# Reproduction — Shared-MAC matrix multiply — flat DSP cost, on silicon

A matrix multiply time-multiplexed through ONE shared multiplier instead of unrolling to N*K*P of them. Bit-identical to the unrolled matmul (so no separate proof — the win is structural and measured): on real fabric the whole product costs 4 DSP48s, and the SAME 4 for a 2x2 and a 3x3 — flat in matrix size.

| stage | artifact | tier |
|---|---|---|
| emit | 14 compiled targets | LOCAL |
| simulate | shared-MAC matmul vs exact real product (max cell error) = 2.01e-05 < 1e-3 (✓) | LOCAL — Q16.16 model, bit-exact to the emitted RTL |
| silicon | `matmul2_pipeline (2x2)` — WNS +1.693 ns @ 100 MHz on Arty A7-100T | REPLAY |

**Simulated and on silicon** (no separate proof needed — bit-identical to the unrolled matmul). One shared multiply-accumulate computes the whole product; the DSP cost stops scaling with matrix size — measured flat at 4 on real fabric.

Silicon evidence (REPLAY): `monogate-research/electronics_intake/matmul_shared_arty/evidence/FINDINGS.md`.