arch: Added an architecture file for the 4-bit Adder Double Chain
architecture
diff --git a/vtr_flow/arch/timing/4bit_adder_double_chain_arch.xml b/vtr_flow/arch/timing/4bit_adder_double_chain_arch.xml
new file mode 100644
index 0000000..896aa63
--- /dev/null
+++ b/vtr_flow/arch/timing/4bit_adder_double_chain_arch.xml
@@ -0,0 +1,968 @@
+<!--
+ This is the architecture file for the 4-bit Adder Double Chain Architecture proposed in [1].
+ Delays for routing and logic blocks come from COFFE runs for a 20 nm technology node.
+ Delays for DSP blocks and BRAMs come from Arria 10 (22 nm) delays.
+
+ This architecture is Stratix-10 like architecture with a modified arithmetic mode
+
+ This architecture has 10 ALMs per cluster, where each ALM is a 6-LUT fracturable into
+ two 5-LUTs. The ALM has 8 inputs and 4 optionally registered outputs.The two 5-LUTs should
+ share at least two inputs. Each two ALM outputs are logically equivalent, which means any
+ output signal that can reach ALM.out[0] can reach ALM.out[1] and the same thing for
+ ALM.out[2] and ALM.out[3]. This architecture has an arithmetic mode where each 5-LUT
+ is fractured into two 4-LUTs and each 4-LUT into two 3-LUTs. This results in a total of
+ eight 3-LUTs per ALM and four bits of addition. This architecture has two separate carry chains
+ with separate start (Cin[0] and Cin[1]) and end (Cout[0] and Cout[1]) points. The adders in each
+ 5 ALMs are connected together with a carry chain; chain 1 is connecting ALM[0] : ALM[4] while
+ chain 2 is connecting ALM[5] : ALM[9].
+
+ The LAB has 60 inputs and 40 outputs. Two outputs of each ALM are fed to the right and
+ left LAB using direct links and are also fed back to the LAB as feedback connections sharing
+ the 60 input ports with the signals coming from the routing channels.
+
+ The architecture also has a 20Kb memory that has true and simple dual port modes. In simple
+ dual port mode the memory can be configured in the following modes: 512x40, 1024x20 and 2048x10,
+ while in true dual port mode it can be configured as: 1024x20 and 2028x10.
+
+ In addition, the architecture has a 27x27 DSP block that can be fractured into two 18x19 DSPs.
+
+
+ [1] M. Eldafrawy, A. Boutros, S. Yazdanshenas, and V. Betz, "FPGA Logic Block Architectures for efficient
+ multiplication and addition to enhance machine learning performance," in Transactions on Reconfigurable
+ Technology and Systems (TRETS), 2019
+
+-->
+<architecture>
+ <!--
+ ODIN II specific config begins
+ Describes the types of user-specified netlist blocks (in blif, this corresponds to
+ ".model [type_of_block]") that this architecture supports.
+
+ Note: Basic LUTs, I/Os, and flip-flops are not included here as there are
+ already special structures in blif (.names, .input, .output, and .latch)
+ that describe them.
+ -->
+ <models>
+ <model name="multiply">
+ <input_ports>
+ <port name="a" combinational_sink_ports="out"/>
+ <port name="b" combinational_sink_ports="out"/>
+ </input_ports>
+ <output_ports>
+ <port name="out"/>
+ </output_ports>
+ </model>
+
+ <model name="single_port_ram">
+ <input_ports>
+ <port name="we" clock="clk"/> <!-- control -->
+ <port name="addr" clock="clk"/> <!-- address lines -->
+ <port name="data" clock="clk"/> <!-- data lines can be broken down into smaller bit widths minimum size 1 -->
+ <port name="clk" is_clock="1"/> <!-- memories are often clocked -->
+ </input_ports>
+ <output_ports>
+ <port name="out" clock="clk"/> <!-- output can be broken down into smaller bit widths minimum size 1 -->
+ </output_ports>
+ </model>
+
+ <model name="dual_port_ram">
+ <input_ports>
+ <port name="we1" clock="clk"/> <!-- write enable -->
+ <port name="we2" clock="clk"/> <!-- write enable -->
+ <port name="addr1" clock="clk"/> <!-- address lines -->
+ <port name="addr2" clock="clk"/> <!-- address lines -->
+ <port name="data1" clock="clk"/> <!-- data lines can be broken down into smaller bit widths minimum size 10 -->
+ <port name="data2" clock="clk"/> <!-- data lines can be broken down into smaller bit widths minimum size 10 -->
+ <port name="clk" is_clock="1"/> <!-- memories are often clocked -->
+ </input_ports>
+ <output_ports>
+ <port name="out1" clock="clk"/> <!-- output can be broken down into smaller bit widths minimum size 10 -->
+ <port name="out2" clock="clk"/> <!-- output can be broken down into smaller bit widths minimum size 10 -->
+ </output_ports>
+ </model>
+
+ <model name="adder">
+ <input_ports>
+ <port name="a" combinational_sink_ports="sumout cout"/>
+ <port name="b" combinational_sink_ports="sumout cout"/>
+ <port name="cin" combinational_sink_ports="sumout cout"/>
+ </input_ports>
+ <output_ports>
+ <port name="cout"/>
+ <port name="sumout"/>
+ </output_ports>
+ </model>
+ </models> <!-- ODIN II specific config ends -->
+
+ <layout> <!-- Physical descriptions begin -->
+ <auto_layout aspect_ratio="1.0">
+ <!--Perimeter of 'io' blocks with 'EMPTY' blocks at corners-->
+ <perimeter type="io" priority="100"/>
+ <corners type="EMPTY" priority="101"/>
+ <!--Fill with 'clb'-->
+ <fill type="clb" priority="10"/>
+ <!--Column of 'mult_27' with 'EMPTY' blocks wherever a 'mult_36' does not fit. Vertical offset by 1 for perimeter.-->
+ <col type="mult_27" startx="6" starty="1" repeatx="8" priority="20"/>
+ <col type="EMPTY" startx="6" repeatx="8" starty="1" priority="19"/>
+ <!--Column of 'memory' with 'EMPTY' blocks wherever a 'memory' does not fit. Vertical offset by 1 for perimeter.-->
+ <col type="memory" startx="2" starty="1" repeatx="8" priority="20"/>
+ <col type="EMPTY" startx="2" repeatx="8" starty="1" priority="19"/>
+ </auto_layout>
+ </layout>
+
+ <device>
+ <!-- Those values are generated using COFFE at 22nm technology -->
+ <sizing R_minW_nmos="13090" R_minW_pmos="19086.83"/>
+ <area grid_logic_tile_area="25201.9"/>
+ <chan_width_distr>
+ <x distr="uniform" peak="1.000000"/>
+ <y distr="uniform" peak="1.000000"/>
+ </chan_width_distr>
+ <switch_block type="wilton" fs="3"/>
+ <connection_block input_switch_name="ipin_cblock"/>
+ </device>
+
+ <switchlist>
+ <!-- Those values are generated using COFFE at 22nm technology -->
+ <switch type="mux" name="0" R="0.0" Cin="0.0" Cout="0.0" Tdel="237e-12" mux_trans_size="2.173" buf_size="34.22"/>
