riscv
/
OpenFPGA
mirror of https://github.com/lnis-uofu/OpenFPGA.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

Merge pull request #277 from lnis-uofu/hetergeneous_arch 

Support Verilog-to-Verification for heterogeneous FPGA
This commit is contained in:
ganeshgore 2021-03-23 19:48:35 -06:00 committed by GitHub
parent 93abfff2bb 44d97ead86
commit 3ea62bb4b3
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
10 changed files with 1482 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
openfpga_flow/benchmarks/micro_benchmark/mac_8/mac_8.v Normal file
View File

@ -0,0 +1,22 @@
//-------------------------------------------------------
// Functionality: A 8-bit multiply-acculumate circuit
// Author: Xifan Tang
//-------------------------------------------------------
module mac_8(a, b, c, out);
parameter DATA_WIDTH = 8; /* declare a parameter. default required */
input [DATA_WIDTH - 1 : 0] a, b, c;
output [DATA_WIDTH - 1 : 0] out;
assign out = a * b + c;
endmodule

96
openfpga_flow/misc/ys_tmpl_yosys_vpr_dsp_flow.ys Normal file
View File

@ -0,0 +1,96 @@
# Yosys synthesis script for ${TOP_MODULE}
#########################
# Parse input files
#########################
# Read verilog files
${READ_VERILOG_FILE}
# Read technology library
read_verilog -lib -specify ${YOSYS_CELL_SIM_VERILOG}
#########################
# Prepare for synthesis
#########################
# Identify top module from hierarchy
hierarchy -check -top ${TOP_MODULE}
# - Convert process blocks to AST
proc
# Flatten all the gates/primitives
flatten
# Identify tri-state buffers from 'z' signal in AST
# with follow-up optimizations to clean up AST
tribuf -logic
opt_expr
opt_clean
# demote inout ports to input or output port
# with follow-up optimizations to clean up AST
deminout
opt
opt_expr
opt_clean
check
opt
wreduce -keepdc
peepopt
pmuxtree
opt_clean
########################
# Map multipliers
# Inspired from synth_xilinx.cc
#########################
# Avoid merging any registers into DSP, reserve memory port registers first
memory_dff
wreduce t:$mul
techmap -map +/mul2dsp.v -map ${YOSYS_DSP_MAP_VERILOG} ${YOSYS_DSP_MAP_PARAMETERS}
select a:mul2dsp
setattr -unset mul2dsp
opt_expr -fine
wreduce
select -clear
chtype -set $mul t:$__soft_mul# Extract arithmetic functions
#########################
# Run coarse synthesis
#########################
# Run a tech map with default library
techmap
alumacc
share
opt
fsm
# Run a quick follow-up optimization to sweep out unused nets/signals
opt -fast
# Optimize any memory cells by merging share-able ports and collecting all the ports belonging to memorcy cells
memory -nomap
opt_clean
#########################
# Map flip-flops
#########################
techmap -map +/adff2dff.v
opt_expr -mux_undef
simplemap
opt_expr
opt_merge
opt_rmdff
opt_clean
opt
#########################
# Map LUTs
#########################
abc -lut ${LUT_SIZE}
#########################
# Check and show statisitics
#########################
hierarchy -check
stat
#########################
# Output netlists
#########################
opt_clean -purge
write_blif ${OUTPUT_BLIF}

290
openfpga_flow/openfpga_arch/k4_frac_N8_reset_softadder_register_scan_chain_dsp8_caravel_io_skywater130nm_fdhd_cc_openfpga.xml Normal file
View File

