Competitor wave 2026-07-11 — segment map, 9-col comparison, moat mapping

[gHashTag]

Competitor wave — 2026-07-11

Market/strategy pass (not a code audit) mapping the drone-mesh / tactical-mesh / wireless-DePIN field around Tri-Net.

Report: docs/WAVE_REPORT_COMPETITORS_2026-07-11.md (branch feat/wave-competitors-2026-07-11).

Segment map (4)

• Tethered comms — Elistair Khronos, AT&T Flying COW, Zenith Aerotech.
• mm-wave mesh — TERASi RU1 (>60 GHz).
• MANET software stacks — Persistent MPU5, TrellisWare TSM/Katana, Silvus StreamCaster, Rajant Kinetic Mesh, Doodle Labs Mesh Rider, Mobilicom SkyHopper, goTenna Pro X2.
• Silicon-bound DePIN — Helium (ATECC608 identity-only), World Mobile, Pollen, Tri-Net.

Honesty state carried into the report

• main does not build (build.rs E0308/E0599).
• HELLO beacon-auth is built-but-not-wired — verify_mac/is_fresh never called on the receive path; daemon sends with mac_key=None.
• Only AX7203 (Artix-7) is proven-on-silicon; Zynq-7020 PL never flashed; all PHY/RF -sim.
• 4 SKY26b dies submitted, not returned. Compute-anchor = [Open conjecture] until the 2026-12-16 tape-out.
• Every competitor cited to a primary URL; vendor perf figures tagged vendor-claimed.

Moat thesis (one line)

Competitors are closed-waveform + firmware-as-abstraction + identity-only-if-any-hardware-binding. Trinity's assets line up on an orthogonal axis: open bit-exact numerics (GF16) + ternary on-board AI (BitNet b1.58) + silicon-anchored (eventually compute-anchored) economics. The axis is real; the compute-anchor tip is unproven pre-tape-out.

Three cooperation lanes

• A — Open-PHY interop with open-sdr/openwifi (GF16 FEC block + conformance vectors).
• B — Open compute-vs-identity attestation schema for wireless DePIN (extends the Helium ATECC608 identity baseline).
• C — Academic co-publication on FANET control-plane resilience (gated on beacon-auth being wired + an on-device run).

Anchor: phi^2 + phi^-2 = 3