Hello people.
Just to inform you that PacoBlaze 1.3b is out. The are no changes to the KCPSMx cores, KCAsm has been enhanced and now supports all the PicoBlaze architectures. Work will continue to merge BlockRAM generation across target platforms.
Thanks for all who have submitted suggestions and bug reports.
Cheers!
-- PabloBleyerKocik /"Computers are like Old Testament gods; pbleyer2004 / lots of rules and no mercy." @embedded.cl / -- Joseph Campbell
He talked about this earlier in another thread.
-- Rick "rickman" Collins rick.collins@XYarius.com Ignore the reply address. To email me use the above address with the XY removed. Arius - A Signal Processing Solutions Company Specializing in DSP and FPGA design URL http://www.arius.com 4 King Ave 301-682-7772 Voice Frederick, MD 21701-3110 301-682-7666 FAX
Do you have any statistics for size and speed in different targets? How does it compare to the Xilinx core?
-- Rick "rickman" Collins rick.collins@XYarius.com Ignore the reply address. To email me use the above address with the XY removed. Arius - A Signal Processing Solutions Company Specializing in DSP and FPGA design URL http://www.arius.com 4 King Ave 301-682-7772 Voice Frederick, MD 21701-3110 301-682-7666 FAX
Here are parts of the synthesis reports for the naked core of an KCPSM3 in a SpartanII (XC2S100):
PacoBlaze: =========================================================================
Design Statistics # IOs : 58
Macro Statistics : # RAM : 3 # 16x8-bit dual-port distributed RAM: 1 # 32x10-bit single-port distributed RAM: 1 # 64x8-bit single-port distributed RAM: 1 # Registers : 19 # 1-bit register : 16 # 10-bit register : 1 # 5-bit register : 2 # Multiplexers : 10 # 1-bit 4-to-1 multiplexer : 1 # 2-to-1 multiplexer : 9 # Adders/Subtractors : 3 # 10-bit adder : 1 # 5-bit addsub : 1 # 8-bit adder carry in/out : 1 # Xors : 1 # 1-bit xor8 : 1
Cell Usage : # BELS : 273 # GND : 1 # LUT1 : 13 # LUT2 : 16 # LUT3 : 108 # LUT4 : 91 # MUXCY : 21 # MUXF5 : 1 # VCC : 1 # XORCY : 21 # FlipFlops/Latches : 46 # FDE : 2 # FDR : 11 # FDRE : 29 # FDRS : 2 # FDS : 2 # RAMS : 34 # RAM16X1D : 8 # RAM32X1S : 26 # Clock Buffers : 1 # BUFGP : 1 # IO Buffers : 57 # IBUF : 28 # OBUF : 29 =========================================================================
Device utilization summary:
---------------------------
Selected Device : 2s100etq144-6
Number of Slices: 200 out of 1200 16% Number of Slice Flip Flops: 46 out of 2400 1% Number of 4 input LUTs: 288 out of 2400 12% Number of bonded IOBs: 57 out of 102 55% Number of GCLKs: 1 out of 4 25%
========================================================================= TIMING REPORT
NOTE: THESE TIMING NUMBERS ARE ONLY A SYNTHESIS ESTIMATE. FOR ACCURATE TIMING INFORMATION PLEASE REFER TO THE TRACE REPORT GENERATED AFTER PLACE-and-ROUTE.