@ -0,0 +1,290 @@
<!-- Architecture annotation for OpenFPGA framework
This annotation supports the k4_frac_cc_sky130nm.xml
- General purpose logic block
- K = 6, N = 10, I = 40
- Single mode
- Routing architecture
- L = 4, fc_in = 0.15, fc_out = 0.1
- Skywater 130nm PDK
- circuit models are binded to the opensource skywater
foundry middle-speed (ms) standard cell library
-->
<openfpga_architecture>
<technology_library>
<device_library>
<device_model name="logic" type="transistor">
<lib type="industry" corner="TOP_TT" ref="M" path="${OPENFPGA_PATH}/openfpga_flow/tech/PTM_45nm/45nm.pm"/>
<design vdd="0.9" pn_ratio="2"/>
<pmos name="pch" chan_length="40e-9" min_width="140e-9" variation="logic_transistor_var"/>
<nmos name="nch" chan_length="40e-9" min_width="140e-9" variation="logic_transistor_var"/>
</device_model>
<device_model name="io" type="transistor">
<lib type="academia" ref="M" path="${OPENFPGA_PATH}/openfpga_flow/tech/PTM_45nm/45nm.pm"/>
<design vdd="2.5" pn_ratio="3"/>
<pmos name="pch_25" chan_length="270e-9" min_width="320e-9" variation="io_transistor_var"/>
<nmos name="nch_25" chan_length="270e-9" min_width="320e-9" variation="io_transistor_var"/>
</device_model>
</device_library>
<variation_library>
<variation name="logic_transistor_var" abs_deviation="0.1" num_sigma="3"/>
<variation name="io_transistor_var" abs_deviation="0.1" num_sigma="3"/>
</variation_library>
</technology_library>
<circuit_library>
<circuit_model type="inv_buf" name="sky130_fd_sc_hd__inv_1" prefix="sky130_fd_sc_hd__inv_1" is_default="true">
<design_technology type="cmos" topology="inverter" size="1"/>
<device_technology device_model_name="logic"/>
<port type="input" prefix="in" lib_name="A" size="1"/>
<port type="output" prefix="out" lib_name="Y" size="1"/>
<delay_matrix type="rise" in_port="in" out_port="out">
10e-12
</delay_matrix>
<delay_matrix type="fall" in_port="in" out_port="out">
10e-12
</delay_matrix>
</circuit_model>
<circuit_model type="inv_buf" name="sky130_fd_sc_hd__buf_2" prefix="sky130_fd_sc_hd__buf_2" is_default="false">
<design_technology type="cmos" topology="buffer" size="1" num_level="2" f_per_stage="2"/>
<device_technology device_model_name="logic"/>
<port type="input" prefix="in" lib_name="A" size="1"/>
<port type="output" prefix="out" lib_name="X" size="1"/>
<delay_matrix type="rise" in_port="in" out_port="out">
10e-12
</delay_matrix>
<delay_matrix type="fall" in_port="in" out_port="out">
10e-12
</delay_matrix>
</circuit_model>
<circuit_model type="inv_buf" name="sky130_fd_sc_hd__buf_4" prefix="sky130_fd_sc_hd__buf_4" is_default="false">
<design_technology type="cmos" topology="buffer" size="1" num_level="2" f_per_stage="4"/>
<device_technology device_model_name="logic"/>
<port type="input" prefix="in" lib_name="A" size="1"/>
<port type="output" prefix="out" lib_name="X" size="1"/>
<delay_matrix type="rise" in_port="in" out_port="out">
10e-12
</delay_matrix>
<delay_matrix type="fall" in_port="in" out_port="out">
10e-12
</delay_matrix>
</circuit_model>
<circuit_model type="inv_buf" name="sky130_fd_sc_hd__inv_2" prefix="sky130_fd_sc_hd__inv_2" is_default="false">
<design_technology type="cmos" topology="buffer" size="1"/>
<device_technology device_model_name="logic"/>
<port type="input" prefix="in" lib_name="A" size="1"/>
<port type="output" prefix="out" lib_name="Y" size="1"/>
<delay_matrix type="rise" in_port="in" out_port="out">
10e-12
</delay_matrix>
<delay_matrix type="fall" in_port="in" out_port="out">
10e-12
</delay_matrix>
</circuit_model>
<circuit_model type="gate" name="sky130_fd_sc_hd__or2_1" prefix="sky130_fd_sc_hd__or2_1" is_default="true">
<design_technology type="cmos" topology="OR"/>
<device_technology device_model_name="logic"/>
<input_buffer exist="false"/>
<output_buffer exist="false"/>
<port type="input" prefix="a" lib_name="A" size="1"/>
<port type="input" prefix="b" lib_name="B" size="1"/>
<port type="output" prefix="out" lib_name="X" size="1"/>
<delay_matrix type="rise" in_port="a b" out_port="out">
10e-12 5e-12
</delay_matrix>
<delay_matrix type="fall" in_port="a b" out_port="out">
10e-12 5e-12
</delay_matrix>
</circuit_model>
<!-- Define a circuit model for the standard cell MUX2
OpenFPGA requires the following truth table for the MUX2
When the select signal sel is enabled, the first input, i.e., in0
will be propagated to the output, i.e., out
If your standard cell provider does not offer the exact truth table,
you can simply swap the inputs as shown in the example below
-->
<circuit_model type="gate" name="sky130_fd_sc_hd__mux2_1" prefix="sky130_fd_sc_hd__mux2_1">
<design_technology type="cmos" topology="MUX2"/>
<device_technology device_model_name="logic"/>
<input_buffer exist="false"/>
<output_buffer exist="false"/>
<port type="input" prefix="in0" lib_name="A1" size="1"/>
<port type="input" prefix="in1" lib_name="A0" size="1"/>
<port type="input" prefix="sel" lib_name="S" size="1"/>
<port type="output" prefix="out" lib_name="X" size="1"/>
</circuit_model>
<circuit_model type="chan_wire" name="chan_segment" prefix="track_seg" is_default="true">
<design_technology type="cmos"/>
<input_buffer exist="false"/>
<output_buffer exist="false"/>
<port type="input" prefix="in" size="1"/>
<port type="output" prefix="out" size="1"/>
<wire_param model_type="pi" R="101" C="22.5e-15" num_level="1"/>
<!-- model_type could be T, res_val and cap_val DON'T CARE -->
</circuit_model>
<circuit_model type="wire" name="direct_interc" prefix="direct_interc" is_default="true">
<design_technology type="cmos"/>
<input_buffer exist="false"/>
<output_buffer exist="false"/>
<port type="input" prefix="in" size="1"/>
<port type="output" prefix="out" size="1"/>
<wire_param model_type="pi" R="0" C="0" num_level="1"/>
<!-- model_type could be T, res_val cap_val should be defined -->
</circuit_model>
<circuit_model type="mux" name="mux_tree" prefix="mux_tree" is_default="true" dump_structural_verilog="true">
<design_technology type="cmos" structure="tree" add_const_input="true" const_input_val="1"/>
<input_buffer exist="false"/>
<output_buffer exist="false"/>
<pass_gate_logic circuit_model_name="sky130_fd_sc_hd__mux2_1"/>
<port type="input" prefix="in" size="1"/>
<port type="output" prefix="out" size="1"/>
<port type="sram" prefix="sram" size="1"/>
</circuit_model>
<circuit_model type="mux" name="mux_tree_tapbuf" prefix="mux_tree_tapbuf" dump_structural_verilog="true">
<design_technology type="cmos" structure="tree" add_const_input="true" const_input_val="1"/>
<input_buffer exist="false"/>
<output_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__buf_4"/>
<pass_gate_logic circuit_model_name="sky130_fd_sc_hd__mux2_1"/>
<port type="input" prefix="in" size="1"/>
<port type="output" prefix="out" size="1"/>
<port type="sram" prefix="sram" size="1"/>
</circuit_model>
<!--DFF subckt ports should be defined as <D> <Q> <CLK> <RESET> <SET> -->
<circuit_model type="ff" name="SDFFRQ" prefix="SDFFRQ" verilog_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/verilog/dff.v">
<design_technology type="cmos"/>
<input_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
<output_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
<port type="input" prefix="D" size="1"/>
<port type="input" prefix="DI" lib_name="SI" size="1"/>
<port type="input" prefix="Test_en" lib_name="SE" size="1" is_global="true" default_val="0"/>
<port type="input" prefix="reset" lib_name="RST" size="1" default_val="0"/>
<port type="output" prefix="Q" size="1"/>
<port type="clock" prefix="clk" lib_name="CK" size="1" default_val="0" />
</circuit_model>
<circuit_model type="lut" name="frac_lut4" prefix="frac_lut4" dump_structural_verilog="true">
<design_technology type="cmos" fracturable_lut="true"/>
<input_buffer exist="false"/>
<output_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__buf_2"/>
<lut_input_inverter exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
<lut_input_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__buf_2"/>
<lut_intermediate_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__buf_2" location_map="-1-"/>
<pass_gate_logic circuit_model_name="sky130_fd_sc_hd__mux2_1"/>
<port type="input" prefix="in" size="4" tri_state_map="---1" circuit_model_name="sky130_fd_sc_hd__or2_1"/>
<port type="output" prefix="lut2_out" size="2" lut_frac_level="2" lut_output_mask="2,3"/>
<port type="output" prefix="lut3_out" size="2" lut_frac_level="3" lut_output_mask="0,1"/>
<port type="output" prefix="lut4_out" size="1" lut_output_mask="0"/>
<port type="sram" prefix="sram" size="16"/>
<port type="sram" prefix="mode" size="1" mode_select="true" circuit_model_name="DFFRQ" default_val="1"/>
</circuit_model>
<!--Scan-chain DFF subckt ports should be defined as <D> <Q> <Qb> <CLK> <RESET> <SET> -->
<circuit_model type="ccff" name="DFFRQ" prefix="DFFRQ" verilog_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/verilog/dff.v">
<design_technology type="cmos"/>
<input_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
<output_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
<port type="input" prefix="D" size="1"/>
<port type="output" prefix="Q" size="1"/>
<port type="clock" prefix="prog_clk" lib_name="CK" size="1" is_global="true" default_val="0" is_prog="true"/>
<port type="input" prefix="pReset" lib_name="RST" size="1" is_global="true" default_val="0" is_prog="true" is_reset="true"/>
</circuit_model>
<circuit_model type="iopad" name="EMBEDDED_IO_ISOLN" prefix="EMBEDDED_IO_ISOLN" is_default="true" verilog_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/verilog/gpio.v">
<design_technology type="cmos"/>
<input_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
<output_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
<port type="input" prefix="SOC_IN" lib_name="SOC_IN" size="1" is_global="true" is_io="true" is_data_io="true"/>
<port type="output" prefix="SOC_OUT" lib_name="SOC_OUT" size="1" is_global="true" is_io="true" is_data_io="true"/>
<port type="output" prefix="SOC_DIR" lib_name="SOC_DIR" size="1" is_global="true" is_io="true"/>
<port type="input" prefix="IO_ISOL_N" lib_name="IO_ISOL_N" size="1" is_global="true" default_val="1"/>
<port type="output" prefix="inpad" lib_name="FPGA_IN" size="1"/>
<port type="input" prefix="outpad" lib_name="FPGA_OUT" size="1"/>
<port type="sram" prefix="en" lib_name="FPGA_DIR" size="1" mode_select="true" circuit_model_name="DFFRQ" default_val="1"/>
</circuit_model>
<circuit_model type="hard_logic" name="CARRY_MUX2" prefix="CARRY_MUX2" verilog_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/verilog/mux2.v">
<design_technology type="cmos"/>
<device_technology device_model_name="logic"/>
<input_buffer exist="false"/>
<output_buffer exist="false"/>
<port type="input" prefix="a" lib_name="A0" size="1"/>
<port type="input" prefix="b" lib_name="A1" size="1"/>
<port type="input" prefix="cin" lib_name="S" size="1"/>
<port type="output" prefix="cout" lib_name="Y" size="1"/>
</circuit_model>
<circuit_model type="hard_logic" name="mult_8x8" prefix="mult_8x8" is_default="true" spice_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/spice/mult_8x8.sp" verilog_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/verilog/mult_8x8.v">
<design_technology type="cmos"/>
<input_buffer exist="true" circuit_model_name="INVTX1"/>
<output_buffer exist="true" circuit_model_name="INVTX1"/>
<port type="input" prefix="A" lib_name="A" size="8"/>
<port type="input" prefix="B" lib_name="B" size="8"/>
<port type="output" prefix="Y" lib_name="Y" size="16"/>
</circuit_model>
</circuit_library>
<configuration_protocol>
<organization type="scan_chain" circuit_model_name="DFFRQ" num_regions="1"/>
</configuration_protocol>
<connection_block>
<switch name="ipin_cblock" circuit_model_name="mux_tree_tapbuf"/>
</connection_block>
<switch_block>
<switch name="L1_mux" circuit_model_name="mux_tree_tapbuf"/>
<switch name="L2_mux" circuit_model_name="mux_tree_tapbuf"/>
<switch name="L4_mux" circuit_model_name="mux_tree_tapbuf"/>
</switch_block>
<routing_segment>
<segment name="L1" circuit_model_name="chan_segment"/>
<segment name="L2" circuit_model_name="chan_segment"/>
<segment name="L4" circuit_model_name="chan_segment"/>
</routing_segment>
<direct_connection>
<direct name="carry_chain" circuit_model_name="direct_interc"/>
<direct name="shift_register" circuit_model_name="direct_interc"/>
<direct name="scan_chain" circuit_model_name="direct_interc" type="column" x_dir="positive" y_dir="positive"/>
</direct_connection>
<tile_annotations>
<global_port name="clk" is_clock="true" default_val="0">
<tile name="clb" port="clk" x="-1" y="-1"/>
</global_port>
<global_port name="Reset" is_reset="true" default_val="1">
<tile name="clb" port="reset" x="-1" y="-1"/>
</global_port>
</tile_annotations>
<pb_type_annotations>
<!-- physical pb_type binding in complex block IO -->
<pb_type name="io" physical_mode_name="physical" idle_mode_name="inpad"/>
<!-- IMPORTANT: must set unused I/Os to operating in INPUT mode !!! -->
<pb_type name="io[physical].iopad" circuit_model_name="EMBEDDED_IO_ISOLN" mode_bits="1"/>
<pb_type name="io[inpad].inpad" physical_pb_type_name="io[physical].iopad" mode_bits="1"/>
<pb_type name="io[outpad].outpad" physical_pb_type_name="io[physical].iopad" mode_bits="0"/>
<!-- End physical pb_type binding in complex block IO -->
<!-- physical pb_type binding in complex block CLB -->
<!-- physical mode will be the default mode if not specified -->
<pb_type name="clb.fle" physical_mode_name="physical"/>
<pb_type name="clb.fle[physical].fabric.frac_logic.frac_lut4" circuit_model_name="frac_lut4" mode_bits="0"/>
<pb_type name="clb.fle[physical].fabric.frac_logic.carry_follower" circuit_model_name="CARRY_MUX2"/>
<pb_type name="clb.fle[physical].fabric.ff" circuit_model_name="SDFFRQ"/>
<!-- Binding operating pb_type to physical pb_type -->
<!-- Binding operating pb_types in mode 'arithmetic' -->
<pb_type name="clb.fle[arithmetic].soft_adder.adder_lut4" physical_pb_type_name="clb.fle[physical].fabric.frac_logic.frac_lut4" mode_bits="1"/>
<pb_type name="clb.fle[arithmetic].soft_adder.ff" physical_pb_type_name="clb.fle[physical].fabric.ff"/>
<pb_type name="clb.fle[arithmetic].soft_adder.carry_follower" physical_pb_type_name="clb.fle[physical].fabric.frac_logic.carry_follower"/>
<!-- Binding operating pb_types in mode 'n2_lut3' -->
<pb_type name="clb.fle[n2_lut3].lut3inter.ble3.lut3" physical_pb_type_name="clb.fle[physical].fabric.frac_logic.frac_lut4" mode_bits="1" physical_pb_type_index_factor="0.5">
<!-- Binding the lut3 to the first 3 inputs of fracturable lut4 -->
<port name="in" physical_mode_port="in[0:2]"/>
<port name="out" physical_mode_port="lut3_out[0:0]" physical_mode_pin_rotate_offset="1"/>
</pb_type>
<pb_type name="clb.fle[n2_lut3].lut3inter.ble3.ff" physical_pb_type_name="clb.fle[physical].fabric.ff"/>
<!-- Binding operating pb_types in mode 'ble4' -->
<pb_type name="clb.fle[n1_lut4].ble4.lut4" physical_pb_type_name="clb.fle[physical].fabric.frac_logic.frac_lut4" mode_bits="0">
<!-- Binding the lut4 to the first 4 inputs of fracturable lut4 -->
<port name="in" physical_mode_port="in[0:3]"/>
<port name="out" physical_mode_port="lut4_out"/>
</pb_type>
<pb_type name="clb.fle[n1_lut4].ble4.ff" physical_pb_type_name="clb.fle[physical].fabric.ff" physical_pb_type_index_factor="2" physical_pb_type_index_offset="0"/>
<!-- Binding operating pb_types in mode 'shift_register' -->
<pb_type name="clb.fle[shift_register].shift_reg.ff" physical_pb_type_name="clb.fle[physical].fabric.ff"/>
<!-- End physical pb_type binding in complex block IO -->
<!-- physical pb_type binding in complex block dsp -->
<pb_type name="mult_8" physical_mode_name="mult_8x8" idle_mode_name="mult_8x8"/>
<!-- Bind the primitive pb_type in the physical mode to a circuit model -->
<pb_type name="mult_8[mult_8x8].mult_8x8_slice.mult_8x8" circuit_model_name="mult_8x8"/>
</pb_type_annotations>
</openfpga_architecture>