+ <switch type="mux" name="ipin_cblock" R="0.0" Cout="0.0" Cin="0.0" Tdel="146.4e-12" mux_trans_size="1.508" buf_size="12.286"/>
+ </switchlist>
+
+ <segmentlist>
+ <segment freq="1.000000" length="4" type="unidir" Rmetal="0.0" Cmetal="0.0">
+ <mux name="0"/>
+ <sb type="pattern">1 1 1 1 1</sb>
+ <cb type="pattern">1 1 1 1</cb>
+ </segment>
+ </segmentlist>
+
+ <directlist>
+ <!-- direct links connecting the two carry chains of on LAB to the LAB below it -->
+ <direct name="adder_carry1" from_pin="clb.cout[0:0]" to_pin="clb.cin[0:0]" x_offset="0" y_offset="-1" z_offset="0"/>
+ <direct name="adder_carry2" from_pin="clb.cout[1:1]" to_pin="clb.cin[1:1]" x_offset="0" y_offset="-1" z_offset="0"/>
+
+ <!-- Direct connect to right LAB -->
+ <direct name="direct_right_1" from_pin="clb.O[4:0]" to_pin="clb.I1[9:5]" x_offset="1" y_offset="0" z_offset="0"/>
+ <direct name="direct_right_2" from_pin="clb.O[24:20]" to_pin="clb.I2[9:5]" x_offset="1" y_offset="0" z_offset="0"/>
+ <direct name="direct_right_3" from_pin="clb.O[9:5]" to_pin="clb.I3[9:5]" x_offset="1" y_offset="0" z_offset="0"/>
+ <direct name="direct_right_4" from_pin="clb.O[29:25]" to_pin="clb.I4[9:5]" x_offset="1" y_offset="0" z_offset="0"/>
+
+ <!-- Direct connect to left LAB -->
+ <direct name="direct_left_1" from_pin="clb.O[14:10]" to_pin="clb.I1[14:10]" x_offset="-1" y_offset="0" z_offset="0"/>
+ <direct name="direct_left_2" from_pin="clb.O[34:30]" to_pin="clb.I2[14:10]" x_offset="-1" y_offset="0" z_offset="0"/>
+ <direct name="direct_left_3" from_pin="clb.O[19:15]" to_pin="clb.I3[14:10]" x_offset="-1" y_offset="0" z_offset="0"/>
+ <direct name="direct_left_4" from_pin="clb.O[39:35]" to_pin="clb.I4[14:10]" x_offset="-1" y_offset="0" z_offset="0"/>
+ </directlist>
+
+ <complexblocklist>
+
+ <!-- Define I/O pads begin -->
+ <!-- Capacity is a unique property of I/Os, it is the maximum number of I/Os that can be placed at the same (X,Y) location on the FPGA -->
+ <!-- Not sure of the area of an I/O (varies widely), and it's not relevant to the design of the FPGA core, so we're setting it to 0. -->
+ <pb_type name="io" capacity="8" area="0">
+ <input name="outpad" num_pins="1"/>
+ <output name="inpad" num_pins="1"/>
+ <clock name="clock" num_pins="1"/>
+
+ <!-- IOs can operate as either inputs or outputs.
+ Delays below come from Ian Kuon. They are small, so they should be interpreted as
+ the delays to and from registers in the I/O (and generally I/Os are registered
+ today and that is when you timing analyze them.
+ -->
+ <mode name="inpad">
+ <pb_type name="inpad" blif_model=".input" num_pb="1">
+ <output name="inpad" num_pins="1"/>
+ </pb_type>
+ <interconnect>
+ <direct name="inpad" input="inpad.inpad" output="io.inpad">
+ <delay_constant max="4.243e-11" in_port="inpad.inpad" out_port="io.inpad"/>
+ </direct>
+ </interconnect>
+ </mode>
+ <mode name="outpad">
+ <pb_type name="outpad" blif_model=".output" num_pb="1">
+ <input name="outpad" num_pins="1"/>
+ </pb_type>
+ <interconnect>
+ <direct name="outpad" input="io.outpad" output="outpad.outpad">
+ <delay_constant max="1.394e-11" in_port="io.outpad" out_port="outpad.outpad"/>
+ </direct>
+ </interconnect>
+ </mode>
+
+ <!-- Every input pin is driven by 15% of the tracks in a channel, every output pin is driven by 10% of the tracks in a channel -->
+ <fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
+
+ <!-- IOs go on the periphery of the FPGA, for consistency,
+ make it physically equivalent on all sides so that only one definition of I/Os is needed.
+ If I do not make a physically equivalent definition, then I need to define 4 different I/Os, one for each side of the FPGA
+ -->
+ <pinlocations pattern="custom">
+ <loc side="left">io.outpad io.inpad io.clock</loc>
+ <loc side="top">io.outpad io.inpad io.clock</loc>
+ <loc side="right">io.outpad io.inpad io.clock</loc>
+ <loc side="bottom">io.outpad io.inpad io.clock</loc>
+ </pinlocations>
+
+ <!-- Place I/Os on the sides of the FPGA -->
+ <power method="ignore"/>
+ </pb_type>
+ <!-- Define I/O pads ends -->
+
+ <!-- Define general purpose logic block (CLB) begin -->
+ <pb_type name="clb">
+ <input name="I1" num_pins="15" equivalent="full"/>
+ <input name="I2" num_pins="15" equivalent="full"/>
+ <input name="I3" num_pins="15" equivalent="full"/>
+ <input name="I4" num_pins="15" equivalent="full"/>
+ <input name="cin" num_pins="2"/>
+ <output name="O" num_pins="40" equivalent="none"/>
+ <output name="cout" num_pins="2"/>
+ <clock name="clk" num_pins="1"/>
+ <pb_type name="lab" num_pb="1">
+ <input name="I1" num_pins="15"/>
+ <input name="I2" num_pins="15"/>
+ <input name="I3" num_pins="15"/>
+ <input name="I4" num_pins="15"/>
+ <input name="cin" num_pins="2"/>
+ <output name="O" num_pins="40"/>
+ <output name="cout" num_pins="2"/>
+ <clock name="clk" num_pins="1"/>
+ <!-- Describe fracturable logic element.
+ Each fracturable logic element has a 6-LUT that can alternatively operate as two 5-LUTs with two shared inputs.
+ The outputs of the fracturable logic element can be optionally registered
+ -->
+ <pb_type name="fle" num_pb="10">
+ <input name="in" num_pins="8"/>
+ <input name="cin" num_pins="1"/>
+ <output name="out" num_pins="4"/>
+ <output name="cout" num_pins="1"/>
+ <clock name="clk" num_pins="1"/>
+ <!--
+ The ALM inputs are as follows:
+ A -> fle[0]
+ B -> fle[1]
+ C -> fle[2]
+ D -> fle[3]
+ E -> fle[4]
+ F -> fle[5]
+ G -> fle[6]
+ H -> fle[7]
+ -->
+ <mode name="n2_lut5">
+ <pb_type name="ble5" num_pb="2">
+ <input name="in" num_pins="5"/>
+ <input name="cin" num_pins="1"/>
+ <output name="out" num_pins="2"/>
+ <output name="cout" num_pins="1"/>
+ <clock name="clk" num_pins="1"/>
+ <mode name="blut5">
+ <pb_type name="flut5" num_pb="1">
+ <input name="in" num_pins="5"/>
+ <output name="out" num_pins="2"/>
+ <clock name="clk" num_pins="1"/>
+ <!-- Regular LUT mode -->
+ <pb_type name="lut5" blif_model=".names" num_pb="1" class="lut">
+ <input name="in" num_pins="5" port_class="lut_in"/>
+ <output name="out" num_pins="1" port_class="lut_out"/>
+ <!-- LUT timing using delay matrix -->
+ <!-- These are the physical delay inputs on a 4bit Adder architecture but because VPR cannot do LUT rebalancing,
+ we instead take the average of these numbers to get more stable results.
