nmos1",
+                                      mod=self.nmos)
+        self.connect_inst(["Z", "A", "gnd", "gnd"])
+
+        self.nmos2_inst=self.add_inst(name="pnor2_nmos2",
+                                      mod=self.nmos)
+        self.connect_inst(["Z", "B", "gnd", "gnd"])
+        
+
+    def place_ptx(self):
         """ 
         Add PMOS and NMOS to the layout at the upper-most and lowest position
         to provide maximum routing in channel
@@ -124,29 +151,16 @@ class pnor2(pgate.pgate):
 
         pmos1_pos = vector(self.pmos.active_offset.x,
                            self.height - self.pmos.active_height - self.top_bottom_space)
-        self.pmos1_inst=self.add_inst(name="pnor2_pmos1",
-                                      mod=self.pmos,
-                                      offset=pmos1_pos)
-        self.connect_inst(["vdd", "A", "net1", "vdd"])
+        self.pmos1_inst.place(pmos1_pos)
 
         self.pmos2_pos = pmos1_pos + self.overlap_offset
-        self.pmos2_inst = self.add_inst(name="pnor2_pmos2",
-                                        mod=self.pmos,
-                                        offset=self.pmos2_pos)
-        self.connect_inst(["net1", "B", "Z", "vdd"])
-
+        self.pmos2_inst.place(self.pmos2_pos)
         
         nmos1_pos = vector(self.pmos.active_offset.x, self.top_bottom_space)
-        self.nmos1_inst=self.add_inst(name="pnor2_nmos1",
-                                      mod=self.nmos,
-                                      offset=nmos1_pos)
-        self.connect_inst(["Z", "A", "gnd", "gnd"])
-
+        self.nmos1_inst.place(nmos1_pos)
+        
         self.nmos2_pos = nmos1_pos + self.overlap_offset
-        self.nmos2_inst=self.add_inst(name="pnor2_nmos2",
-                                      mod=self.nmos,
-                                      offset=self.nmos2_pos)
-        self.connect_inst(["Z", "B", "gnd", "gnd"])
+        self.nmos2_inst.place(self.nmos2_pos)
         
         # Output position will be in between the PMOS and NMOS        
         self.output_pos = vector(0,0.5*(pmos1_pos.y+nmos1_pos.y+self.nmos.active_height))
@@ -213,16 +227,6 @@ class pnor2(pgate.pgate):
                                         width=contact.m1m2.first_layer_height,
                                         height=contact.m1m2.first_layer_width)
 
-
-
-    def input_load(self):
-        return ((self.nmos_size+self.pmos_size)/parameter["min_tx_size"])*spice["min_tx_gate_c"]
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        r = spice["min_tx_r"]/(self.nmos_size/parameter["min_tx_size"])
-        c_para = spice["min_tx_drain_c"]*(self.nmos_size/parameter["min_tx_size"])#ff
-        return self.cal_delay_with_rc(corner, r = r, c =  c_para+load, slew = slew)
-        
     def analytical_power(self, corner, load):
         """Returns dynamic and leakage power. Results in nW"""
         c_eff = self.calculate_effective_capacitance(load)
diff --git a/compiler/pgates/ptristate_inv.py b/compiler/pgates/ptristate_inv.py
index 55a248a7..0564bf86 100644
--- a/compiler/pgates/ptristate_inv.py
+++ b/compiler/pgates/ptristate_inv.py
@@ -210,18 +210,11 @@ class ptristate_inv(pgate.pgate):
         self.connect_pin_to_rail(self.nmos1_inst,"S","gnd")
         self.connect_pin_to_rail(self.pmos1_inst,"S","vdd")
         
-
-    def analytical_delay(self, corner, slew, load=0.0):
-        from tech import spice
-        r = spice["min_tx_r"]
-        c_para = spice["min_tx_drain_c"]
-        return self.cal_delay_with_rc(corner, r = r, c =  c_para+load, slew = slew)
-        
     def analytical_power(self, corner, load):
         """Returns dynamic and leakage power. Results in nW"""
         #Power in this module currently not defined. Returns 0 nW (leakage and dynamic).
         total_power = self.return_power() 
         return total_power
 
-    def input_load(self):
+    def get_cin(self):
         return 9*spice["min_tx_gate_c"]
diff --git a/compiler/pgates/single_level_column_mux.py b/compiler/pgates/single_level_column_mux.py
index 7c21c0a8..999d3ccd 100644
--- a/compiler/pgates/single_level_column_mux.py
+++ b/compiler/pgates/single_level_column_mux.py
@@ -189,7 +189,7 @@ class single_level_column_mux(pgate.pgate):
                       width=self.bitcell.width,
                       height=self.height)
         