16
openfpga_flow/openfpga_cell_library/verilog/mult_8x8.v Normal file
View File

@ -0,0 +1,16 @@
//-----------------------------------------------------
// Design Name : mult_8x8
// File Name : mult_8x8.v
// Function : A 8-bit multiplier
// Coder : Xifan Tang
//-----------------------------------------------------
module mult_8x8 (
input [0:7] A,
input [0:7] B,
output [0:15] Y
);
assign Y = A * B;
endmodule

74
openfpga_flow/openfpga_shell_scripts/fix_heterogeneous_device_example_script.openfpga Normal file
View File

@ -0,0 +1,74 @@
# Run VPR for the 'and' design
#--write_rr_graph example_rr_graph.xml
vpr ${VPR_ARCH_FILE} ${VPR_TESTBENCH_BLIF} --clock_modeling route --device ${OPENFPGA_VPR_DEVICE_LAYOUT}
# Read OpenFPGA architecture definition
read_openfpga_arch -f ${OPENFPGA_ARCH_FILE}
# Read OpenFPGA simulation settings
read_openfpga_simulation_setting -f ${OPENFPGA_SIM_SETTING_FILE}
# Annotate the OpenFPGA architecture to VPR data base
# to debug use --verbose options
link_openfpga_arch --sort_gsb_chan_node_in_edges
# Check and correct any naming conflicts in the BLIF netlist
check_netlist_naming_conflict --fix --report ./netlist_renaming.xml
# Apply fix-up to clustering nets based on routing results
pb_pin_fixup --verbose
# Apply fix-up to Look-Up Table truth tables based on packing results
lut_truth_table_fixup
# Build the module graph
# - Enabled compression on routing architecture modules
# - Enable pin duplication on grid modules
build_fabric --compress_routing #--verbose
# Write the fabric hierarchy of module graph to a file
# This is used by hierarchical PnR flows
write_fabric_hierarchy --file ./fabric_hierarchy.txt
# Repack the netlist to physical pbs
# This must be done before bitstream generator and testbench generation
# Strongly recommend it is done after all the fix-up have been applied
repack #--verbose
# Build the bitstream
# - Output the fabric-independent bitstream to a file
build_architecture_bitstream --verbose --write_file fabric_independent_bitstream.xml
# Build fabric-dependent bitstream
build_fabric_bitstream --verbose
# Write fabric-dependent bitstream
write_fabric_bitstream --file fabric_bitstream.xml --format xml
# Write the Verilog netlist for FPGA fabric
# - Enable the use of explicit port mapping in Verilog netlist
write_fabric_verilog --file ./SRC --explicit_port_mapping --include_timing --print_user_defined_template --verbose
# Write the Verilog testbench for FPGA fabric
# - We suggest the use of same output directory as fabric Verilog netlists
# - Must specify the reference benchmark file if you want to output any testbenches
# - Enable top-level testbench which is a full verification including programming circuit and core logic of FPGA
# - Enable pre-configured top-level testbench which is a fast verification skipping programming phase
# - Simulation ini file is optional and is needed only when you need to interface different HDL simulators using openfpga flow-run scripts
write_verilog_testbench --file ./SRC --reference_benchmark_file_path ${REFERENCE_VERILOG_TESTBENCH} --print_top_testbench --print_preconfig_top_testbench --print_simulation_ini ./SimulationDeck/simulation_deck.ini --include_signal_init --support_icarus_simulator #--explicit_port_mapping
# Write the SDC files for PnR backend
# - Turn on every options here
write_pnr_sdc --file ./SDC
# Write SDC to disable timing for configure ports
write_sdc_disable_timing_configure_ports --file ./SDC/disable_configure_ports.sdc
# Write the SDC to run timing analysis for a mapped FPGA fabric
write_analysis_sdc --file ./SDC_analysis
# Finish and exit OpenFPGA
exit
# Note :
# To run verification at the end of the flow maintain source in ./SRC directory

12
openfpga_flow/openfpga_yosys_techlib/k4_frac_N8_tileable_reset_softadder_register_scan_chain_dsp8_nonLR_caravel_io_skywater130nm_cell_sim.v Normal file
View File

@ -0,0 +1,12 @@
//-----------------------------
// 8-bit multiplier
//-----------------------------
module mult_8(
input [0:7] A,
input [0:7] B,
output [0:15] Y
);
assign Y = A * B;
endmodule

20
openfpga_flow/openfpga_yosys_techlib/k4_frac_N8_tileable_reset_softadder_register_scan_chain_dsp8_nonLR_caravel_io_skywater130nm_dsp_map.v Normal file
View File

@ -0,0 +1,20 @@
//-----------------------------
// 8-bit multiplier
//-----------------------------
module mult_8x8 (
input [0:7] A,
input [0:7] B,
output [0:15] Y
);
parameter A_SIGNED = 0;
parameter B_SIGNED = 0;
parameter A_WIDTH = 0;
parameter B_WIDTH = 0;
parameter Y_WIDTH = 0;
mult_8 #() _TECHMAP_REPLACE_ (
.A (A),
.B (B),
.Y (Y) );
endmodule

3
openfpga_flow/regression_test_scripts/fpga_verilog_reg_test.sh
View File

@ -44,6 +44,9 @@ run-task fpga_verilog/bram/dpram16k --debug --show_thread_logs
echo -e "Testing Verilog generation with 16k block RAMs spanning two columns ";
run-task fpga_verilog/bram/wide_dpram16k --debug --show_thread_logs
echo -e "Testing Verilog generation with heterogeneous fabric using 8-bit single-mode multipliers ";
run-task fpga_verilog/dsp/single_mode_mult_8x8 --debug --show_thread_logs
echo -e "Testing Verilog generation with different I/O capacities on each side of an FPGA ";
run-task fpga_verilog/io/multi_io_capacity --debug --show_thread_logs

43
openfpga_flow/tasks/fpga_verilog/dsp/single_mode_mult_8x8/config/task.conf Normal file
View File