+ Note that those are the same delays for inputs A - E as the ones used for the 6-LUT, however, we have
+ subtracted the delay of the last mux stage to get the delay of inputs A - E till the 5-LUT output
+ 219.86e-12
+ 216.03e-12
+ 197.86e-12
+ 110.77e-12
+ 71.48e-12
+ -->
+ <delay_matrix type="max" in_port="lut5.in" out_port="lut5.out">
+ 163.2e-12
+ 163.2e-12
+ 163.2e-12
+ 163.2e-12
+ 163.2e-12
+ </delay_matrix>
+ </pb_type>
+ <pb_type name="ff" blif_model=".latch" num_pb="2" class="flipflop">
+ <input name="D" num_pins="1" port_class="D"/>
+ <output name="Q" num_pins="1" port_class="Q"/>
+ <clock name="clk" num_pins="1" port_class="clock"/>
+ <T_setup value="18.91e-12" port="ff.D" clock="clk"/>
+ <T_clock_to_Q max="60.32e-12" port="ff.Q" clock="clk"/>
+ </pb_type>
+ <interconnect>
+ <direct name="lut5_in" input="flut5.in" output="lut5.in"/>
+ <direct name="reg_in" input="flut5.in[0]" output="ff[0].D"/>
+ <direct name="lut5_ff" input="lut5.out" output="ff[1].D">
+ <pack_pattern name="ble5" in_port="lut5.out" out_port="ff[1].D"/>
+ <delay_constant max="18.1e-12" in_port="lut5.out" out_port="ff[1].D"/>
+ </direct>
+ <complete name="clock" input="flut5.clk" output="ff.clk"/>
+ <complete name="out_mux" input="ff.Q lut5.out" output="flut5.out">
+ <delay_constant max="43.87e-12" in_port="lut5.out" out_port="flut5.out"/>
+ <delay_constant max="43.87e-12" in_port="ff.Q" out_port="flut5.out"/>
+ </complete>
+ </interconnect>
+ </pb_type>
+ <interconnect>
+ <direct name="direct1" input="ble5.in" output="flut5.in"/>
+ <direct name="direct2" input="ble5.clk" output="flut5.clk"/>
+ <direct name="direct3" input="flut5.out" output="ble5.out"/>
+ </interconnect>
+ </mode>
+ <!-- Special dual-LUT mode that drives adder only -->
+ <mode name="arithmetic">
+ <pb_type name="arithmetic" num_pb="1">
+ <input name="in" num_pins="5"/>
+ <input name="cin" num_pins="1"/>
+ <output name="out" num_pins="2"/>
+ <output name="cout" num_pins="1"/>
+ <clock name="clk" num_pins="1"/>
+ <pb_type name="lut4" num_pb="2">
+ <input name="in" num_pins="4"/>
+ <input name="cin" num_pins="1"/>
+ <output name="out" num_pins="1"/>
+ <output name="cout" num_pins="1"/>
+ <clock name="clk" num_pins="1"/>
+ <pb_type name="lut3" blif_model=".names" num_pb="2" class="lut">
+ <input name="in" num_pins="3" port_class="lut_in"/>
+ <output name="out" num_pins="1" port_class="lut_out"/>
+ <!-- LUT timing using delay matrix -->
+ <!-- These are the physical delay inputs on a 4bit Adder architecture but because VPR cannot do LUT rebalancing,
+ we instead take the average of these numbers to get more stable results. Those are the delays from input
+ A - C to the 3-LUT output
+ 138.45e-12
+ 134.62e-12
+ 116.45e-12
+ -->
+ <!-- taking the average of the three long delays: 232e-12 -->
+ <delay_matrix type="max" in_port="lut3.in" out_port="lut3.out">
+ 129.84e-12
+ 129.84e-12
+ 129.84e-12
+ </delay_matrix>
+ </pb_type>
+ <pb_type name="adder" blif_model=".subckt adder" num_pb="1">
+ <input name="a" num_pins="1"/>
+ <input name="b" num_pins="1"/>
+ <input name="cin" num_pins="1"/>
+ <output name="cout" num_pins="1"/>
+ <output name="sumout" num_pins="1"/>
+ <delay_constant max="65.36e-12" in_port="adder.a" out_port="adder.sumout"/>
+ <delay_constant max="65.36e-12" in_port="adder.b" out_port="adder.sumout"/>
+ <delay_constant max="36.93e-12" in_port="adder.cin" out_port="adder.sumout"/>
+ <delay_constant max="44.01e-12" in_port="adder.a" out_port="adder.cout"/>
+ <delay_constant max="44.01e-12" in_port="adder.b" out_port="adder.cout"/>
+ <delay_constant max="23.18e-12" in_port="adder.cin" out_port="adder.cout"/>
+ </pb_type>
+ <pb_type name="ff" blif_model=".latch" num_pb="1" class="flipflop">
+ <input name="D" num_pins="1" port_class="D"/>
+ <output name="Q" num_pins="1" port_class="Q"/>
+ <clock name="clk" num_pins="1" port_class="clock"/>
+ <T_setup value="18.91e-12" port="ff.D" clock="clk"/>
+ <T_clock_to_Q max="60.32e-12" port="ff.Q" clock="clk"/>
+ </pb_type>
+ <interconnect> <!-- arithmetic interconnect -->
+ <direct name="clock" input="lut4.clk" output="ff.clk"/>
+ <direct name="lut_in1" input="lut4.in[1:0]" output="lut3[0:0].in[1:0]"/>
+ <mux name="input_mux1" input="lut4.in[2:2] lut4.in[3:3]" output="lut3[0:0].in[2:2]">
+ <delay_constant max="5.262e-12" in_port="lut4.in[2:2]" out_port="lut3[0:0].in[2:2]"/>
+ <delay_constant max="5.262e-12" in_port="lut4.in[3:3]" out_port="lut3[0:0].in[2:2]"/>
+ </mux>
+ <direct name="lut_in2" input="lut4.in[1:0]" output="lut3[1:1].in[1:0]"/>
+ <mux name="input_mux2" input="lut4.in[2:2] lut4.in[3:3]" output="lut3[1:1].in[2:2]">
+ <delay_constant max="5.262e-12" in_port="lut4.in[2:2]" out_port="lut3[1:1].in[2:2]"/>
+ <delay_constant max="5.262e-12" in_port="lut4.in[3:3]" out_port="lut3[1:1].in[2:2]"/>
+ </mux>
+ <direct name="lut_to_add1" input="lut3[0:0].out" output="adder.a"/>
+ <direct name="lut_to_add2" input="lut3[1:1].out" output="adder.b"/>
+ <direct name="add_to_ff" input="adder.sumout" output="ff.D">
+ <delay_constant max="18.1e-12" in_port="adder.sumout" out_port="ff.D"/>
+ <!--pack_pattern name="chain" in_port="adder.sumout" out_port="ff.D"/-->
+ </direct>
+ <direct name="carry_in" input="lut4.cin" output="adder.cin">
+ <pack_pattern name="chain" in_port="lut4.cin" out_port="adder.cin"/>
+ </direct>
+ <direct name="carry_out" input="adder.cout" output="lut4.cout">
+ <pack_pattern name="chain" in_port="adder.cout" out_port="lut4.cout"/>
+ </direct>
+ <mux name="sumout" input="ff.Q adder.sumout" output="lut4.out">
+ <delay_constant max="43.87e-12" in_port="adder.sumout" out_port="lut4.out"/>
+ <delay_constant max="43.87e-12" in_port="ff.Q" out_port="lut4.out"/>
+ </mux>
+ </interconnect>
+ </pb_type>
+ <interconnect>
+ <direct name="direct1" input="arithmetic.in[3:0]" output="lut4[0:0].in[3:0]"/>
+ <direct name="direct2" input="arithmetic.in[2:0]" output="lut4[1:1].in[2:0]"/>
+ <direct name="direct22" input="arithmetic.in[4:4]" output="lut4[1:1].in[3:3]"/>
+ <direct name="carry_in" input="arithmetic.cin" output="lut4[0:0].cin">
