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
82 lines
3.8 KiB
Python
82 lines
3.8 KiB
Python
#!/usr/bin/env python3
|
|
"""
|
|
Independent Python oracle for rtl/mac_unit.v and rtl/mac8.v (aspect C.1 of
|
|
the re-certification campaign, docs/validation/01-datapath.md).
|
|
|
|
Deliberately reimplements the arithmetic from first principles (two's
|
|
complement wraparound, sign extension) -- NOT by reading the Verilog and
|
|
transcribing it. Every function below is checked against Python's own
|
|
built-in arbitrary-precision integers, which have no width and cannot
|
|
share a truncation/sign-extension bug with the RTL.
|
|
|
|
Run standalone to regenerate the golden vectors used by
|
|
sim/mac_unit_tb.v / sim/mac8_tree_tb.v:
|
|
python3 tools/validation/mac_oracle.py
|
|
"""
|
|
|
|
def to_signed(val: int, width: int) -> int:
|
|
"""Interpret the low `width` bits of `val` as two's complement."""
|
|
val &= (1 << width) - 1
|
|
if val >= (1 << (width - 1)):
|
|
val -= (1 << width)
|
|
return val
|
|
|
|
def to_unsigned(val: int, width: int) -> int:
|
|
return val & ((1 << width) - 1)
|
|
|
|
def mac_unit(x: int, w: int, acc_in: int, data_width: int = 8, acc_width: int = 32) -> int:
|
|
"""Bit-exact model of rtl/mac_unit.v: acc_out = acc_in + sign_extend(x*w)."""
|
|
assert -(1 << (data_width - 1)) <= x < (1 << (data_width - 1))
|
|
assert -(1 << (data_width - 1)) <= w < (1 << (data_width - 1))
|
|
product = x * w # Python int, exact, no width -- the whole point of an independent oracle
|
|
prod_width = 2 * data_width
|
|
assert -(1 << (prod_width - 1)) <= product < (1 << (prod_width - 1)), \
|
|
"product overflowed PROD_WIDTH -- data_width assumption violated"
|
|
raw = to_unsigned(acc_in, acc_width) + to_unsigned(product, prod_width if product >= 0 else prod_width)
|
|
# acc_in + sign_extend(product) computed directly in signed arithmetic,
|
|
# then wrapped to acc_width bits (matches Verilog's silent wraparound
|
|
# on a fixed-width signed reg/wire -- confirmed intentional, not a
|
|
# bug, because callers rely on it: see docs/validation/01-datapath.md).
|
|
return to_signed(acc_in + product, acc_width)
|
|
|
|
def mac8_tree(products: list[int], acc_in: int, acc_width: int = 32) -> int:
|
|
"""
|
|
Bit-exact model of rtl/mac8.v's balanced binary adder tree +
|
|
final accumulator add. `products` must have length PARALLEL (a power
|
|
of two, matching the RTL's $clog2-based tree construction).
|
|
"""
|
|
n = len(products)
|
|
assert n > 0 and (n & (n - 1)) == 0, "PARALLEL must be a power of two"
|
|
level = list(products)
|
|
while len(level) > 1:
|
|
level = [to_signed(level[i] + level[i + 1], acc_width) for i in range(0, len(level), 2)]
|
|
return to_signed(acc_in + level[0], acc_width)
|
|
|
|
def mac8_full(x_vals: list[int], w_vals: list[int], acc_in: int,
|
|
data_width: int = 8, acc_width: int = 32) -> int:
|
|
"""End-to-end oracle: PARALLEL independent products -> balanced tree -> + acc_in."""
|
|
assert len(x_vals) == len(w_vals)
|
|
products = [mac_unit(x, w, 0, data_width, acc_width) for x, w in zip(x_vals, w_vals)]
|
|
return mac8_tree(products, acc_in, acc_width)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
# Self-check: a handful of hand-verifiable cases, printed for a human
|
|
# to eyeball before trusting this file as an oracle for anything else.
|
|
cases = [
|
|
(5, 7, 0, 35),
|
|
(-128, -128, 0, 16384), # the one INT8 x INT8 case that does NOT fit in INT16 magnitude terms symmetrically
|
|
(127, 127, 0, 16129),
|
|
(-128, 127, 0, -16256),
|
|
(0, 0, 0, 0),
|
|
(-1, -1, 0, 1),
|
|
]
|
|
ok = True
|
|
for x, w, acc_in, expected in cases:
|
|
got = mac_unit(x, w, acc_in)
|
|
status = "OK" if got == expected else "MISMATCH"
|
|
if got != expected:
|
|
ok = False
|
|
print(f"mac_unit(x={x}, w={w}, acc_in={acc_in}) = {got} (expected {expected}) {status}")
|
|
print("ALL SELF-CHECKS PASSED" if ok else "SELF-CHECK FAILURE -- oracle itself is wrong, fix before using it")
|