@ -0,0 +1,43 @@
# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
# Configuration file for running experiments
# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
# timeout_each_job : FPGA Task script splits fpga flow into multiple jobs
# Each job execute fpga_flow script on combination of architecture & benchmark
# timeout_each_job is timeout for each job
# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
[GENERAL]
run_engine=openfpga_shell
power_tech_file = ${PATH:OPENFPGA_PATH}/openfpga_flow/tech/PTM_45nm/45nm.xml
power_analysis = false
spice_output=false
verilog_output=true
timeout_each_job = 20*60
fpga_flow=yosys_vpr
[OpenFPGA_SHELL]
openfpga_shell_template=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_shell_scripts/fix_heterogeneous_device_example_script.openfpga
openfpga_arch_file=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_arch/k4_frac_N8_reset_softadder_register_scan_chain_dsp8_caravel_io_skywater130nm_fdhd_cc_openfpga.xml
openfpga_sim_setting_file=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_simulation_settings/fixed_sim_openfpga.xml
# Yosys script parameters
yosys_cell_sim_verilog=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_yosys_techlib/k4_frac_N8_tileable_reset_softadder_register_scan_chain_dsp8_nonLR_caravel_io_skywater130nm_cell_sim.v
yosys_dsp_map_verilog=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_yosys_techlib/k4_frac_N8_tileable_reset_softadder_register_scan_chain_dsp8_nonLR_caravel_io_skywater130nm_dsp_map.v
yosys_dsp_map_parameters=-D DSP_A_MAXWIDTH=8 -D DSP_B_MAXWIDTH=8 -D DSP_A_MINWIDTH=2 -D DSP_B_MINWIDTH=2 -D DSP_NAME=mult_8x8
# VPR parameter
openfpga_vpr_device_layout=3x2
[ARCHITECTURES]
arch0=${PATH:OPENFPGA_PATH}/openfpga_flow/vpr_arch/k4_frac_N8_tileable_reset_softadder_register_scan_chain_dsp8_nonLR_caravel_io_skywater130nm.xml
[BENCHMARKS]
bench0=${PATH:OPENFPGA_PATH}/openfpga_flow/benchmarks/micro_benchmark/mac_8/mac_8.v
[SYNTHESIS_PARAM]
bench_yosys_common=${PATH:OPENFPGA_PATH}/openfpga_flow/misc/ys_tmpl_yosys_vpr_dsp_flow.ys
bench_yosys_rewrite_common=${PATH:OPENFPGA_PATH}/openfpga_flow/misc/ys_tmpl_yosys_vpr_flow_with_rewrite.ys;${PATH:OPENFPGA_PATH}/openfpga_flow/misc/ys_tmpl_rewrite_flow.ys
bench0_top = mac_8
[SCRIPT_PARAM_MIN_ROUTE_CHAN_WIDTH]
end_flow_with_test=
vpr_fpga_verilog_formal_verification_top_netlist=

906
openfpga_flow/vpr_arch/k4_frac_N8_tileable_reset_softadder_register_scan_chain_dsp8_nonLR_caravel_io_skywater130nm.xml Normal file
View File