-    def analytical_delay(self, corner, slew, load):
+    def get_stage_effort(self, corner, slew, load):
         """Returns relative delay that the column mux. Difficult to convert to LE model."""
         parasitic_delay = 1
         cin = 2*self.tx_size #This is not CMOS, so using this may be incorrect.
diff --git a/compiler/sram/sram_1bank.py b/compiler/sram/sram_1bank.py
index 355ca769..51cfeada 100644
--- a/compiler/sram/sram_1bank.py
+++ b/compiler/sram/sram_1bank.py
@@ -301,7 +301,7 @@ class sram_1bank(sram_base):
                 dest_pin = self.bank_inst.get_pin(signal+"{}".format(port))                
                 self.connect_vbus_m2m3(src_pin, dest_pin)
 
-        for port in self.read_ports:
+        for port in self.all_ports:
             # Only input (besides pins) is the replica bitline
             src_pin = self.control_logic_insts[port].get_pin("rbl_bl")
             dest_pin = self.bank_inst.get_pin("rbl_bl{}".format(port))                
diff --git a/compiler/sram/sram_base.py b/compiler/sram/sram_base.py
index f34d058e..f06a56f0 100644
--- a/compiler/sram/sram_base.py
+++ b/compiler/sram/sram_base.py
@@ -339,7 +339,7 @@ class sram_base(design, verilog, lef):
         for port in self.read_ports:
             for bit in range(self.word_size):
                 temp.append("DOUT{0}[{1}]".format(port,bit))
-        for port in self.read_ports:
+        for port in self.all_ports:
             temp.append("rbl_bl{0}".format(port))
         for port in self.write_ports:
             for bit in range(self.word_size):
@@ -499,9 +499,7 @@ class sram_base(design, verilog, lef):
             if port in self.readwrite_ports:
                 temp.append("web{}".format(port))
             temp.append("clk{}".format(port))
-
-            if port in self.read_ports:
-                temp.append("rbl_bl{}".format(port))
+            temp.append("rbl_bl{}".format(port))
 
             # Ouputs
             if port in self.read_ports:
@@ -564,31 +562,7 @@ class sram_base(design, verilog, lef):
         self.sp_write_file(sp, usedMODS)
         del usedMODS
         sp.close()
-
-        
-    def analytical_delay(self, corner, slew,load):
-        """ Estimates the delay from clk -> DOUT 
-            LH and HL are the same in analytical model. """
-        delays = {}
-        for port in self.all_ports:
-            if port in self.readonly_ports:
-                control_logic = self.control_logic_r
-            elif port in self.readwrite_ports:
-                control_logic = self.control_logic_rw
-            else:
-                delays[port] = self.return_delay(0,0) #Write ports do not have a lib defined delay, marked as 0 
-                continue
-            clk_to_wlen_delays = control_logic.analytical_delay(corner, slew, load)
-            wlen_to_dout_delays = self.bank.analytical_delay(corner,slew,load,port) #port should probably be specified...
-            all_delays = clk_to_wlen_delays+wlen_to_dout_delays
-            total_delay = logical_effort.calculate_absolute_delay(all_delays)
-            total_delay = self.apply_corners_analytically(total_delay, corner)
-            last_slew = .1*all_delays[-1].get_absolute_delay() #slew approximated as 10% of delay
-            last_slew = self.apply_corners_analytically(last_slew, corner)
-            delays[port] = self.return_delay(delay=total_delay, slew=last_slew)
-
-        return delays
-        
+ 
     def get_wordline_stage_efforts(self, inp_is_rise=True):
         """Get the all the stage efforts for each stage in the path from clk_buf to a wordline"""
         stage_effort_list = []
diff --git a/compiler/tests/04_pand3_test.py b/compiler/tests/04_pand3_test.py
new file mode 100755
index 00000000..4408f6e8
--- /dev/null
+++ b/compiler/tests/04_pand3_test.py
@@ -0,0 +1,38 @@
+#!/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 pand3_test(openram_test):
+
+    def runTest(self):
+        globals.init_openram("config_{0}".format(OPTS.tech_name))
+        global verify
+        import verify
+
+        import pand3
+
+        debug.info(2, "Testing pand3 gate 4x")
+        a = pand3.pand3(name="pand3x4", size=4)
+        self.local_check(a)
+
+        globals.end_openram()
+
+# instantiate a copdsay of the class to actually run the test
+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/21_model_delay_test.py b/compiler/tests/21_model_delay_test.py
index 3b43e789..114f0c6e 100755
--- a/compiler/tests/21_model_delay_test.py
+++ b/compiler/tests/21_model_delay_test.py
@@ -15,7 +15,7 @@ from globals import OPTS
 from sram_factory import factory
 import debug
 
-@unittest.skip("SKIPPING 21_model_delay_test")
+# @unittest.skip("SKIPPING 21_model_delay_test")
 class model_delay_test(openram_test):
     """ Compare the accuracy of the analytical model with a spice simulation. """
     
@@ -51,23 +51,28 @@ class model_delay_test(openram_test):
         import tech
         loads = [tech.spice["msflop_in_cap"]*4]
         slews = [tech.spice["rise_time"]*2]
+     
+        # Run a spice characterization
         spice_data, port_data = d.analyze(probe_address, probe_data, slews, loads)
         spice_data.update(port_data[0])
      