+ <pack_pattern name="chain" in_port="arithmetic.cin" out_port="lut4[0:0].cin"/>
+ </direct>
+ <direct name="carry_link" input="lut4[0:0].cout" output="lut4[1:1].cin">
+ <pack_pattern name="chain" in_port="lut4[0:0].cout" out_port="lut4[1:1].cin"/>
+ </direct>
+ <direct name="carry_out" input="lut4[1:1].cout" output="arithmetic.cout">
+ <pack_pattern name="chain" in_port="lut4[1:1].cout" out_port="arithmetic.cout"/>
+ </direct>
+ <complete name="complete1" input="arithmetic.clk" output="lut4[1:0].clk"/>
+ <direct name="direct4" input="lut4[0:0].out" output="arithmetic.out[0:0]"/>
+ <direct name="direct5" input="lut4[1:1].out" output="arithmetic.out[1:1]"/>
+ </interconnect>
+ </pb_type>
+ <interconnect> <!-- ble5 interconnect @ mode arithmetic -->
+ <direct name="direct1" input="ble5.in" output="arithmetic.in"/>
+ <direct name="carry_in" input="ble5.cin" output="arithmetic.cin">
+ <pack_pattern name="chain" in_port="ble5.cin" out_port="arithmetic.cin"/>
+ </direct>
+ <direct name="carry_out" input="arithmetic.cout" output="ble5.cout">
+ <pack_pattern name="chain" in_port="arithmetic.cout" out_port="ble5.cout"/>
+ </direct>
+ <direct name="direct2" input="ble5.clk" output="arithmetic.clk"/>
+ <direct name="direct3" input="arithmetic.out" output="ble5.out"/>
+ </interconnect>
+ </mode>
+ </pb_type>
+ <interconnect>
+ <!-- Shared inputs between the two 5-LUTs -->
+ <complete name="lut5_reg1" input="fle.in[0]" output="ble5[0].in[0] ble5[1].in[1]"/>
+ <complete name="lut5_reg2" input="fle.in[1]" output="ble5[0].in[1] ble5[1].in[0]"/>
+
+ <!-- Rest of the 5-LUT inputs -->
+ <direct name="lut5_inputs_1" input="fle.in[4:2]" output="ble5[0].in[4:2]"/>
+ <direct name="lut5_inputs_22" input="fle.in[7:5]" output="ble5[1].in[4:2]"/>
+
+ <direct name="lut5_outputs_1" input="ble5[0].out" output="fle.out[1:0]"/>
+ <direct name="lut5_outputs_2" input="ble5[1].out" output="fle.out[3:2]"/>
+
+ <direct name="carry_in" input="fle.cin" output="ble5[0].cin">
+ <pack_pattern name="chain" in_port="fle.cin" out_port="ble5[0].cin"/>
+ </direct>
+ <direct name="carry_out" input="ble5[1].cout" output="fle.cout">
+ <pack_pattern name="chain" in_port="ble5[1].cout" out_port="fle.cout"/>
+ </direct>
+ <direct name="carry_link" input="ble5[0].cout" output="ble5[1].cin">
+ <pack_pattern name="chain" in_port="ble5[0].cout" out_port="ble5[1].cout"/>
+ </direct>
+ <complete name="clock" input="fle.clk" output="ble5[1:0].clk"/>
+ </interconnect>
+ </mode> <!-- n2_lut5 -->
+ <mode name="n1_lut6">
+ <pb_type name="ble6" num_pb="1">
+ <input name="in" num_pins="6"/>
+ <output name="out" num_pins="4"/>
+ <clock name="clk" num_pins="1"/>
+ <pb_type name="lut6" blif_model=".names" num_pb="1" class="lut">
+ <input name="in" num_pins="6" port_class="lut_in"/>
+ <output name="out" num_pins="1" port_class="lut_out"/>
+ <!-- LUT timing using delay matrix -->
+ <!-- These are the physical delay inputs on a 4bit Adder architecture LUT but because VPR cannot do LUT rebalancing,
+ we instead take the average of these numbers to get more stable results
+ 264.27e-12
+ 260.44e-12
+ 242.27e-12
+ 155.18e-12
+ 115.89e-12
+ 78.67e-12
+ -->
+ <delay_matrix type="max" in_port="lut6.in" out_port="lut6.out">
+ 186.12e-12
+ 186.12e-12
+ 186.12e-12
+ 186.12e-12
+ 186.12e-12
+ 186.12e-12
+ </delay_matrix>
+ </pb_type>
+ <pb_type name="ff" blif_model=".latch" num_pb="2" class="flipflop">
+ <input name="D" num_pins="1" port_class="D"/>
+ <output name="Q" num_pins="1" port_class="Q"/>
+ <clock name="clk" num_pins="1" port_class="clock"/>
+ <T_setup value="18.91e-12" port="ff.D" clock="clk"/>
+ <T_clock_to_Q max="60.32e-12" port="ff.Q" clock="clk"/>
+ </pb_type>
+ <interconnect>
+ <direct name="lut6_inputs" input="ble6.in" output="lut6.in"/>
+ <direct name="lut6_ff" input="lut6.out" output="ff[1].D">
+ <delay_constant max="18.1e-12" in_port="lut6.out" out_port="ff[1].D"/>
+ <pack_pattern name="ble6" in_port="lut6.out" out_port="ff[1].D"/>
+ </direct>
+ <complete name="clock" input="ble6.clk" output="ff.clk"/>
+ <direct name="input_to_ff" input="ble6.in[0]" output="ff[0].D"/>
+ <complete name="mux1" input="ff[0].Q lut6.out" output="ble6.out[1:0]">
+ <delay_constant max="43.87e-12" in_port="lut6.out" out_port="ble6.out[1:0]"/>
+ <delay_constant max="43.87e-12" in_port="ff[0].Q" out_port="ble6.out[1:0]"/>
+ </complete>
+ <complete name="mux2" input="ff[1].Q lut6.out" output="ble6.out[3:2]">
+ <delay_constant max="43.87e-12" in_port="lut6.out" out_port="ble6.out[3:2]"/>
+ <delay_constant max="43.87e-12" in_port="ff[1].Q" out_port="ble6.out[3:2]"/>
+ </complete>
+ </interconnect>
+ </pb_type>
+ <interconnect>
+ <!-- ble6 takes inputs A, B, C, D, E, & F; where F is fle[7] -->
+ <direct name="lut6_inputs1" input="fle.in[4:0]" output="ble6.in[4:0]"/>
+ <direct name="lut6_inputs2" input="fle.in[7]" output="ble6.in[5]"/>
+ <direct name="direct2" input="ble6.out" output="fle.out"/>
+ <direct name="direct4" input="fle.clk" output="ble6.clk"/>
+ </interconnect>
+ </mode> <!-- n1_lut6 -->
+ </pb_type>
+ <interconnect>
+ <!-- We use a 50% depop crossbar built using small full xbars to get sets of logically equivalent pins at inputs of CLB -->
+ <complete name="lutA" input="lab.I4 lab.I3" output="fle[9:0].in[0:0]">
+ <delay_constant max="75.11e-12" in_port="lab.I4" out_port="fle.in[0:0]"/>
+ <delay_constant max="75.11e-12" in_port="lab.I3" out_port="fle.in[0:0]"/>
+ </complete>
+ <complete name="lutB" input="lab.I3 lab.I2" output="fle[9:0].in[1:1]">
+ <delay_constant max="75.11e-12" in_port="lab.I3" out_port="fle.in[1:1]"/>
+ <delay_constant max="75.11e-12" in_port="lab.I2" out_port="fle.in[1:1]"/>
+ </complete>
+ <complete name="lutC" input="lab.I2 lab.I1" output="fle[9:0].in[2:2]">
+ <delay_constant max="75.11e-12" in_port="lab.I2" out_port="fle.in[2:2]"/>
+ <delay_constant max="75.11e-12" in_port="lab.I1" out_port="fle.in[2:2]"/>
+ </complete>
+ <complete name="lutD" input="lab.I4 lab.I2" output="fle[9:0].in[3:3]">
+ <delay_constant max="75.11e-12" in_port="lab.I4" out_port="fle.in[3:3]"/>