@ -0,0 +1,906 @@
<!--
Low-cost homogeneous FPGA Architecture.
- Skywater 130 nm technology
- General purpose logic block:
K = 4, N = 8, fracturable 4 LUTs (can operate as one 4-LUT or two 3-LUTs with all 3 inputs shared)
with optionally registered outputs
- Heterogeneous block
8-bit multiplier
- Routing architecture:
- 10% L = 1, fc_in = 0.15, Fc_out = 0.10
- 10% L = 2, fc_in = 0.15, Fc_out = 0.10
- 80% L = 4, fc_in = 0.15, Fc_out = 0.10
- 100 routing tracks per channel
Authors: Xifan Tang
-->
<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="mult_8">
<input_ports>
<port name="A" combinational_sink_ports="Y"/>
<port name="B" combinational_sink_ports="Y"/>
</input_ports>
<output_ports>
<port name="Y"/>
</output_ports>
</model>
<!-- A virtual model for I/O to be used in the physical mode of io block -->
<model name="io">
<input_ports>
<port name="outpad"/>
</input_ports>
<output_ports>
<port name="inpad"/>
</output_ports>
</model>
<model name="adder_lut4">
<input_ports>
<port name="in" combinational_sink_ports="lut2_out lut4_out"/>
</input_ports>
<output_ports>
<port name="lut2_out"/>
<port name="lut4_out"/>
</output_ports>
</model>
<model name="carry_follower">
<input_ports>
<port name="a" combinational_sink_ports="cout"/>
<port name="b" combinational_sink_ports="cout"/>
<port name="cin" combinational_sink_ports="cout"/>
</input_ports>
<output_ports>
<port name="cout"/>
</output_ports>
</model>
<model name="frac_lut4">
<input_ports>
<port name="in"/>
</input_ports>
<output_ports>
<port name="lut2_out"/>
<port name="lut3_out"/>
<port name="lut4_out"/>
</output_ports>
</model>
<model name="carry_follower_physical">
<input_ports>
<port name="a" combinational_sink_ports="cout"/>
<port name="b" combinational_sink_ports="cout"/>
<port name="cin" combinational_sink_ports="cout"/>
</input_ports>
<output_ports>
<port name="cout"/>
</output_ports>
</model>
<!-- A virtual model for scan-chain flip-flop to be used in the physical mode of FF -->
<model name="scff">
<input_ports>
<port name="D" clock="clk"/>
<port name="DI" clock="clk"/>
<port name="reset" clock="clk"/>
<port name="clk" is_clock="1"/>
</input_ports>
<output_ports>
<port name="Q" clock="clk"/>
</output_ports>
</model>
</models>
<tiles>
<!-- Do NOT add clock pins to I/O here!!! VPR does not build clock network in the way that OpenFPGA can support
If you need to register the I/O, define clocks in the circuit models
These clocks can be handled in back-end
-->
<!-- Top-side has 1 I/O per tile -->
<tile name="io_top" capacity="1" area="0">
<equivalent_sites>
<site pb_type="io"/>
</equivalent_sites>
<input name="outpad" num_pins="1"/>
<output name="inpad" num_pins="1"/>
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
<pinlocations pattern="custom">
<loc side="bottom">io_top.outpad io_top.inpad</loc>
</pinlocations>
</tile>
<!-- Right-side has 1 I/O per tile -->
<tile name="io_right" capacity="1" area="0">
<equivalent_sites>
<site pb_type="io"/>
</equivalent_sites>
<input name="outpad" num_pins="1"/>
<output name="inpad" num_pins="1"/>
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
<pinlocations pattern="custom">
<loc side="left">io_right.outpad io_right.inpad</loc>
</pinlocations>
</tile>
<!-- Bottom-side has 9 I/O per tile -->
<tile name="io_bottom" capacity="9" area="0">
<equivalent_sites>
<site pb_type="io"/>
</equivalent_sites>
<input name="outpad" num_pins="1"/>
<output name="inpad" num_pins="1"/>
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
<pinlocations pattern="custom">
<loc side="top">io_bottom.outpad io_bottom.inpad</loc>
</pinlocations>
</tile>
<!-- Left-side has 1 I/O per tile -->
<tile name="io_left" capacity="1" area="0">
<equivalent_sites>
<site pb_type="io"/>
</equivalent_sites>
<input name="outpad" num_pins="1"/>
<output name="inpad" num_pins="1"/>
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
<pinlocations pattern="custom">
<loc side="right">io_left.outpad io_left.inpad</loc>
</pinlocations>
</tile>
<!-- CLB has most pins on the top and right sides -->
<tile name="clb" area="53894">
<equivalent_sites>
<site pb_type="clb"/>
</equivalent_sites>
<input name="I0" num_pins="2" equivalent="full"/>
<input name="I0i" num_pins="2" equivalent="none"/>
<input name="I1" num_pins="2" equivalent="full"/>
<input name="I1i" num_pins="2" equivalent="none"/>
<input name="I2" num_pins="2" equivalent="full"/>
<input name="I2i" num_pins="2" equivalent="none"/>
<input name="I3" num_pins="2" equivalent="full"/>
<input name="I3i" num_pins="2" equivalent="none"/>
<input name="I4" num_pins="2" equivalent="full"/>
<input name="I4i" num_pins="2" equivalent="none"/>
<input name="I5" num_pins="2" equivalent="full"/>
<input name="I5i" num_pins="2" equivalent="none"/>
<input name="I6" num_pins="2" equivalent="full"/>
<input name="I6i" num_pins="2" equivalent="none"/>
<input name="I7" num_pins="2" equivalent="full"/>
<input name="I7i" num_pins="2" equivalent="none"/>
<input name="reg_in" num_pins="1"/>
<input name="sc_in" num_pins="1"/>
<input name="cin" num_pins="1"/>
<input name="reset" num_pins="1" is_non_clock_global="true"/>
<output name="O" num_pins="16" equivalent="none"/>
<output name="reg_out" num_pins="1"/>
<output name="sc_out" num_pins="1"/>
<output name="cout" num_pins="1"/>
<clock name="clk" num_pins="1"/>
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10">
<fc_override port_name="reg_in" fc_type="frac" fc_val="0"/>
<fc_override port_name="reg_out" fc_type="frac" fc_val="0"/>
<fc_override port_name="sc_in" fc_type="frac" fc_val="0"/>
<fc_override port_name="sc_out" fc_type="frac" fc_val="0"/>
<fc_override port_name="cin" fc_type="frac" fc_val="0"/>
<fc_override port_name="cout" fc_type="frac" fc_val="0"/>
<fc_override port_name="clk" fc_type="frac" fc_val="0"/>
<fc_override port_name="reset" fc_type="frac" fc_val="0"/>
</fc>
<!--pinlocations pattern="spread"/-->
<pinlocations pattern="custom">
<loc side="left">clb.clk clb.reset</loc>
<loc side="top">clb.reg_in clb.sc_in clb.cin clb.O[7:0] clb.I0 clb.I0i clb.I1 clb.I1i clb.I2 clb.I2i clb.I3 clb.I3i</loc>
<loc side="right">clb.O[15:8] clb.I4 clb.I4i clb.I5 clb.I5i clb.I6 clb.I6i clb.I7 clb.I7i</loc>
<loc side="bottom">clb.reg_out clb.sc_out clb.cout</loc>
</pinlocations>
</tile>
<tile name="mult_8" height="2" area="396000">
<equivalent_sites>
<site pb_type="mult_8" pin_mapping="direct"/>
</equivalent_sites>
<input name="a" num_pins="8"/>
<input name="b" num_pins="8"/>
<output name="out" num_pins="16"/>
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
<!-- Highly recommand to customize pin location when direct connection is used!!! -->
<!--pinlocations pattern="spread"/-->
<pinlocations pattern="custom">
<loc side="left"></loc>
<loc side="top"></loc>
<loc side="right">mult_8.a[0:5] mult_8.b[0:5] mult_8.out[0:10]</loc>
<loc side="bottom">mult_8.a[6:7] mult_8.b[6:7] mult_8.out[11:15]</loc>
</pinlocations>
</tile>
</tiles>
<!-- ODIN II specific config ends -->
<!-- Physical descriptions begin -->
<layout tileable="true">
<auto_layout aspect_ratio="1.0">
<!--Perimeter of 'io' blocks with 'EMPTY' blocks at corners-->
<row type="io_top" starty="H-1" priority="100"/>
<row type="io_bottom" starty="0" priority="100"/>
<col type="io_left" startx="0" priority="100"/>
<col type="io_right" startx="W-1" priority="100"/>
<corners type="EMPTY" priority="101"/>
<!--Fill with 'clb'-->
<fill type="clb" priority="10"/>
<!--Column of 'mult_8' with 'EMPTY' blocks wherever a 'mult_8' does not fit. Vertical offset by 1 for perimeter.-->
<col type="mult_8" startx="2" starty="1" repeatx="8" priority="20"/>
</auto_layout>
<fixed_layout name="3x2" width="5" height="4">
<!--Perimeter of 'io' blocks with 'EMPTY' blocks at corners-->
<row type="io_top" starty="H-1" priority="100"/>
<row type="io_bottom" starty="0" priority="100"/>
<col type="io_left" startx="0" priority="100"/>
<col type="io_right" startx="W-1" priority="100"/>
<corners type="EMPTY" priority="101"/>
<!--Fill with 'clb'-->
<fill type="clb" priority="10"/>
<!--Column of 'mult_8' with 'EMPTY' blocks wherever a 'mult_8' does not fit. Vertical offset by 1 for perimeter.-->
<col type="mult_8" startx="2" starty="1" repeatx="8" priority="20"/>
</fixed_layout>
<fixed_layout name="12x12" width="14" height="14">
<!--Perimeter of 'io' blocks with 'EMPTY' blocks at corners-->
<row type="io_top" starty="H-1" priority="100"/>
<row type="io_bottom" starty="0" priority="100"/>
<col type="io_left" startx="0" priority="100"/>
<col type="io_right" startx="W-1" priority="100"/>
<corners type="EMPTY" priority="101"/>
<!--Fill with 'clb'-->
<fill type="clb" priority="10"/>
<!--Column of 'mult_8' with 'EMPTY' blocks wherever a 'mult_8' does not fit. Vertical offset by 1 for perimeter.-->
<col type="mult_8" startx="2" starty="1" repeatx="8" priority="20"/>
</fixed_layout>
</layout>
<device>
<!-- VB & JL: Using Ian Kuon's transistor sizing and drive strength data for routing, at 40 nm. Ian used BPTM
models. We are modifying the delay values however, to include metal C and R, which allows more architecture
experimentation. We are also modifying the relative resistance of PMOS to be 1.8x that of NMOS
(vs. Ian's 3x) as 1.8x lines up with Jeff G's data from a 45 nm process (and is more typical of
