Begins the V2 Neural Multiprocessor / Dataflow architecture per docs/v2-description.md, per explicit user request to freeze V1 and start V2 development, copying from V1 what's needed. Scaffold: - hardware/v1/: byte-exact, read-only copy of the current V1 codebase (rtl, testbenches, tools, constraints, a representative subset of synthesis results, and reference docs) -- verified identical via diff/cmp against the live top-level tree before being made filesystem-read-only. The live top-level tree is untouched and remains the project's "production" V1 (see hardware/v1/README.md and hardware/v2/logs/decisions.log DEC-0001 for why copy-not-move). - hardware/v2/: mandatory structure (rtl/sim/constraints/synthesis/ reports/scripts/logs/docs) plus the full logging system required by the spec (development/architecture/simulation/synthesis/timing/ benchmark/decisions/experiments/errors.log). M1 -- Neural Processor (hardware/v2/rtl/neural_processor.v): - 8-stage pipelined perceptron unit (P_IN=8): input align, 8 multipliers, 3-level adder tree, accumulator, bias+activation, INT8 saturation. Genuine 1-tile/cycle throughput, not just a wider combinational datapath. - 7-state FSM (NP_IDLE..NP_ERROR per docs/v2-description.md §6, with 4 baseline states merged into NP_WAIT_OPERANDS -- see decisions.log DEC-0002); valid/ready/data/last stream interfaces per §7. - Bit-exact vs the frozen hardware/v1/rtl/neuron_parallel.v + mac8.v + mac_unit.v: 7/7 tests pass (hardware/v2/sim/tb_neural_processor.v), covering regular/mixed-sign/extreme-INT8 vectors, both activations, a zero-idle-gap back-to-back-tiles throughput check, and an 8-tile job -- verified with Verilator (see below for why). - Real synthesis + place&route (Yosys + nextpnr-ecp5): 0 CHECK problems, Fmax 183.12 MHz at ACC_WIDTH=32 (PASS at 80MHz, ~3x V1's isolated PARALLEL=8 Fmax of 61.71 MHz) and 176.21 MHz at ACC_WIDTH=24 (a user-requested comparison experiment, also bit-exact-verified; see experiments.log EXP-0001/EXP-0002 and benchmark.log). Three real bugs found and resolved during M1 development (full diagnostic record in errors.log): - Two independent, reproducible Icarus Verilog v13.0 scheduling defects (ERR-0001, ERR-0002) that silently produced wrong simulation results for standard sequential Verilog -- confirmed via Verilator 5.050 giving correct results on the same minimal repros. Verilator is now the trusted simulator for hardware/v2/ (decisions.log DEC-0004); Icarus's affected protocol-violation check was removed from the RTL and deferred architecturally to the Neural Director (DEC-0003) rather than chased further. - One real RTL bug (ERR-0003): last0 wasn't gated like valid0, letting a "last tile" tag leak into the pipeline ahead of its actual valid tile on back-to-back jobs. Fixed and verified. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013xXuuRUWZScuo1DeYJxs3v