+ <delay_constant max="75.11e-12" in_port="lab.I2" out_port="fle.in[3:3]"/>
+ </complete>
+ <complete name="lutE" input="lab.I3 lab.I1" output="fle[9:0].in[4:4]">
+ <delay_constant max="75.11e-12" in_port="lab.I3" out_port="fle.in[4:4]"/>
+ <delay_constant max="75.11e-12" in_port="lab.I1" out_port="fle.in[4:4]"/>
+ </complete>
+ <complete name="lutF" input="lab.I4 lab.I1" output="fle[9:0].in[5:5]">
+ <delay_constant max="75.11e-12" in_port="lab.I4" out_port="fle.in[5:5]"/>
+ <delay_constant max="75.11e-12" in_port="lab.I1" out_port="fle.in[5:5]"/>
+ </complete>
+ <complete name="lutG" input="lab.I4 lab.I3" output="fle[9:0].in[6:6]">
+ <delay_constant max="75.11e-12" in_port="lab.I4" out_port="fle.in[6:6]"/>
+ <delay_constant max="75.11e-12" in_port="lab.I3" out_port="fle.in[6:6]"/>
+ </complete>
+ <complete name="lutH" input="lab.I3 lab.I2" output="fle[9:0].in[7:7]">
+ <delay_constant max="75.11e-12" in_port="lab.I3" out_port="fle.in[7:7]"/>
+ <delay_constant max="75.11e-12" in_port="lab.I2" out_port="fle.in[7:7]"/>
+ </complete>
+
+ <complete name="clks" input="lab.clk" output="fle[9:0].clk"/>
+
+ <!-- This way of specifying direct connection to clb outputs is important because this architecture uses automatic spreading of opins.
+ By grouping to output pins in this fashion, if a logic block is completely filled by 6-LUTs,
+ then the outputs those 6-LUTs take get evenly distributed across all four sides of the CLB instead of clumped on two sides (which is what happens with a more
+ naive specification).
+ -->
+ <direct name="labouts11" input="fle[9:0].out[0]" output="lab.O[9:0]"/>
+ <direct name="labouts12" input="fle[9:0].out[1]" output="lab.O[19:10]"/>
+ <direct name="labouts13" input="fle[9:0].out[2]" output="lab.O[29:20]"/>
+ <direct name="labouts14" input="fle[9:0].out[3]" output="lab.O[39:30]"/>
+
+ <!-- Carry chain links -->
+ <direct name="carry_in1" input="lab.cin[0:0]" output="fle[0:0].cin">
+ <!-- Put all inter-block carry chain delay on this one edge -->
+ <delay_constant max="17.8e-12" in_port="lab.cin[0:0]" out_port="fle[0:0].cin"/>
+ <pack_pattern name="chain" in_port="lab.cin[0:0]" out_port="fle[0:0].cin"/>
+ </direct>
+ <direct name="carry_in2" input="lab.cin[1:1]" output="fle[5:5].cin">
+ <!-- Put all inter-block carry chain delay on this one edge -->
+ <delay_constant max="18.47e-12" in_port="lab.cin[1:1]" out_port="fle[5:5].cin"/>
+ <pack_pattern name="chain" in_port="lab.cin[1:1]" out_port="fle[5:5].cin"/>
+ </direct>
+ <direct name="carry_out1" input="fle[4:4].cout" output="lab.cout[0:0]">
+ <pack_pattern name="chain" in_port="fle[4:4].cout" out_port="lab.cout[0:0]"/>
+ </direct>
+ <direct name="carry_out2" input="fle[9:9].cout" output="lab.cout[1:1]">
+ <pack_pattern name="chain" in_port="fle[9:9].cout" out_port="lab.cout[1:1]"/>
+ </direct>
+ <direct name="carry_link1" input="fle[3:0].cout" output="fle[4:1].cin">
+ <pack_pattern name="chain" in_port="fle[3:0].cout" out_port="fle[4:1].cin"/>
+ </direct>
+ <direct name="carry_link2" input="fle[8:5].cout" output="fle[9:6].cin">
+ <pack_pattern name="chain" in_port="fle[8:5].cout" out_port="fle[9:6].cin"/>
+ </direct>
+ </interconnect>
+ </pb_type>
+ <interconnect>
+
+ <direct name="carry_in1" input="clb.cin[0:0]" output="lab.cin[0:0]"/>
+ <direct name="carry_out1" input="lab.cout[0:0]" output="clb.cout[0:0]"/>
+ <direct name="carry_in2" input="clb.cin[1:1]" output="lab.cin[1:1]"/>
+ <direct name="carry_out2" input="lab.cout[1:1]" output="clb.cout[1:1]"/>
+
+ <direct name="clock" input="clb.clk" output="lab.clk"/>
+
+ <!-- local feedback signals from two outputs per ALM sharing LAB inputs with
+ routing channel inputs -->
+ <complete name="Input_feedback_I1" input="clb.I1 lab.O[4:0]" output="lab.I1"/>
+ <complete name="Input_feedback_I2" input="clb.I2 lab.O[24:20]" output="lab.I2"/>
+ <complete name="Input_feedback_I3" input="clb.I3 lab.O[9:5]" output="lab.I3"/>
+ <complete name="Input_feedback_I4" input="clb.I4 lab.O[29:25]" output="lab.I4"/>
+
+ <direct name="output" input="lab.O" output="clb.O"/>
+ </interconnect>
+ <fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10">
+ <fc_override port_name="cin" fc_type="frac" fc_val="0"/>
+ <fc_override port_name="cout" fc_type="frac" fc_val="0"/>
+ </fc>
+ <pinlocations pattern="spread"/>
+ </pb_type>
+ <!-- Define general purpose logic block (CLB) ends -->
+
+ <!-- Define fracturable multiplier begin -->
+ <!-- This multiplier can operate as a 27x27 multiplier that can fracture to two 18x19 multipliers
+ the delays of thsi multiplier are from Arria 10 which is a 22nm chip -->
+ <pb_type name="mult_27" height="2">
+ <input name="datain" num_pins="74"/>
+ <output name="dataout" num_pins="74"/>
+
+ <mode name="two_mult_18x19">
+ <pb_type name="two_mult_18x19" num_pb="2">
+ <input name="a" num_pins="18"/>
+ <input name="b" num_pins="19"/>
+ <output name="out" num_pins="37"/>
+ <pb_type name="mult_18x19" blif_model=".subckt multiply" num_pb="1">
+ <input name="a" num_pins="18"/>
+ <input name="b" num_pins="19"/>
+ <output name="out" num_pins="37"/>
+ <!-- Using the numbers from Arria 10 which is a 22nm technology, an 18x19 multiplier
+ can operate at 548 MHz which maps to a delay of 1.825e-9 -->
+ <delay_constant max="1.825e-9" in_port="mult_18x19.a" out_port="mult_18x19.out"/>
+ <delay_constant max="1.825e-9" in_port="mult_18x19.b" out_port="mult_18x19.out"/>
+ </pb_type>
+ <interconnect>
+ <direct name="a2a" input="two_mult_18x19.a" output="mult_18x19.a">
+ </direct>
+ <direct name="b2b" input="two_mult_18x19.b" output="mult_18x19.b">
+ </direct>
+ <direct name="out2out" input="mult_18x19.out" output="two_mult_18x19.out">
+ </direct>
+ </interconnect>
+ <power method="pin-toggle">
+ <port name="a" energy_per_toggle="1.09e-12"/>
+ <port name="b" energy_per_toggle="1.09e-12"/>
+ <static_power power_per_instance="0.0"/>
+ </power>
+ </pb_type>
+ <interconnect>
+ <!-- Stratix IV input delay of 207ps is conservative for this architecture because this architecture does not have an input crossbar in the multiplier.