45 nm in general). I'm upping the Rmin_nmos from Ian's just over 6k to nearly 9k, and dropping
RminW_pmos from 18k to 16k to hit this 1.8x ratio, while keeping the delays of buffers approximately
lined up with Stratix IV.
We are using Jeff G.'s capacitance data for 45 nm (in tech/ptm_45nm).
Jeff's tables list C in for transistors with widths in multiples of the minimum feature size (45 nm).
The minimum contactable transistor is 2.5 * 45 nm, so I need to multiply drive strength sizes in this file
by 2.5x when looking up in Jeff's tables.
The delay values are lined up with Stratix IV, which has an architecture similar to this
proposed FPGA, and which is also 40 nm
C_ipin_cblock: input capacitance of a track buffer, which VPR assumes is a single-stage
4x minimum drive strength buffer. -->
<sizing R_minW_nmos="8926" R_minW_pmos="16067"/>
<!-- The grid_logic_tile_area below will be used for all blocks that do not explicitly set their own (non-routing)
area; set to 0 since we explicitly set the area of all blocks currently in this architecture file.
-->
<area grid_logic_tile_area="0"/>
<chan_width_distr>
<x distr="uniform" peak="1.000000"/>
<y distr="uniform" peak="1.000000"/>
</chan_width_distr>
<switch_block type="wilton" fs="3" sub_type="subset" sub_fs="3"/>
<connection_block input_switch_name="ipin_cblock"/>
</device>
<switchlist>
<!-- VB: the mux_trans_size and buf_size data below is in minimum width transistor *areas*, assuming the purple
book area formula. This means the mux transistors are about 5x minimum drive strength.
We assume the first stage of the buffer is 3x min drive strength to be reasonable given the large
mux transistors, and this gives a reasonable stage ratio of a bit over 5x to the second stage. We assume
the n and p transistors in the first stage are equal-sized to lower the buffer trip point, since it's fed
by a pass transistor mux. We can then reverse engineer the buffer second stage to hit the specified
buf_size (really buffer area) - 16.2x minimum drive nmos and 1.8*16.2 = 29.2x minimum drive.
I then took the data from Jeff G.'s PTM modeling of 45 nm to get the Cin (gate of first stage) and Cout
(diff of second stage) listed below. Jeff's models are in tech/ptm_45nm, and are in min feature multiples.
The minimum contactable transistor is 2.5 * 45 nm, so I need to multiply the drive strength sizes above by
2.5x when looking up in Jeff's tables.
Finally, we choose a switch delay (58 ps) that leads to length 4 wires having a delay equal to that of SIV of 126 ps.
This also leads to the switch being 46% of the total wire delay, which is reasonable. -->
<switch type="mux" name="L1_mux" R="551" Cin=".77e-15" Cout="4e-15" Tdel="58e-12" mux_trans_size="2.630740" buf_size="27.645901"/>
<switch type="mux" name="L2_mux" R="551" Cin=".77e-15" Cout="4e-15" Tdel="58e-12" mux_trans_size="2.630740" buf_size="27.645901"/>
<switch type="mux" name="L4_mux" R="551" Cin=".77e-15" Cout="4e-15" Tdel="58e-12" mux_trans_size="2.630740" buf_size="27.645901"/>
<!--switch ipin_cblock resistance set to yeild for 4x minimum drive strength buffer-->
<switch type="mux" name="ipin_cblock" R="2231.5" Cout="0." Cin="1.47e-15" Tdel="7.247000e-11" mux_trans_size="1.222260" buf_size="auto"/>
</switchlist>
<segmentlist>
<!--- VB & JL: using ITRS metal stack data, 96 nm half pitch wires, which are intermediate metal width/space.
With the 96 nm half pitch, such wires would take 60 um of height, vs. a 90 nm high (approximated as square) Stratix IV tile so this seems
reasonable. Using a tile length of 90 nm, corresponding to the length of a Stratix IV tile if it were square. -->
<!-- GIVE a specific name for the segment! OpenFPGA appreciate that! -->
<segment name="L1" freq="0.10" length="1" type="unidir" Rmetal="101" Cmetal="22.5e-15">
<mux name="L1_mux"/>
<sb type="pattern">1 1</sb>
<cb type="pattern">1</cb>
</segment>
<segment name="L2" freq="0.10" length="2" type="unidir" Rmetal="101" Cmetal="22.5e-15">
<mux name="L2_mux"/>
<sb type="pattern">1 1 1</sb>
<cb type="pattern">1 1</cb>
</segment>
<segment name="L4" freq="0.80" length="4" type="unidir" Rmetal="101" Cmetal="22.5e-15">
<mux name="L4_mux"/>
<sb type="pattern">1 1 1 1 1</sb>
<cb type="pattern">1 1 1 1</cb>
</segment>
</segmentlist>
<directlist>
<direct name="carry_chain" from_pin="clb.cout" to_pin="clb.cin" x_offset="0" y_offset="-1" z_offset="0"/>
<direct name="shift_register" from_pin="clb.reg_out" to_pin="clb.reg_in" x_offset="0" y_offset="-1" z_offset="0"/>
<direct name="scan_chain" from_pin="clb.sc_out" to_pin="clb.sc_in" x_offset="0" y_offset="-1" z_offset="0"/>
</directlist>
<complexblocklist>
<!-- Define input pads begin -->
<pb_type name="io">
<input name="outpad" num_pins="1"/>
<output name="inpad" num_pins="1"/>
<!-- Do NOT add clock pins to I/O here!!! VPR does not build clock network in the way that OpenFPGA can support
If you need to register the I/O, define clocks in the circuit models
These clocks can be handled in back-end
-->
<!-- A mode denotes the physical implementation of an I/O
This mode will be not packable but is mainly used for fabric verilog generation
-->
<mode name="physical" disable_packing="true">
<pb_type name="iopad" blif_model=".subckt io" num_pb="1">
<input name="outpad" num_pins="1"/>
<output name="inpad" num_pins="1"/>
</pb_type>
<interconnect>
<direct name="outpad" input="io.outpad" output="iopad.outpad">
<delay_constant max="1.394e-11" in_port="io.outpad" out_port="iopad.outpad"/>
</direct>
<direct name="inpad" input="iopad.inpad" output="io.inpad">
<delay_constant max="4.243e-11" in_port="iopad.inpad" out_port="io.inpad"/>
</direct>
</interconnect>
</mode>
<!-- 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>
<power method="ignore"/>
</pb_type>
<!-- Define I/O pads ends -->
<!-- Define general purpose logic block (CLB) begin -->
<!-- -Due to the absence of local routing,
the 4 inputs of fracturable LUT4 are no longer equivalent,
because the 4th input can not be switched when the dual-LUT3 modes are used.
So pin equivalence should be applied to the first 3 inputs only
-->
<pb_type name="clb">
<input name="I0" num_pins="2" equivalent="full"/>
<input name="I0i" num_pins="2" equivalent="none"/>
<input name="I1" num_pins="2" equivalent="full"/>
<input name="I1i" num_pins="2" equivalent="none"/>
<input name="I2" num_pins="2" equivalent="full"/>
<input name="I2i" num_pins="2" equivalent="none"/>
<input name="I3" num_pins="2" equivalent="full"/>
<input name="I3i" num_pins="2" equivalent="none"/>
<input name="I4" num_pins="2" equivalent="full"/>
<input name="I4i" num_pins="2" equivalent="none"/>
<input name="I5" num_pins="2" equivalent="full"/>
<input name="I5i" num_pins="2" equivalent="none"/>
<input name="I6" num_pins="2" equivalent="full"/>
<input name="I6i" num_pins="2" equivalent="none"/>
<input name="I7" num_pins="2" equivalent="full"/>
<input name="I7i" num_pins="2" equivalent="none"/>
<input name="reg_in" num_pins="1"/>
<input name="sc_in" num_pins="1"/>
<input name="cin" num_pins="1"/>
<input name="reset" num_pins="1" is_non_clock_global="true"/>
<output name="O" num_pins="16" equivalent="none"/>
<output name="reg_out" num_pins="1"/>
<output name="sc_out" num_pins="1"/>
<output name="cout" num_pins="1"/>
<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 shared inputs.
The outputs of the fracturable logic element can be optionally registered
-->
<pb_type name="fle" num_pb="8">
<input name="in" num_pins="4"/>
<input name="reg_in" num_pins="1"/>
<input name="sc_in" num_pins="1"/>
<input name="cin" num_pins="1"/>
<input name="reset" num_pins="1"/>
<output name="out" num_pins="2"/>
<output name="reg_out" num_pins="1"/>
<output name="sc_out" num_pins="1"/>
<output name="cout" num_pins="1"/>
<clock name="clk" num_pins="1"/>
<!-- Physical mode definition begin (physical implementation of the fle) -->
<mode name="physical" disable_packing="true">
<pb_type name="fabric" num_pb="1">
<input name="in" num_pins="4"/>
<input name="reg_in" num_pins="1"/>
<input name="sc_in" num_pins="1"/>
<input name="cin" num_pins="1"/>
<input name="reset" num_pins="1"/>
<output name="out" num_pins="2"/>
<output name="reg_out" num_pins="1"/>
<output name="sc_out" num_pins="1"/>
<output name="cout" num_pins="1"/>
<clock name="clk" num_pins="1"/>
<pb_type name="frac_logic" num_pb="1">
<input name="in" num_pins="4"/>
<input name="cin" num_pins="1"/>
<output name="out" num_pins="2"/>
<output name="cout" num_pins="1"/>
<!-- Define LUT -->
<pb_type name="frac_lut4" blif_model=".subckt frac_lut4" num_pb="1">
<input name="in" num_pins="4"/>
<output name="lut2_out" num_pins="2"/>
<output name="lut3_out" num_pins="2"/>
<output name="lut4_out" num_pins="1"/>
</pb_type>
<pb_type name="carry_follower" blif_model=".subckt carry_follower_physical" 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"/>
<delay_constant max="0.3e-9" in_port="carry_follower.a" out_port="carry_follower.cout"/>
<delay_constant max="0.3e-9" in_port="carry_follower.b" out_port="carry_follower.cout"/>
<delay_constant max="0.3e-9" in_port="carry_follower.cin" out_port="carry_follower.cout"/>
</pb_type>
<interconnect>