Clock Information:
------------------
-----------------------------------+------------------------+-------+ Clock Signal | Clock buffer(FF name) | Load |
-----------------------------------+------------------------+-------+ clk | BUFGP | 80 |
-----------------------------------+------------------------+-------+
Timing Summary:
--------------- Speed Grade: -6
Minimum period: 21.627ns (Maximum Frequency: 46.238MHz) Minimum input arrival time before clock: 24.700ns Maximum output required time after clock: 8.320ns Maximum combinational path delay: 10.734ns
Timing Detail:
-------------- All values displayed in nanoseconds (ns)
------------------------------------------------------------------------- Timing constraint: Default period analysis for Clock 'clk' Delay: 21.627ns (Levels of Logic = 16) Source: register_Mram_dpr_inst_ramx_3 (RAM) Destination: zero (FF) Source Clock: clk rising Destination Clock: clk rising
Data Path: register_Mram_dpr_inst_ramx_3 to zero Gate Net Cell:in->out fanout Delay Delay Logical Name (Net Name)
---------------------------------------- ------------ RAM16X1D:WCLK->DPO 2 1.568 1.150 register_Mram_dpr_inst_ramx_3 (register_y_data_out) LUT3:I2->O 35 0.468 3.375 scratch_address1 (scratch_address) LUT3:I2->O 1 0.468 0.000 alu_Madd__n0021_inst_lut2_171 (alu_Madd__n0021_inst_lut2_17) MUXCY:S->O 1 0.515 0.000 alu_Madd__n0021_inst_cy_18 (alu_Madd__n0021_inst_cy_18) MUXCY:CI->O 1 0.058 0.000 alu_Madd__n0021_inst_cy_19 (alu_Madd__n0021_inst_cy_19) MUXCY:CI->O 1 0.058 0.000 alu_Madd__n0021_inst_cy_20 (alu_Madd__n0021_inst_cy_20) MUXCY:CI->O 1 0.058 0.000 alu_Madd__n0021_inst_cy_21 (alu_Madd__n0021_inst_cy_21) XORCY:CI->O 1 0.648 0.920 alu_Madd__n0021_inst_sum_22 (alu_addsub_result) LUT4:I1->O 1 0.468 0.920 alu__old_result_653 (CHOICE603) LUT4:I3->O 3 0.468 1.320 alu__old_result_661 (alu_result) LUT3:I1->O 1 0.468 0.000 alu_Mmux__n0004_inst_mux_f5_0111_G (N11696) MUXF5:I1->O 2 0.403 1.150 alu_Mmux__n0004_inst_mux_f5_0111 (alu_shift_bit) LUT4:I1->O 1 0.468 0.920 alu__old_result_629 (CHOICE463) LUT4:I2->O 3 0.468 1.320 alu__old_result_661 (alu_result) LUT3:I1->O 1 0.468 0.920 Mmux__n0016_Result24 (CHOICE447) LUT3:I1->O 1 0.468 0.920 Mmux__n0016_Result49_SW0 (N11635) LUT4:I3->O 1 0.468 0.000 Mmux__n0016_Result49 (N10428) FDRE:D 0.724 zero
---------------------------------------- Total 21.627ns (8.712ns logic, 12.915ns route) (40.3% logic, 59.7% route)
------------------------------------------------------------------------- Timing constraint: Default OFFSET IN BEFORE for Clock 'clk' Offset: 24.700ns (Levels of Logic = 14) Source: instruction (PAD) Destination: zero (FF) Destination Clock: clk rising
Data Path: instruction to zero Gate Net Cell:in->out fanout Delay Delay Logical Name (Net Name)
---------------------------------------- ------------ IBUF:I->O 30 0.797 3.250 instruction_17_IBUF (instruction_17_IBUF) LUT3:I1->O 3 0.468 1.320 alu_Ker71491 (alu_N7151) LUT3:I0->O 9 0.468 2.150 alu__n00191 (alu__n0019) LUT3:I0->O 8 0.468 2.050 alu_Ker707067 (N9028) LUT3:I2->O 1 0.468 0.920 alu__old_result_653_SW0 (N11655) LUT4:I2->O 1 0.468 0.920 alu__old_result_653 (CHOICE603) LUT4:I3->O 3 0.468 1.320 alu__old_result_661 (alu_result) LUT3:I1->O 1 0.468 0.000 alu_Mmux__n0004_inst_mux_f5_0111_G (N11696) MUXF5:I1->O 2 0.403 1.150 alu_Mmux__n0004_inst_mux_f5_0111 (alu_shift_bit) LUT4:I1->O 1 0.468 0.920 alu__old_result_629 (CHOICE463) LUT4:I2->O 3 0.468 1.320 alu__old_result_661 (alu_result) LUT3:I1->O 1 0.468 0.920 Mmux__n0016_Result24 (CHOICE447) LUT3:I1->O 1 0.468 0.920 Mmux__n0016_Result49_SW0 (N11635) LUT4:I3->O 1 0.468 0.000 Mmux__n0016_Result49 (N10428) FDRE:D 0.724 zero
---------------------------------------- Total 24.700ns (7.540ns logic, 17.160ns route) (30.5% logic, 69.5% route)
------------------------------------------------------------------------- Timing constraint: Default OFFSET OUT AFTER for Clock 'clk' Offset: 8.320ns (Levels of Logic = 1) Source: register_Mram_dpr_inst_ramx_7 (RAM) Destination: out_port (PAD) Source Clock: clk rising
Data Path: register_Mram_dpr_inst_ramx_7 to out_port Gate Net Cell:in->out fanout Delay Delay Logical Name (Net Name)
