Files
FPGA-Neural/hardware/v1/sim/int8_memory_access_tb.v
T
micheleandClaude Sonnet 5 dc0b331d3e feat(v2): scaffold hardware/v1 frozen baseline + M1 Neural Processor
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
2026-09-05 14:06:53 +02:00

405 lines
9.2 KiB
Verilog

`timescale 1ns/1ps
module tb;
localparam ADDR_WIDTH = 23;
localparam CLK_PERIOD = 12.5;
reg clk;
reg rst;
// ============================================================
// INT8 interface
// ============================================================
reg req;
reg wr;
reg [ADDR_WIDTH-1:0] addr;
reg signed [7:0] wdata;
wire signed [7:0] rdata;
wire ready;
// ============================================================
// Memory interface
// ============================================================
wire mem_req;
wire mem_wr;
wire [ADDR_WIDTH-1:0] mem_addr;
wire [15:0] mem_wdata;
wire mem_lb_n;
wire mem_ub_n;
wire [15:0] mem_rdata;
wire mem_ready;
// ============================================================
// DUT
// ============================================================
int8_memory_access #(
.ADDR_WIDTH(ADDR_WIDTH)
) dut (
.clk (clk),
.rst (rst),
.req (req),
.wr (wr),
.addr (addr),
.wdata (wdata),
.rdata (rdata),
.ready (ready),
.mem_req (mem_req),
.mem_wr (mem_wr),
.mem_addr (mem_addr),
.mem_wdata (mem_wdata),
.mem_lb_n (mem_lb_n),
.mem_ub_n (mem_ub_n),
.mem_rdata (mem_rdata),
.mem_ready (mem_ready)
);
// ============================================================
// Simple memory model
// ============================================================
reg [15:0] memory [0:1023];
reg [15:0] model_rdata;
reg model_ready;
integer i;
assign mem_rdata = model_rdata;
assign mem_ready = model_ready;
// ============================================================
// Clock
// ============================================================
initial begin
clk = 1'b0;
forever #(CLK_PERIOD / 2.0)
clk = ~clk;
end
// ============================================================
// VCD
// ============================================================
initial begin
$dumpfile("sim/int8_memory_access.vcd");
$dumpvars(0, tb);
end
// ============================================================
// Memory model
// ============================================================
always @(posedge clk) begin
model_ready <= 1'b0;
if (mem_req) begin
if (mem_wr) begin
if (!mem_lb_n)
memory[mem_addr][7:0] <= mem_wdata[7:0];
if (!mem_ub_n)
memory[mem_addr][15:8] <= mem_wdata[15:8];
end else begin
model_rdata <= memory[mem_addr];
end
model_ready <= 1'b1;
end
end
// ============================================================
// Write helper
// ============================================================
task write_byte;
input [ADDR_WIDTH-1:0] byte_addr;
input signed [7:0] data;
begin
@(posedge clk);
addr <= byte_addr;
wdata <= data;
wr <= 1'b1;
req <= 1'b1;
@(posedge clk);
req <= 1'b0;
wait (ready);
$display(
"WRITE BYTE addr=0x%08x data=0x%02x PASS",
byte_addr,
data
);
@(posedge clk);
end
endtask
// ============================================================
// Read helper
// ============================================================
task read_byte;
input [ADDR_WIDTH-1:0] byte_addr;
input signed [7:0] expected;
begin
@(posedge clk);
addr <= byte_addr;
wr <= 1'b0;
req <= 1'b1;
@(posedge clk);
req <= 1'b0;
wait (ready);
if (rdata !== expected) begin
$display(
"READ BYTE addr=0x%08x FAIL got=0x%02x expected=0x%02x",
byte_addr,
rdata,
expected
);
$fatal;
end else begin
$display(
"READ BYTE addr=0x%08x data=0x%02x PASS",
byte_addr,
rdata
);
end
@(posedge clk);
end
endtask
// ============================================================
// Test
// ============================================================
initial begin
req = 1'b0;
wr = 1'b0;
addr = 0;
wdata = 0;
model_rdata = 16'h0000;
model_ready = 1'b0;
for (i = 0; i < 1024; i = i + 1)
memory[i] = 16'h0000;
rst = 1'b1;
repeat (5)
@(posedge clk);
rst = 1'b0;
$display("");
$display("========================================");
$display("INT8 MEMORY ACCESS TEST");
$display("========================================");
$display("");
// ========================================================
// Basic byte writes
// ========================================================
write_byte(22'h000000, 8'h12);
write_byte(22'h000001, 8'h34);
write_byte(22'h000002, 8'h56);
write_byte(22'h000003, 8'h78);
// ========================================================
// Reads
// ========================================================
read_byte(22'h000000, 8'h12);
read_byte(22'h000001, 8'h34);
read_byte(22'h000002, 8'h56);
read_byte(22'h000003, 8'h78);
// ========================================================
// Verify physical packing
// ========================================================
if (memory[0] !== 16'h3412) begin
$display(
"PACKING ERROR word[0]=0x%04x expected=0x3412",
memory[0]
);
$fatal;
end else begin
$display(
"PACKING word[0] = 0x%04x PASS",
memory[0]
);
end
if (memory[1] !== 16'h7856) begin
$display(
"PACKING ERROR word[1]=0x%04x expected=0x7856",
memory[1]
);
$fatal;
end else begin
$display(
"PACKING word[1] = 0x%04x PASS",
memory[1]
);
end
// ========================================================
// Byte preservation test
// ========================================================
// Start with 0x1234
write_byte(22'h000010, 8'h34);
write_byte(22'h000011, 8'h12);
if (memory[8] !== 16'h1234) begin
$display(
"INITIAL WORD ERROR = 0x%04x",
memory[8]
);
$fatal;
end
// Change low byte only:
// 0x1234 -> 0x12AA
write_byte(22'h000010, 8'hAA);
if (memory[8] !== 16'h12AA) begin
$display(
"LOW BYTE PRESERVE ERROR = 0x%04x expected=0x12AA",
memory[8]
);
$fatal;
end else begin
$display(
"LOW BYTE PRESERVE 0x1234 -> 0x12AA PASS"
);
end
// Change high byte only:
// 0x12AA -> 0xBBAA
write_byte(22'h000011, 8'hBB);
if (memory[8] !== 16'hBBAA) begin
$display(
"HIGH BYTE PRESERVE ERROR = 0x%04x expected=0xBBAA",
memory[8]
);
$fatal;
end else begin
$display(
"HIGH BYTE PRESERVE 0x12AA -> 0xBBAA PASS"
);
end
// ========================================================
// Signed INT8 values
// ========================================================
write_byte(22'h000020, -8'sd1);
write_byte(22'h000021, -8'sd128);
write_byte(22'h000022, 8'sd127);
read_byte(22'h000020, -8'sd1);
read_byte(22'h000021, -8'sd128);
read_byte(22'h000022, 8'sd127);
// ========================================================
// Final result
// ========================================================
$display("");
$display("========================================");
$display("INT8 MEMORY ACCESS TEST PASSED");
$display("BYTE ADDRESSING : PASS");
$display("LOW BYTE : PASS");
$display("HIGH BYTE : PASS");
$display("INT8 PACKING : PASS");
$display("BYTE PRESERVATION : PASS");
$display("SIGNED INT8 : PASS");
$display("========================================");
$display("");
$finish;
end
endmodule