+        # Run analytical characterization
         model_data, port_data = d.analytical_delay(slews, loads)
         model_data.update(port_data[0])
         
-        #Only compare the delays
+        # Only compare the delays
         spice_delays = {key:value for key, value in spice_data.items() if 'delay' in key}
         model_delays = {key:value for key, value in model_data.items() if 'delay' in key}
         debug.info(1,"Spice Delays={}".format(spice_delays))
         debug.info(1,"Model Delays={}".format(model_delays))
+        
         if OPTS.tech_name == "freepdk45":
             error_tolerance = 0.25
         elif OPTS.tech_name == "scn4m_subm":
             error_tolerance = 0.25
         else:
             self.assertTrue(False) # other techs fail
+            
         # Check if no too many or too few results
         self.assertTrue(len(spice_delays.keys())==len(model_delays.keys()))
 
diff --git a/compiler/tests/21_ngspice_delay_test.py b/compiler/tests/21_ngspice_delay_test.py
index 06a61d55..39f0d5e0 100755
--- a/compiler/tests/21_ngspice_delay_test.py
+++ b/compiler/tests/21_ngspice_delay_test.py
@@ -54,27 +54,27 @@ class timing_sram_test(openram_test):
         data.update(port_data[0])
 
         if OPTS.tech_name == "freepdk45":
-            golden_data = {'delay_hl': [0.22609590000000002],
-                           'delay_lh': [0.22609590000000002],
-                           'leakage_power': 0.003317743,
+            golden_data = {'delay_hl': [0.2264205],
+                           'delay_lh': [0.2264205],
+                           'leakage_power': 0.0021017429999999997,
                            'min_period': 0.859,
-                           'read0_power': [0.3271056],
-                           'read1_power': [0.3064244],
-                           'slew_hl': [0.2153979],
-                           'slew_lh': [0.2153979],
-                           'write0_power': [0.3532067],
-                           'write1_power': [0.3381259]}
+                           'read0_power': [0.3339161],
+                           'read1_power': [0.31329440000000003],
+                           'slew_hl': [0.2590786],
+                           'slew_lh': [0.2590786],
+                           'write0_power': [0.36360849999999995],
+                           'write1_power': [0.3486931]}
         elif OPTS.tech_name == "scn4m_subm":
-            golden_data = {'delay_hl': [1.709791],
-                           'delay_lh': [1.709791],
-                           'leakage_power': 0.06803324999999999,
+            golden_data = {'delay_hl': [1.7083549999999998],
+                           'delay_lh': [1.7083549999999998],
+                           'leakage_power': 0.001119657,
                            'min_period': 7.812,
-                           'read0_power': [7.9499070000000005],
-                           'read1_power': [7.619662999999999],
-                           'slew_hl': [1.390261],
-                           'slew_lh': [1.390261],
-                           'write0_power': [8.913003],
-                           'write1_power': [8.166687000000001]}
+                           'read0_power': [8.013845],
+                           'read1_power': [7.6889389999999995],
+                           'slew_hl': [1.31918],
+                           'slew_lh': [1.31918],
+                           'write0_power': [8.791557000000001],
+                           'write1_power': [8.70443]}
         else:
             self.assertTrue(False) # other techs fail
 
diff --git a/compiler/tests/23_lib_sram_model_corners_test.py b/compiler/tests/23_lib_sram_model_corners_test.py
index 51fcbca6..84b54e3d 100755
--- a/compiler/tests/23_lib_sram_model_corners_test.py
+++ b/compiler/tests/23_lib_sram_model_corners_test.py
@@ -14,7 +14,7 @@ import globals
 from globals import OPTS
 import debug
 
-@unittest.skip("SKIPPING 23_lib_sram_model_corners_test")
+#@unittest.skip("SKIPPING 23_lib_sram_model_corners_test")
 class lib_model_corners_lib_test(openram_test):
 
     def runTest(self):
diff --git a/compiler/tests/23_lib_sram_model_test.py b/compiler/tests/23_lib_sram_model_test.py
index fd75dc25..9fdda7e7 100755
--- a/compiler/tests/23_lib_sram_model_test.py
+++ b/compiler/tests/23_lib_sram_model_test.py
@@ -14,7 +14,7 @@ import globals
 from globals import OPTS
 import debug
 
-@unittest.skip("SKIPPING 23_lib_sram_model_test")
+#@unittest.skip("SKIPPING 23_lib_sram_model_test")
 class lib_sram_model_test(openram_test):
 
     def runTest(self):
diff --git a/compiler/tests/30_openram_front_end_test.py b/compiler/tests/30_openram_front_end_test.py
index 2be14001..db75fe0a 100755
--- a/compiler/tests/30_openram_front_end_test.py
+++ b/compiler/tests/30_openram_front_end_test.py
@@ -16,7 +16,7 @@ from sram_factory import factory
 import debug
 import getpass
 
-@unittest.skip("SKIPPING 30_openram_front_end_test")
+#@unittest.skip("SKIPPING 30_openram_front_end_test")
 class openram_front_end_test(openram_test):
 
     def runTest(self):