+ Subtract 72.5 ps delay, which is already in the connection block input mux, leading
+ -->
+ <direct name="datain2a1" input="mult_27.datain[17:0]" output="two_mult_18x19[0].a">
+ <delay_constant max="134e-12" in_port="mult_27.datain[17:0]" out_port="two_mult_18x19[0].a"/>
+ </direct>
+ <direct name="datain2b1" input="mult_27.datain[36:18]" output="two_mult_18x19[0].b">
+ <delay_constant max="134e-12" in_port="mult_27.datain[36:18]" out_port="two_mult_18x19[0].b"/>
+ </direct>
+ <direct name="datain2a2" input="mult_27.datain[54:37]" output="two_mult_18x19[1].a">
+ <delay_constant max="134e-12" in_port="mult_27.datain[54:37]" out_port="two_mult_18x19[1].a"/>
+ </direct>
+ <direct name="datain2b2" input="mult_27.datain[73:55]" output="two_mult_18x19[1].b">
+ <delay_constant max="134e-12" in_port="mult_27.datain[73:55]" out_port="two_mult_18x19[1].b"/>
+ </direct>
+ <direct name="out2dataout" input="two_mult_18x19[1:0].out" output="mult_27.dataout">
+ <delay_constant max="1.09e-9" in_port="two_mult_18x19[1:0].out" out_port="mult_27.dataout"/>
+ </direct>
+ </interconnect>
+ </mode>
+
+ <mode name="mult_27x27">
+ <pb_type name="one_mult_27x27" num_pb="1">
+ <input name="a" num_pins="27"/>
+ <input name="b" num_pins="27"/>
+ <output name="out" num_pins="54"/>
+
+ <pb_type name="mult_27x27" blif_model=".subckt multiply" num_pb="1">
+ <input name="a" num_pins="27"/>
+ <input name="b" num_pins="27"/>
+ <output name="out" num_pins="54"/>
+ <!-- Using the numbers from Arria 10 which is a 22nm technology, an 27x27 multiplier
+ can operate at 541 MHz which maps to a delay of 1.848e-9 -->
+ <delay_constant max="1.848e-9" in_port="mult_27x27.a" out_port="mult_27x27.out"/>
+ <delay_constant max="1.848e-9" in_port="mult_27x27.b" out_port="mult_27x27.out"/>
+ </pb_type>
+ <interconnect>
+ <direct name="a2a" input="one_mult_27x27.a" output="mult_27x27.a">
+ </direct>
+ <direct name="b2b" input="one_mult_27x27.b" output="mult_27x27.b">
+ </direct>
+ <direct name="out2out" input="mult_27x27.out" output="one_mult_27x27.out">
+ </direct>
+ </interconnect>
+ <power method="pin-toggle">
+ <port name="a" energy_per_toggle="2.13e-12"/>
+ <port name="b" energy_per_toggle="2.13e-12"/>
+ <static_power power_per_instance="0.0"/>
+ </power>
+ </pb_type>
+ <interconnect>
+ <!-- Stratix IV input delay of 207ps is conservative for this architecture because this architecture does not have an input crossbar in the multiplier.
+ Subtract 72.5 ps delay, which is already in the connection block input mux, leading
+ to a 134 ps delay.