---------------------------------------- ------------ RAM16X1D:WCLK->SPO 9 1.568 2.150 register_Mram_dpr_inst_ramx_7 (out_port_7_OBUF) OBUF:I->O 4.602 out_port_7_OBUF (out_port)
---------------------------------------- Total 8.320ns (6.170ns logic, 2.150ns route) (74.2% logic, 25.8% route)
------------------------------------------------------------------------- Timing constraint: Default path analysis Delay: 10.734ns (Levels of Logic = 3) Source: instruction (PAD) Destination: out_port (PAD)
Data Path: instruction to out_port Gate Net Cell:in->out fanout Delay Delay Logical Name (Net Name)
---------------------------------------- ------------ IBUF:I->O 10 0.797 2.250 instruction_10_IBUF (instruction_10_IBUF) RAM16X1D:A2->SPO 9 0.935 2.150 register_Mram_dpr_inst_ramx_1 (out_port_1_OBUF) OBUF:I->O 4.602 out_port_1_OBUF (out_port)
---------------------------------------- Total 10.734ns (6.334ns logic, 4.400ns route) (59.0% logic, 41.0% route)
PicoBlaze: =========================================================================
Design Statistics # IOs : 58
Cell Usage : # BELS : 196 # GND : 1 # INV : 3 # LUT1 : 2 # LUT2 : 6 # LUT3 : 71 # LUT4 : 30 # MUXCY : 39 # MUXF5 : 9 # VCC : 1 # XORCY : 34 # FlipFlops/Latches : 86 # FD : 21 # FDE : 2 # FDR : 33 # FDRE : 8 # FDRSE : 20 # FDS : 2 # RAMS : 18 # RAM16X1D : 8 # RAM32X1S : 10 # Clock Buffers : 1 # BUFGP : 1 # IO Buffers : 57 # IBUF : 28 # OBUF : 29 # Others : 8 # RAM64X1S : 8 =========================================================================
Device utilization summary:
---------------------------
Selected Device : 2s100etq144-6
Number of Slices: 129 out of 1200 10% Number of Slice Flip Flops: 86 out of 2400 3% Number of 4 input LUTs: 169 out of 2400 7% Number of bonded IOBs: 57 out of 102 55% Number of GCLKs: 1 out of 4 25%
========================================================================= TIMING REPORT
NOTE: THESE TIMING NUMBERS ARE ONLY A SYNTHESIS ESTIMATE. FOR ACCURATE TIMING INFORMATION PLEASE REFER TO THE TRACE REPORT GENERATED AFTER PLACE-and-ROUTE.
Clock Information:
------------------
-----------------------------------+------------------------+-------+ Clock Signal | Clock buffer(FF name) | Load |
-----------------------------------+------------------------+-------+ clk | BUFGP | 104 |
-----------------------------------+------------------------+-------+
Timing Summary:
--------------- Speed Grade: -6
Minimum period: 9.767ns (Maximum Frequency: 102.386MHz) Minimum input arrival time before clock: 10.997ns Maximum output required time after clock: 8.878ns Maximum combinational path delay: 11.292ns
Timing Detail:
-------------- All values displayed in nanoseconds (ns)
------------------------------------------------------------------------- Timing constraint: Default period analysis for Clock 'clk' Delay: 9.767ns (Levels of Logic = 13) Source: carry_flag_flop (FF) Destination: register_bit9 (FF) Source Clock: clk rising Destination Clock: clk rising
Data Path: carry_flag_flop to register_bit9 Gate Net Cell:in->out fanout Delay Delay Logical Name (Net Name)
---------------------------------------- ------------ FDRE:C->Q 4 0.992 1.520 carry_flag_flop (carry_flag_flop) LUT4:I0->O 2 0.468 1.150 condition_met_lut (condition_met) LUT3:I1->O 11 0.468 2.350 normal_count_lut (normal_count) LUT3:I0->O 1 0.468 0.000 value_select_mux0 (pc_value) MUXCY:S->O 1 0.515 0.000 pc_value_muxcy0 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy1 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy2 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy3 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy4 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy5 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy6 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy7 (pc_value_carry) MUXCY:CI->O 0 0.058 0.000 pc_value_muxcy8 (pc_value_carry) XORCY:CI->O 2 0.648 0.000 pc_value_xor9 (inc_pc_value) FDRSE:D 0.724 register_bit9
---------------------------------------- Total 9.767ns (4.747ns logic, 5.020ns route) (48.6% logic, 51.4% route)