258 lines
5.7 KiB
Verilog
258 lines
5.7 KiB
Verilog
`timescale 1ns/1ps
|
|
|
|
module tb;
|
|
|
|
parameter DATA_WIDTH = 8;
|
|
parameter N_INPUTS = 256;
|
|
parameter N_NEURONS = 4;
|
|
parameter PARALLEL = 32;
|
|
parameter ACC_WIDTH = 32;
|
|
|
|
reg clk;
|
|
reg rst;
|
|
reg start;
|
|
|
|
reg signed [DATA_WIDTH*N_INPUTS-1:0] x_bus;
|
|
|
|
reg signed [DATA_WIDTH*N_INPUTS*N_NEURONS-1:0]
|
|
weights_bus;
|
|
|
|
reg signed [DATA_WIDTH*N_NEURONS-1:0]
|
|
bias_bus;
|
|
|
|
wire signed [DATA_WIDTH*N_NEURONS-1:0] y_bus;
|
|
|
|
wire busy;
|
|
wire done;
|
|
|
|
integer i;
|
|
integer errors;
|
|
integer expected;
|
|
|
|
layer #(
|
|
.DATA_WIDTH(DATA_WIDTH),
|
|
.N_INPUTS(N_INPUTS),
|
|
.N_NEURONS(N_NEURONS),
|
|
.PARALLEL(PARALLEL),
|
|
.ACC_WIDTH(ACC_WIDTH)
|
|
) dut (
|
|
.clk(clk),
|
|
.rst(rst),
|
|
.start(start),
|
|
.x_bus(x_bus),
|
|
.weights_bus(weights_bus),
|
|
.bias_bus(bias_bus),
|
|
.y_bus(y_bus),
|
|
.busy(busy),
|
|
.done(done)
|
|
);
|
|
|
|
initial begin
|
|
clk = 0;
|
|
forever #5 clk = ~clk;
|
|
end
|
|
|
|
task run_layer;
|
|
begin
|
|
@(posedge clk);
|
|
start = 1;
|
|
|
|
@(posedge clk);
|
|
start = 0;
|
|
|
|
wait(done == 1);
|
|
|
|
@(posedge clk);
|
|
end
|
|
endtask
|
|
|
|
initial begin
|
|
|
|
$dumpfile("sim/parametric.vcd");
|
|
$dumpvars(0, tb);
|
|
|
|
rst = 1;
|
|
start = 0;
|
|
x_bus = 0;
|
|
weights_bus = 0;
|
|
bias_bus = 0;
|
|
errors = 0;
|
|
|
|
repeat (2) @(posedge clk);
|
|
|
|
rst = 0;
|
|
|
|
/*
|
|
* Tutti gli input = 1
|
|
*/
|
|
for (i = 0; i < N_INPUTS; i = i + 1)
|
|
x_bus[i*DATA_WIDTH +: DATA_WIDTH] = 8'sd1;
|
|
|
|
/*
|
|
* N0:
|
|
* primi 32 pesi = 1
|
|
* restanti 224 = 0
|
|
*
|
|
* Risultato = 32
|
|
*
|
|
* Questo verifica che l'accumulatore
|
|
* attraversi correttamente tutti gli 8 gruppi.
|
|
*/
|
|
for (i = 0; i < N_INPUTS; i = i + 1)
|
|
weights_bus[
|
|
0*N_INPUTS*DATA_WIDTH +
|
|
i*DATA_WIDTH +:
|
|
DATA_WIDTH
|
|
] = (i < 32) ? 8'sd1 : 8'sd0;
|
|
|
|
/*
|
|
* N1:
|
|
* tutti i 256 pesi = 0
|
|
* bias = 10
|
|
*
|
|
* Risultato = 10
|
|
*/
|
|
for (i = 0; i < N_INPUTS; i = i + 1)
|
|
weights_bus[
|
|
1*N_INPUTS*DATA_WIDTH +
|
|
i*DATA_WIDTH +:
|
|
DATA_WIDTH
|
|
] = 8'sd0;
|
|
|
|
bias_bus[1*DATA_WIDTH +: DATA_WIDTH] = 8'sd10;
|
|
|
|
/*
|
|
* N2:
|
|
* tutti i pesi = -1
|
|
*
|
|
* Risultato = -256
|
|
* ReLU -> 0
|
|
*/
|
|
for (i = 0; i < N_INPUTS; i = i + 1)
|
|
weights_bus[
|
|
2*N_INPUTS*DATA_WIDTH +
|
|
i*DATA_WIDTH +:
|
|
DATA_WIDTH
|
|
] = -8'sd1;
|
|
|
|
/*
|
|
* N3:
|
|
* primi 32 pesi = 4
|
|
* restanti = 0
|
|
*
|
|
* Risultato = 128
|
|