+ -->
+ <direct name="datain2a" input="mult_27.datain[26:0]" output="one_mult_27x27.a">
+ <delay_constant max="134e-12" in_port="mult_27.datain[26:0]" out_port="one_mult_27x27.a"/>
+ </direct>
+ <direct name="datain2b" input="mult_27.datain[53:27]" output="one_mult_27x27.b">
+ <delay_constant max="134e-12" in_port="mult_27.datain[53:27]" out_port="one_mult_27x27.b"/>
+ </direct>
+ <direct name="out2dataout" input="one_mult_27x27.out" output="mult_27.dataout[53:0]">
+ <delay_constant max="1.93e-9" in_port="one_mult_27x27.out" out_port="mult_27.dataout[53:0]"/>
+ </direct>
+ </interconnect>
+
+ </mode>
+
+ <fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
+ <pinlocations pattern="spread"/>
+
+ <!-- Place this multiplier block every 8 columns from (and including) the sixth column -->
+ <power method="sum-of-children"/>
+ </pb_type>
+ <!-- Define fracturable multiplier end -->
+
+ <!-- Define fracturable memory begin -->
+ <!-- The architecture also has a 20Kb memory that has true and simple dual port modes. In simple
+ dual port mode the memory can be configured in the following modes: 512x40, 1024x20 and 2048x10
+ While in true dual port mode it can be configured as: 1024x20 and 2028x10. -->
+ <pb_type name="memory" height="4">
+ <input name="addr1" num_pins="11"/>
+ <input name="addr2" num_pins="11"/>
+ <input name="data" num_pins="40"/>
+ <input name="we1" num_pins="1"/>
+ <input name="we2" num_pins="1"/>
+ <output name="out" num_pins="40"/>
+ <clock name="clk" num_pins="1"/>
+
+ <!-- Specify single port mode first -->
+ <mode name="mem_512x40_sp">
+ <pb_type name="mem_512x40_sp" blif_model=".subckt single_port_ram" class="memory" num_pb="1">
+ <input name="addr" num_pins="9" port_class="address"/>
+ <input name="data" num_pins="40" port_class="data_in"/>
+ <input name="we" num_pins="1" port_class="write_en"/>
+ <output name="out" num_pins="40" port_class="data_out"/>
+ <clock name="clk" num_pins="1" port_class="clock"/>
+ <T_setup value="509e-12" port="mem_512x40_sp.addr" clock="clk"/>
+ <T_setup value="509e-12" port="mem_512x40_sp.data" clock="clk"/>
+ <T_setup value="509e-12" port="mem_512x40_sp.we" clock="clk"/>
+ <T_clock_to_Q max="1.234e-9" port="mem_512x40_sp.out" clock="clk"/>
+ <power method="pin-toggle">
+ <port name="clk" energy_per_toggle="9.0e-12"/>
+ <static_power power_per_instance="0.0"/>
+ </power>
+ </pb_type>
+ <interconnect>
+ <direct name="address1" input="memory.addr1[8:0]" output="mem_512x40_sp.addr">
+ <delay_constant max="132e-12" in_port="memory.addr1[8:0]" out_port="mem_512x40_sp.addr"/>
+ </direct>
+ <direct name="data1" input="memory.data" output="mem_512x40_sp.data">
+ <delay_constant max="132e-12" in_port="memory.data" out_port="mem_512x40_sp.data"/>
+ </direct>
+ <direct name="writeen1" input="memory.we1" output="mem_512x40_sp.we">
+ <delay_constant max="132e-12" in_port="memory.we1" out_port="mem_512x40_sp.we"/>
+ </direct>
+ <direct name="dataout1" input="mem_512x40_sp.out" output="memory.out">
+ <delay_constant max="40e-12" in_port="mem_512x40_sp.out" out_port="memory.out"/>
+ </direct>
+ <direct name="clk" input="memory.clk" output="mem_512x40_sp.clk">
+ </direct>
+ </interconnect>
+ </mode>
+
+ <mode name="mem_1024x20_sp">
+ <pb_type name="mem_1024x20_sp" blif_model=".subckt single_port_ram" class="memory" num_pb="1">
+ <input name="addr" num_pins="10" port_class="address"/>
+ <input name="data" num_pins="20" port_class="data_in"/>
+ <input name="we" num_pins="1" port_class="write_en"/>
+ <output name="out" num_pins="20" port_class="data_out"/>
+ <clock name="clk" num_pins="1" port_class="clock"/>
+ <T_setup value="509e-12" port="mem_1024x20_sp.addr" clock="clk"/>
+ <T_setup value="509e-12" port="mem_1024x20_sp.data" clock="clk"/>
+ <T_setup value="509e-12" port="mem_1024x20_sp.we" clock="clk"/>
+ <T_clock_to_Q max="1.234e-9" port="mem_1024x20_sp.out" clock="clk"/>
+ <power method="pin-toggle">
+ <port name="clk" energy_per_toggle="9.0e-12"/>
+ <static_power power_per_instance="0.0"/>
+ </power>
+ </pb_type>
+ <interconnect>
+ <direct name="address1" input="memory.addr1[9:0]" output="mem_1024x20_sp.addr">
+ <delay_constant max="132e-12" in_port="memory.addr1[9:0]" out_port="mem_1024x20_sp.addr"/>
+ </direct>
+ <direct name="data1" input="memory.data[19:0]" output="mem_1024x20_sp.data">
+ <delay_constant max="132e-12" in_port="memory.data[19:0]" out_port="mem_1024x20_sp.data"/>
+ </direct>
+ <direct name="writeen1" input="memory.we1" output="mem_1024x20_sp.we">
+ <delay_constant max="132e-12" in_port="memory.we1" out_port="mem_1024x20_sp.we"/>
+ </direct>
+ <direct name="dataout1" input="mem_1024x20_sp.out" output="memory.out[19:0]">
+ <delay_constant max="40e-12" in_port="mem_1024x20_sp.out" out_port="memory.out[19:0]"/>
+ </direct>
+ <direct name="clk" input="memory.clk" output="mem_1024x20_sp.clk">
+ </direct>
+ </interconnect>
+ </mode>
+
+ <mode name="mem_2048x10_sp">
+ <pb_type name="mem_2048x10_sp" blif_model=".subckt single_port_ram" class="memory" num_pb="1">
+ <input name="addr" num_pins="11" port_class="address"/>
+ <input name="data" num_pins="10" port_class="data_in"/>
+ <input name="we" num_pins="1" port_class="write_en"/>
+ <output name="out" num_pins="10" port_class="data_out"/>
+ <clock name="clk" num_pins="1" port_class="clock"/>
+ <T_setup value="509e-12" port="mem_2048x10_sp.addr" clock="clk"/>
+ <T_setup value="509e-12" port="mem_2048x10_sp.data" clock="clk"/>
+ <T_setup value="509e-12" port="mem_2048x10_sp.we" clock="clk"/>
+ <T_clock_to_Q max="1.234e-9" port="mem_2048x10_sp.out" clock="clk"/>
+ <power method="pin-toggle">
+ <port name="clk" energy_per_toggle="9.0e-12"/>
+ <static_power power_per_instance="0.0"/>
+ </power>
+ </pb_type>
+ <interconnect>
+ <direct name="address1" input="memory.addr1[10:0]" output="mem_2048x10_sp.addr">
+ <delay_constant max="132e-12" in_port="memory.addr1[10:0]" out_port="mem_2048x10_sp.addr"/>
+ </direct>
+ <direct name="data1" input="memory.data[9:0]" output="mem_2048x10_sp.data">
+ <delay_constant max="132e-12" in_port="memory.data[9:0]" out_port="mem_2048x10_sp.data"/>
+ </direct>
+ <direct name="writeen1" input="memory.we1" output="mem_2048x10_sp.we">
+ <delay_constant max="132e-12" in_port="memory.we1" out_port="mem_2048x10_sp.we"/>
+ </direct>
+ <direct name="dataout1" input="mem_2048x10_sp.out" output="memory.out[9:0]">
+ <delay_constant max="40e-12" in_port="mem_2048x10_sp.out" out_port="memory.out[9:0]"/>
+ </direct>
+ <direct name="clk" input="memory.clk" output="mem_2048x10_sp.clk">
+ </direct>
+ </interconnect>
+ </mode>
+
+ <!-- Specify true dual port mode next -->