------------------------------------------------------------------------- Timing constraint: Default OFFSET IN BEFORE for Clock 'clk' Offset: 10.997ns (Levels of Logic = 14) Source: instruction (PAD) Destination: register_bit9 (FF) Destination Clock: clk rising
Data Path: instruction to register_bit9 Gate Net Cell:in->out fanout Delay Delay Logical Name (Net Name)
---------------------------------------- ------------ IBUF:I->O 27 0.797 3.175 instruction_14_IBUF (instruction_14_IBUF) LUT4:I0->O 1 0.468 0.920 move_group_lut (move_group) LUT3:I2->O 11 0.468 2.350 normal_count_lut (normal_count) LUT3:I0->O 1 0.468 0.000 value_select_mux0 (pc_value) MUXCY:S->O 1 0.515 0.000 pc_value_muxcy0 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy1 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy2 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy3 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy4 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy5 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy6 (pc_value_carry) MUXCY:CI->O 1 0.058 0.000 pc_value_muxcy7 (pc_value_carry) MUXCY:CI->O 0 0.058 0.000 pc_value_muxcy8 (pc_value_carry) XORCY:CI->O 2 0.648 0.000 pc_value_xor9 (inc_pc_value) FDRSE:D 0.724 register_bit9
---------------------------------------- Total 10.997ns (4.552ns logic, 6.445ns route) (41.4% logic, 58.6% route)
------------------------------------------------------------------------- Timing constraint: Default OFFSET OUT AFTER for Clock 'clk' Offset: 8.878ns (Levels of Logic = 2) Source: register_bit70 (RAM) Destination: port_id (PAD) Source Clock: clk rising
Data Path: register_bit70 to port_id Gate Net Cell:in->out fanout Delay Delay Logical Name (Net Name)
---------------------------------------- ------------ RAM16X1D:WCLK->DPO 1 1.568 0.920 register_bit70 (sy) LUT3:I2->O 3 0.468 1.320 operand_select_mux7 (port_id_7_OBUF) OBUF:I->O 4.602 port_id_7_OBUF (port_id)
---------------------------------------- Total 8.878ns (6.638ns logic, 2.240ns route) (74.8% logic, 25.2% route)
------------------------------------------------------------------------- Timing constraint: Default path analysis Delay: 11.292ns (Levels of Logic = 4) Source: instruction (PAD) Destination: port_id (PAD)
Data Path: instruction to port_id Gate Net Cell:in->out fanout Delay Delay Logical Name (Net Name)
---------------------------------------- ------------ IBUF:I->O 10 0.797 2.250 instruction_4_IBUF (instruction_4_IBUF) RAM16X1D:DPRA0->DPO 1 0.935 0.920 register_bit00 (sy) LUT3:I2->O 3 0.468 1.320 operand_select_mux0 (port_id_0_OBUF) OBUF:I->O 4.602 port_id_0_OBUF (port_id)
---------------------------------------- Total 11.292ns (6.802ns logic, 4.490ns route) (60.2% logic, 39.8% route)
WOW! That is one small CPU. I didn't realize that the pB was that small. I will have to take a look to see why it is so much smaller than a stack machine. Does it include interrupts?
I notice that the PacoB uses less FFs, but more LUTs and runs slower than the pB. Have you analyzed it to see how they differ? I belive the source is available on the pB vs. the uB which must be bought, no?
Pablo Bleyer Kocik wrote:
-- Rick "rickman" Collins rick.collins@XYarius.com Ignore the reply address. To email me use the above address with the XY removed. Arius - A Signal Processing Solutions Company Specializing in DSP and FPGA design URL http://www.arius.com 4 King Ave 301-682-7772 Voice Frederick, MD 21701-3110 301-682-7666 FAX
It has a single interrupt pin, when asserted it vectors to the top of program memory - you place a jump there to branch to your actual interrupt routine.
If you need more than one interrupt channel you can easily craft a simple interrupt controller and hang it off one of picoblaze's IO ports, the IRQ handler queries the controller for the actual IRQ number, and vectors accordingly.
As Ken Chapman (Picoblaze creator) so insightfully says - what's most interesting about Picoblaze is not the processor itself, but the context that you embed it in.
That's correct - Picoblaze is distributed in source form, Microblaze is encrypted VHDL unless you buy the source for extra (see recent discussions).
Regards,
John
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.