<direct name="direct1" input="frac_logic.in[0:1]" output="frac_lut4.in[0:1]"/>
<direct name="direct2" input="frac_logic.in[3:3]" output="frac_lut4.in[3:3]"/>
<direct name="direct3" input="frac_logic.cin" output="carry_follower.b"/>
<direct name="direct4" input="frac_lut4.lut2_out[1:1]" output="carry_follower.a"/>
<direct name="direct5" input="frac_lut4.lut2_out[0:0]" output="carry_follower.cin"/>
<direct name="direct6" input="carry_follower.cout" output="frac_logic.cout"/>
<direct name="direct7" input="frac_lut4.lut3_out[1]" output="frac_logic.out[1]"/>
<!-- Xifan Tang: I use out[0] because the output of lut6 in lut6 mode is wired to the out[0] -->
<mux name="mux1" input="frac_lut4.lut4_out frac_lut4.lut3_out[0]" output="frac_logic.out[0]"/>
<mux name="mux2" input="frac_logic.cin frac_logic.in[2:2]" output="frac_lut4.in[2:2]"/>
</interconnect>
</pb_type>
<!-- Define flip-flop with scan-chain capability, DI is the scan-chain data input -->
<pb_type name="ff" blif_model=".subckt scff" num_pb="2">
<input name="D" num_pins="1"/>
<input name="DI" num_pins="1"/>
<input name="reset" num_pins="1"/>
<output name="Q" num_pins="1"/>
<clock name="clk" num_pins="1"/>
<T_setup value="66e-12" port="ff.D" clock="clk"/>
<T_setup value="66e-12" port="ff.DI" clock="clk"/>
<T_setup value="66e-12" port="ff.reset" clock="clk"/>
<T_clock_to_Q max="124e-12" port="ff.Q" clock="clk"/>
</pb_type>
<interconnect>
<direct name="direct1" input="fabric.in" output="frac_logic.in"/>
<direct name="direct2" input="fabric.cin" output="frac_logic.cin"/>
<direct name="direct3" input="fabric.sc_in" output="ff[0].DI"/>
<direct name="direct4" input="ff[0].Q" output="ff[1].DI"/>
<direct name="direct5" input="ff[1].Q" output="fabric.sc_out"/>
<direct name="direct6" input="ff[1].Q" output="fabric.reg_out"/>
<direct name="direct7" input="frac_logic.cout" output="fabric.cout"/>
<complete name="complete1" input="fabric.clk" output="ff[1:0].clk"/>
<complete name="complete2" input="fabric.reset" output="ff[1:0].reset"/>
<mux name="mux1" input="frac_logic.out[0:0] fabric.reg_in" output="ff[0:0].D">
<delay_constant max="25e-12" in_port="frac_logic.out[0:0]" out_port="ff[0:0].D"/>
<delay_constant max="45e-12" in_port="fabric.reg_in" out_port="ff[0:0].D"/>
</mux>
<mux name="mux2" input="frac_logic.out[1:1] ff[0:0].Q" output="ff[1:1].D">
<delay_constant max="25e-12" in_port="frac_logic.out[1:1]" out_port="ff[1:1].D"/>
<delay_constant max="45e-12" in_port="ff[0:0].Q" out_port="ff[1:1].D"/>
</mux>
<mux name="mux3" input="ff[0].Q frac_logic.out[0]" output="fabric.out[0]">
<!-- LUT to output is faster than FF to output on a Stratix IV -->
<delay_constant max="25e-12" in_port="frac_logic.out[0]" out_port="fabric.out[0]"/>
<delay_constant max="45e-12" in_port="ff[0].Q" out_port="fabric.out[0]"/>
</mux>
<mux name="mux4" input="ff[1].Q frac_logic.out[1]" output="fabric.out[1]">
<!-- LUT to output is faster than FF to output on a Stratix IV -->
<delay_constant max="25e-12" in_port="frac_logic.out[1]" out_port="fabric.out[1]"/>
<delay_constant max="45e-12" in_port="ff[1].Q" out_port="fabric.out[1]"/>
</mux>
</interconnect>
</pb_type>
<interconnect>
<direct name="direct1" input="fle.in" output="fabric.in"/>
<direct name="direct2" input="fle.reg_in" output="fabric.reg_in"/>
<direct name="direct3" input="fle.sc_in" output="fabric.sc_in"/>
<direct name="direct4" input="fle.cin" output="fabric.cin"/>
<direct name="direct5" input="fabric.out" output="fle.out"/>
<direct name="direct6" input="fabric.reg_out" output="fle.reg_out"/>
<direct name="direct7" input="fabric.sc_out" output="fle.sc_out"/>
<direct name="direct8" input="fabric.cout" output="fle.cout"/>
<direct name="direct9" input="fle.clk" output="fabric.clk"/>
<direct name="direct10" input="fle.reset" output="fabric.reset"/>
</interconnect>
</mode>
<!-- Physical mode definition end (physical implementation of the fle) -->
<!-- Arithmetic mode definition begin -->
<mode name="arithmetic">
<pb_type name="soft_adder" num_pb="1">
<input name="in" num_pins="4"/>
<input name="cin" num_pins="1"/>
<output name="sumout" num_pins="1"/>
<output name="cout" num_pins="1"/>
<!-- Define special LUT marco to be used as adder -->
<pb_type name="adder_lut4" blif_model=".subckt adder_lut4" num_pb="1">
<input name="in" num_pins="4"/>
<output name="lut2_out" num_pins="2"/>
<output name="lut4_out" num_pins="1"/>
<delay_constant max="0.3e-9" in_port="adder_lut4.in" out_port="adder_lut4.lut2_out"/>
<delay_constant max="0.3e-9" in_port="adder_lut4.in" out_port="adder_lut4.lut4_out"/>
</pb_type>
<pb_type name="carry_follower" blif_model=".subckt carry_follower" 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"/>
<delay_constant max="0.3e-9" in_port="carry_follower.a" out_port="carry_follower.cout"/>
<delay_constant max="0.3e-9" in_port="carry_follower.b" out_port="carry_follower.cout"/>
<delay_constant max="0.3e-9" in_port="carry_follower.cin" out_port="carry_follower.cout"/>
</pb_type>
<interconnect>
<direct name="direct1" input="soft_adder.in[0:1]" output="adder_lut4.in[0:1]"/>
<direct name="direct2" input="soft_adder.in[3:3]" output="adder_lut4.in[3:3]"/>
<direct name="direct3" input="soft_adder.cin" output="carry_follower.b">
<!-- Pack pattern to build an adder chain connection considered by packer -->
<pack_pattern name="chain" in_port="soft_adder.cin" out_port="carry_follower.b"/>
</direct>
<direct name="direct4" input="adder_lut4.lut2_out[1:1]" output="carry_follower.a">
<!-- Pack pattern to pair adder_lut4 and carry_follower into a molecule
considered by packer -->
<pack_pattern name="lut_follower" in_port="adder_lut4.lut2_out[1:1]" out_port="carry_follower.a"/>
</direct>
<direct name="direct5" input="adder_lut4.lut2_out[0:0]" output="carry_follower.cin">
</direct>
<direct name="direct6" input="carry_follower.cout" output="soft_adder.cout">
<!-- Pack pattern to build an adder chain connection considered by packer -->
<pack_pattern name="chain" in_port="carry_follower.cout" out_port="soft_adder.cout"/>
</direct>
<direct name="direct7" input="adder_lut4.lut4_out" output="soft_adder.sumout[0:0]">
</direct>
<mux name="mux1" input="soft_adder.cin soft_adder.in[2:2]" output="adder_lut4.in[2:2]">
</mux>
</interconnect>
</pb_type>
<interconnect>
<direct name="direct1" input="fle.in" output="soft_adder.in"/>
<direct name="direct2" input="fle.cin" output="soft_adder.cin">
<!-- Pack pattern to build an adder chain connection considered by packer -->
<pack_pattern name="chain" in_port="fle.cin" out_port="soft_adder.cin"/>
</direct>
<direct name="direct3" input="soft_adder.sumout" output="fle.out[0:0]"/>
<direct name="direct4" input="soft_adder.cout" output="fle.cout">
<!-- Pack pattern to build an adder chain connection considered by packer -->
<pack_pattern name="chain" in_port="soft_adder.cout" out_port="fle.cout"/>
</direct>
</interconnect>
</mode>
<!-- Arithmetic mode definition end -->
<!-- Dual 3-LUT mode definition begin -->
<mode name="n2_lut3">
<pb_type name="lut3inter" num_pb="1">
<input name="in" num_pins="3"/>
<output name="out" num_pins="2"/>
<clock name="clk" num_pins="1"/>
<pb_type name="ble3" num_pb="2">
<input name="in" num_pins="3"/>
<output name="out" num_pins="1"/>
<clock name="clk" num_pins="1"/>
<!-- Define the LUT -->
<pb_type name="lut3" blif_model=".names" num_pb="1" 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 Stratix IV LUT but because VPR cannot do LUT rebalancing,
we instead take the average of these numbers to get more stable results
82e-12
173e-12
261e-12
263e-12
398e-12
-->
<delay_matrix type="max" in_port="lut3.in" out_port="lut3.out">
235e-12
235e-12
235e-12
</delay_matrix>
</pb_type>
<!-- Define the flip-flop -->
<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="66e-12" port="ff.D" clock="clk"/>
<T_clock_to_Q max="124e-12" port="ff.Q" clock="clk"/>
</pb_type>
<interconnect>
<direct name="direct1" input="ble3.in[2:0]" output="lut3[0:0].in[2:0]"/>
<direct name="direct2" input="lut3[0:0].out" output="ff[0:0].D">
<!-- Advanced user option that tells CAD tool to find LUT+FF pairs in netlist -->
<pack_pattern name="ble3" in_port="lut3[0:0].out" out_port="ff[0:0].D"/>
</direct>
<direct name="direct3" input="ble3.clk" output="ff[0:0].clk"/>
<mux name="mux1" input="ff[0:0].Q lut3.out[0:0]" output="ble3.out[0:0]">
<!-- LUT to output is faster than FF to output on a Stratix IV -->
<delay_constant max="25e-12" in_port="lut3.out[0:0]" out_port="ble3.out[0:0]"/>
<delay_constant max="45e-12" in_port="ff[0:0].Q" out_port="ble3.out[0:0]"/>
</mux>
</interconnect>
</pb_type>
<interconnect>
<direct name="direct1" input="lut3inter.in" output="ble3[0:0].in"/>
<direct name="direct2" input="lut3inter.in" output="ble3[1:1].in"/>
<direct name="direct3" input="ble3[1:0].out" output="lut3inter.out"/>
<complete name="complete1" input="lut3inter.clk" output="ble3[1:0].clk"/>
</interconnect>
</pb_type>
<interconnect>
<direct name="direct1" input="fle.in[2:0]" output="lut3inter.in"/>
<direct name="direct2" input="lut3inter.out" output="fle.out"/>
<direct name="direct3" input="fle.clk" output="lut3inter.clk"/>
</interconnect>
</mode>
<!-- Dual 3-LUT mode definition end -->
<!-- 4-LUT mode definition begin -->
<mode name="n1_lut4">
<!-- Define 4-LUT mode -->