+ <mode name="mem_1024x20_dp">
+ <pb_type name="mem_1024x20_dp" blif_model=".subckt dual_port_ram" class="memory" num_pb="1">
+ <input name="addr1" num_pins="10" port_class="address1"/>
+ <input name="addr2" num_pins="10" port_class="address2"/>
+ <input name="data1" num_pins="20" port_class="data_in1"/>
+ <input name="data2" num_pins="20" port_class="data_in2"/>
+ <input name="we1" num_pins="1" port_class="write_en1"/>
+ <input name="we2" num_pins="1" port_class="write_en2"/>
+ <output name="out1" num_pins="20" port_class="data_out1"/>
+ <output name="out2" num_pins="20" port_class="data_out2"/>
+ <clock name="clk" num_pins="1" port_class="clock"/>
+ <T_setup value="509e-12" port="mem_1024x20_dp.addr1" clock="clk"/> <T_setup value="509e-12" port="mem_1024x20_dp.data1" clock="clk"/>
+ <T_setup value="509e-12" port="mem_1024x20_dp.we1" clock="clk"/>
+ <T_setup value="509e-12" port="mem_1024x20_dp.addr2" clock="clk"/>
+ <T_setup value="509e-12" port="mem_1024x20_dp.data2" clock="clk"/>
+ <T_setup value="509e-12" port="mem_1024x20_dp.we2" clock="clk"/>
+ <T_clock_to_Q max="1.234e-9" port="mem_1024x20_dp.out1" clock="clk"/>
+ <T_clock_to_Q max="1.234e-9" port="mem_1024x20_dp.out2" clock="clk"/>
+ <power method="pin-toggle">
+ <port name="clk" energy_per_toggle="17.9e-12"/>
+ <static_power power_per_instance="0.0"/>
+ </power>
+ </pb_type>
+ <interconnect>
+ <direct name="address1" input="memory.addr1[9:0]" output="mem_1024x20_dp.addr1">
+ <delay_constant max="132e-12" in_port="memory.addr1[9:0]" out_port="mem_1024x20_dp.addr1"/>
+ </direct>
+ <direct name="address2" input="memory.addr2[9:0]" output="mem_1024x20_dp.addr2">
+ <delay_constant max="132e-12" in_port="memory.addr2[9:0]" out_port="mem_1024x20_dp.addr2"/>
+ </direct>
+ <direct name="data1" input="memory.data[19:0]" output="mem_1024x20_dp.data1">
+ <delay_constant max="132e-12" in_port="memory.data[19:0]" out_port="mem_1024x20_dp.data1"/>
+ </direct>
+ <direct name="data2" input="memory.data[39:20]" output="mem_1024x20_dp.data2">
+ <delay_constant max="132e-12" in_port="memory.data[39:20]" out_port="mem_1024x20_dp.data2"/>
+ </direct>
+ <direct name="writeen1" input="memory.we1" output="mem_1024x20_dp.we1">
+ <delay_constant max="132e-12" in_port="memory.we1" out_port="mem_1024x20_dp.we1"/>
+ </direct>
+ <direct name="writeen2" input="memory.we2" output="mem_1024x20_dp.we2">
+ <delay_constant max="132e-12" in_port="memory.we2" out_port="mem_1024x20_dp.we2"/>
+ </direct>
+ <direct name="dataout1" input="mem_1024x20_dp.out1" output="memory.out[19:0]">
+ <delay_constant max="40e-12" in_port="mem_1024x20_dp.out1" out_port="memory.out[19:0]"/>
+ </direct>
+ <direct name="dataout2" input="mem_1024x20_dp.out2" output="memory.out[39:20]">
+ <delay_constant max="40e-12" in_port="mem_1024x20_dp.out2" out_port="memory.out[39:20]"/>
+ </direct>
+ <direct name="clk" input="memory.clk" output="mem_1024x20_dp.clk">
+ </direct>
+ </interconnect>
+ </mode>
+
+ <mode name="mem_2048x10_dp">
+ <pb_type name="mem_2048x10_dp" blif_model=".subckt dual_port_ram" class="memory" num_pb="1">
+ <input name="addr1" num_pins="11" port_class="address1"/>
+ <input name="addr2" num_pins="11" port_class="address2"/>
+ <input name="data1" num_pins="10" port_class="data_in1"/>
+ <input name="data2" num_pins="10" port_class="data_in2"/>
+ <input name="we1" num_pins="1" port_class="write_en1"/>
+ <input name="we2" num_pins="1" port_class="write_en2"/>
+ <output name="out1" num_pins="10" port_class="data_out1"/>
+ <output name="out2" num_pins="10" port_class="data_out2"/>
+ <clock name="clk" num_pins="1" port_class="clock"/>
+ <T_setup value="509e-12" port="mem_2048x10_dp.addr1" clock="clk"/>
+ <T_setup value="509e-12" port="mem_2048x10_dp.data1" clock="clk"/>
+ <T_setup value="509e-12" port="mem_2048x10_dp.we1" clock="clk"/>
+ <T_setup value="509e-12" port="mem_2048x10_dp.addr2" clock="clk"/>
+ <T_setup value="509e-12" port="mem_2048x10_dp.data2" clock="clk"/>
+ <T_setup value="509e-12" port="mem_2048x10_dp.we2" clock="clk"/>
+ <T_clock_to_Q max="1.234e-9" port="mem_2048x10_dp.out1" clock="clk"/>
+ <T_clock_to_Q max="1.234e-9" port="mem_2048x10_dp.out2" clock="clk"/>
+ <power method="pin-toggle">
+ <port name="clk" energy_per_toggle="17.9e-12"/>
+ <static_power power_per_instance="0.0"/>
+ </power>
+ </pb_type>
+ <interconnect>
+ <direct name="address1" input="memory.addr1[10:0]" output="mem_2048x10_dp.addr1">
+ <delay_constant max="132e-12" in_port="memory.addr1[10:0]" out_port="mem_2048x10_dp.addr1"/>
+ </direct>
+ <direct name="address2" input="memory.addr2[10:0]" output="mem_2048x10_dp.addr2">
+ <delay_constant max="132e-12" in_port="memory.addr2[10:0]" out_port="mem_2048x10_dp.addr2"/>
+ </direct>
+ <direct name="data1" input="memory.data[9:0]" output="mem_2048x10_dp.data1">
+ <delay_constant max="132e-12" in_port="memory.data[9:0]" out_port="mem_2048x10_dp.data1"/>
+ </direct>
+ <direct name="data2" input="memory.data[19:10]" output="mem_2048x10_dp.data2">
+ <delay_constant max="132e-12" in_port="memory.data[19:10]" out_port="mem_2048x10_dp.data2"/>
+ </direct>
+ <direct name="writeen1" input="memory.we1" output="mem_2048x10_dp.we1">
+ <delay_constant max="132e-12" in_port="memory.we1" out_port="mem_2048x10_dp.we1"/>
+ </direct>
+ <direct name="writeen2" input="memory.we2" output="mem_2048x10_dp.we2">
+ <delay_constant max="132e-12" in_port="memory.we2" out_port="mem_2048x10_dp.we2"/>
+ </direct>
+ <direct name="dataout1" input="mem_2048x10_dp.out1" output="memory.out[9:0]">
+ <delay_constant max="40e-12" in_port="mem_2048x10_dp.out1" out_port="memory.out[9:0]"/>
+ </direct>
+ <direct name="dataout2" input="mem_2048x10_dp.out2" output="memory.out[19:10]">
+ <delay_constant max="40e-12" in_port="mem_2048x10_dp.out2" out_port="memory.out[19:10]"/>
+ </direct>
+ <direct name="clk" input="memory.clk" output="mem_2048x10_dp.clk">
+ </direct>
+ </interconnect>
+ </mode>
+
+ <!-- Every input pin is driven by 15% of the tracks in a channel, every output pin is driven by 10% of the tracks in a channel -->
+ <fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
+ <pinlocations pattern="spread"/>
+
+ <!-- Place this memory block every 8 columns from (and including) the second column -->
+ <power method="sum-of-children"/>
+ </pb_type>
+ <!-- Define fracturable memory end -->
+
+
+ </complexblocklist>
+
+ <power>
+ <local_interconnect C_wire="2.5e-10"/>
+ <mux_transistor_size mux_transistor_size="3"/>
+ <FF_size FF_size="4"/>
+ <LUT_transistor_size LUT_transistor_size="4"/>
+ </power>
+
+ <clocks>
+ <clock buffer_size="auto" C_wire="2.5e-10"/>
+ </clocks>
+</architecture>