* Saturazione INT8 -> 127
|
|
*/
|
|
for (i = 0; i < N_INPUTS; i = i + 1)
|
|
weights_bus[
|
|
3*N_INPUTS*DATA_WIDTH +
|
|
i*DATA_WIDTH +:
|
|
DATA_WIDTH
|
|
] = (i < 32) ? 8'sd4 : 8'sd0;
|
|
|
|
$display("");
|
|
$display("========================================");
|
|
$display("INT8 / INT32 PARAMETRIC LAYER TEST");
|
|
$display("N_INPUTS = %0d", N_INPUTS);
|
|
$display("N_NEURONS = %0d", N_NEURONS);
|
|
$display("PARALLEL = %0d", PARALLEL);
|
|
$display("DATA_WIDTH = %0d", DATA_WIDTH);
|
|
$display("ACC_WIDTH = %0d", ACC_WIDTH);
|
|
$display("========================================");
|
|
|
|
run_layer;
|
|
|
|
/*
|
|
* Neuron 0
|
|
*/
|
|
expected = 32;
|
|
|
|
if ($signed(y_bus[0*DATA_WIDTH +: DATA_WIDTH]) !== expected) begin
|
|
$display("FAIL N0: got %0d expected %0d",
|
|
$signed(y_bus[0*DATA_WIDTH +: DATA_WIDTH]),
|
|
expected);
|
|
errors = errors + 1;
|
|
end
|
|
else
|
|
$display("PASS N0: %0d", $signed(y_bus[0*DATA_WIDTH +: DATA_WIDTH]));
|
|
|
|
/*
|
|
* Neuron 1
|
|
*/
|
|
if (N_NEURONS > 1) begin
|
|
|
|
expected = 10;
|
|
|
|
if ($signed(y_bus[1*DATA_WIDTH +: DATA_WIDTH]) !== expected) begin
|
|
$display("FAIL N1: got %0d expected %0d",
|
|
$signed(y_bus[1*DATA_WIDTH +: DATA_WIDTH]),
|
|
expected);
|
|
errors = errors + 1;
|
|
end
|
|
else
|
|
$display("PASS N1: %0d",
|
|
$signed(y_bus[1*DATA_WIDTH +: DATA_WIDTH]));
|
|
end
|
|
|
|
/*
|
|
* Neuron 2
|
|
*/
|
|
if (N_NEURONS > 2) begin
|
|
|
|
expected = 0;
|
|
|
|
if ($signed(y_bus[2*DATA_WIDTH +: DATA_WIDTH]) !== 0) begin
|
|
$display("FAIL N2: got %0d expected 0",
|
|
$signed(y_bus[2*DATA_WIDTH +: DATA_WIDTH]));
|
|
errors = errors + 1;
|
|
end
|
|
else
|
|
$display("PASS N2: 0 (ReLU)");
|
|
end
|
|
|
|
/*
|
|
* Neuron 3
|
|
*
|
|
* 32 * 4 = 128
|
|
* INT8 positive saturation -> 127
|
|
*/
|
|
if (N_NEURONS > 3) begin
|
|
|
|
expected = 127;
|
|
|
|
if ($signed(y_bus[3*DATA_WIDTH +: DATA_WIDTH]) !== expected) begin
|
|
$display("FAIL N3: got %0d expected %0d",
|
|
$signed(y_bus[3*DATA_WIDTH +: DATA_WIDTH]),
|
|
expected);
|
|
errors = errors + 1;
|
|
end
|
|
else
|
|
$display("PASS N3: %0d (saturation)",
|
|
$signed(y_bus[3*DATA_WIDTH +: DATA_WIDTH]));
|
|
end
|
|
|
|
if (busy !== 0) begin
|
|
$display("FAIL: busy still active");
|
|
errors = errors + 1;
|
|
end
|
|
|
|
if (done !== 1) begin
|
|
$display("FAIL: done not asserted");
|
|
errors = errors + 1;
|
|
end
|
|
|
|
$display("");
|
|
$display("========================================");
|
|
|
|
if (errors == 0)
|
|
$display("INT8 PARAMETRIC TEST PASSED");
|
|
else
|
|
$display("INT8 PARAMETRIC TEST FAILED: %0d errors", errors);
|
|
|
|
$display("========================================");
|
|
$display("");
|
|
|
|
$finish;
|
|
end
|
|
|
|
endmodule |