<pb_type name="ble4" num_pb="1">
<input name="in" num_pins="4"/>
<output name="out" num_pins="1"/>
<clock name="clk" num_pins="1"/>
<!-- Define LUT -->
<pb_type name="lut4" blif_model=".names" num_pb="1" class="lut">
<input name="in" num_pins="4" 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 Stratix IV LUT but because VPR cannot do LUT rebalancing,
we instead take the average of these numbers to get more stable results
82e-12
173e-12
261e-12
263e-12
398e-12
397e-12
-->
<delay_matrix type="max" in_port="lut4.in" out_port="lut4.out">
261e-12
261e-12
261e-12
261e-12
</delay_matrix>
</pb_type>
<!-- Define flip-flop -->
<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="66e-12" port="ff.D" clock="clk"/>
<T_clock_to_Q max="124e-12" port="ff.Q" clock="clk"/>
</pb_type>
<interconnect>
<direct name="direct1" input="ble4.in" output="lut4[0:0].in"/>
<direct name="direct2" input="lut4.out" output="ff.D">
<!-- Advanced user option that tells CAD tool to find LUT+FF pairs in netlist -->
<pack_pattern name="ble4" in_port="lut4.out" out_port="ff.D"/>
</direct>
<direct name="direct3" input="ble4.clk" output="ff.clk"/>
<mux name="mux1" input="ff.Q lut4.out" output="ble4.out">
<!-- LUT to output is faster than FF to output on a Stratix IV -->
<delay_constant max="25e-12" in_port="lut4.out" out_port="ble4.out"/>
<delay_constant max="45e-12" in_port="ff.Q" out_port="ble4.out"/>
</mux>
</interconnect>
</pb_type>
<interconnect>
<direct name="direct1" input="fle.in" output="ble4.in"/>
<direct name="direct2" input="ble4.out" output="fle.out[0:0]"/>
<direct name="direct3" input="fle.clk" output="ble4.clk"/>
</interconnect>
</mode>
<!-- 4-LUT mode definition end -->
<!-- Define shift register begin -->
<mode name="shift_register">
<pb_type name="shift_reg" num_pb="1">
<input name="reg_in" num_pins="1"/>
<output name="ff_out" num_pins="2"/>
<output name="reg_out" num_pins="1"/>
<clock name="clk" num_pins="1"/>
<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="66e-12" port="ff.D" clock="clk"/>
<T_clock_to_Q max="124e-12" port="ff.Q" clock="clk"/>
</pb_type>
<interconnect>
<direct name="direct1" input="shift_reg.reg_in" output="ff[0].D"/>
<direct name="direct2" input="ff[0].Q" output="ff[1].D"/>
<direct name="direct3" input="ff[1].Q" output="shift_reg.reg_out"/>
<direct name="direct4" input="ff[0].Q" output="shift_reg.ff_out[0:0]"/>
<direct name="direct5" input="ff[1].Q" output="shift_reg.ff_out[1:1]"/>
<complete name="complete1" input="shift_reg.clk" output="ff.clk"/>
</interconnect>
</pb_type>
<interconnect>
<direct name="direct1" input="fle.reg_in" output="shift_reg.reg_in"/>
<direct name="direct2" input="shift_reg.reg_out" output="fle.reg_out"/>
<direct name="direct3" input="shift_reg.ff_out" output="fle.out"/>
<direct name="direct4" input="fle.clk" output="shift_reg.clk"/>
</interconnect>
</mode>
<!-- Define shift register end -->
</pb_type>
<interconnect>
<!-- We use direct connections to reduce the area to the most
The global local routing is going to compensate the loss in routability
-->
<!-- FIXME: The implicit port definition results in I0[0] connected to
in[2]. Such twisted connection is not expected.
I[0] should be connected to in[0]
-->
<direct name="direct_fle0" input="clb.I0[0:1]" output="fle[0:0].in[0:1]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle0i" input="clb.I0i[0:1]" output="fle[0:0].in[2:3]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle1" input="clb.I1[0:1]" output="fle[1:1].in[0:1]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle1i" input="clb.I1i[0:1]" output="fle[1:1].in[2:3]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle2" input="clb.I2[0:1]" output="fle[2:2].in[0:1]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle2i" input="clb.I2i[0:1]" output="fle[2:2].in[2:3]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle3" input="clb.I3[0:1]" output="fle[3:3].in[0:1]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle3i" input="clb.I3i[0:1]" output="fle[3:3].in[2:3]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle4" input="clb.I4[0:1]" output="fle[4:4].in[0:1]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle4i" input="clb.I4i[0:1]" output="fle[4:4].in[2:3]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle5" input="clb.I5[0:1]" output="fle[5:5].in[0:1]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle5i" input="clb.I5i[0:1]" output="fle[5:5].in[2:3]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle6" input="clb.I6[0:1]" output="fle[6:6].in[0:1]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle6i" input="clb.I6i[0:1]" output="fle[6:6].in[2:3]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle7" input="clb.I7[0:1]" output="fle[7:7].in[0:1]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<direct name="direct_fle7i" input="clb.I7i[0:1]" output="fle[7:7].in[2:3]">
<!-- TODO: Timing should be backannotated from post-PnR results -->
</direct>
<complete name="clks" input="clb.clk" output="fle[7:0].clk">
</complete>
<complete name="resets" input="clb.reset" output="fle[7:0].reset">
</complete>
<!-- 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="clbouts1" input="fle[3:0].out[0:1]" output="clb.O[7:0]"/>
<direct name="clbouts2" input="fle[7:4].out[0:1]" output="clb.O[15:8]"/>
<!-- Shift register chain links -->
<direct name="shift_register_in" input="clb.reg_in" output="fle[0:0].reg_in">
<!-- Put all inter-block carry chain delay on this one edge -->
<delay_constant max="0.16e-9" in_port="clb.reg_in" out_port="fle[0:0].reg_in"/>
<!--pack_pattern name="chain" in_port="clb.reg_in" out_port="fle[0:0].reg_in"/-->
</direct>
<direct name="shift_register_out" input="fle[7:7].reg_out" output="clb.reg_out">
<!--pack_pattern name="chain" in_port="fle[7:7].reg_out" out_port="clb.reg_out"/-->
</direct>
<direct name="shift_register_link" input="fle[6:0].reg_out" output="fle[7:1].reg_in">
<!--pack_pattern name="chain" in_port="fle[6:0].reg_out" out_port="fle[7:1].reg_in"/-->
</direct>
<!-- Scan chain links -->
<direct name="scan_chain_in" input="clb.sc_in" output="fle[0:0].sc_in">
<!-- Put all inter-block carry chain delay on this one edge -->
<delay_constant max="0.16e-9" in_port="clb.sc_in" out_port="fle[0:0].sc_in"/>
</direct>
<direct name="scan_chain_out" input="fle[7:7].sc_out" output="clb.sc_out">
</direct>
<direct name="scan_chain_link" input="fle[6:0].sc_out" output="fle[7:1].sc_in">
</direct>
<!-- Carry chain links -->
<direct name="carry_chain_in" input="clb.cin" output="fle[0:0].cin">
<!-- Put all inter-block carry chain delay on this one edge -->
<pack_pattern name="chain" in_port="clb.cin" out_port="fle[0:0].cin"/>
<delay_constant max="0.16e-9" in_port="clb.cin" out_port="fle[0:0].cin"/>
</direct>
<direct name="carry_chain_out" input="fle[7:7].cout" output="clb.cout">
<pack_pattern name="chain" in_port="fle[7:7].cout" out_port="clb.cout"/>
</direct>
<direct name="carry_chain_link" input="fle[6:0].cout" output="fle[7:1].cin">
<pack_pattern name="chain" in_port="fle[6:0].cout" out_port="fle[7:1].cin"/>
</direct>
</interconnect>
<!-- 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 -->
<!-- Place this general purpose logic block in any unspecified column -->
</pb_type>
<!-- Define general purpose logic block (CLB) ends -->
<!-- Define fracturable multiplier begin -->
<pb_type name="mult_8">
<input name="a" num_pins="8"/>
<input name="b" num_pins="8"/>
<output name="out" num_pins="16"/>
<mode name="mult_8x8">
<pb_type name="mult_8x8_slice" num_pb="1">
<input name="A_cfg" num_pins="8"/>
<input name="B_cfg" num_pins="8"/>
<output name="OUT_cfg" num_pins="16"/>
<pb_type name="mult_8x8" blif_model=".subckt mult_8" num_pb="1">
<input name="A" num_pins="8"/>
<input name="B" num_pins="8"/>
<output name="Y" num_pins="16"/>
<delay_constant max="1.523e-9" min="0.776e-9" in_port="mult_8x8.A" out_port="mult_8x8.Y"/>
<delay_constant max="1.523e-9" min="0.776e-9" in_port="mult_8x8.B" out_port="mult_8x8.Y"/>
</pb_type>
<interconnect>
<direct name="a2a" input="mult_8x8_slice.A_cfg" output="mult_8x8.A">
</direct>
<direct name="b2b" input="mult_8x8_slice.B_cfg" output="mult_8x8.B">
</direct>
<direct name="out2out" input="mult_8x8.Y" output="mult_8x8_slice.OUT_cfg">
</direct>
</interconnect>
<power method="pin-toggle">
<port name="A_cfg" energy_per_toggle="2.13e-12"/>
<port name="B_cfg" energy_per_toggle="2.13e-12"/>
<static_power power_per_instance="0.0"/>
</power>
</pb_type>
<interconnect>
<direct name="a2a" input="mult_8.a" output="mult_8x8_slice.A_cfg">
<delay_constant max="134e-12" min="74e-12" in_port="mult_8.a" out_port="mult_8x8_slice.A_cfg"/>
</direct>
<direct name="b2b" input="mult_8.b" output="mult_8x8_slice.B_cfg">
<delay_constant max="134e-12" min="74e-12" in_port="mult_8.b" out_port="mult_8x8_slice.B_cfg"/>
</direct>
<direct name="out2out" input="mult_8x8_slice.OUT_cfg" output="mult_8.out">
<delay_constant max="1.93e-9" min="74e-12" in_port="mult_8x8_slice.OUT_cfg" out_port="mult_8.out"/>
</direct>
</interconnect>
</mode>
<!-- Place this multiplier block every 8 columns from (and including) the sixth column -->
<power method="sum-of-children"/>
</pb_type>
<!-- Define fracturable multiplier end -->
</complexblocklist>
</architecture>