🔒
Confidential · Internal & Partner Access

Restricted framework reference

This page contains proposal-grade QUASAR field-theoretic scaffolding — speculative design language, chip lineup diligence, and never-state guardrails kept out of public marketing. Enter the access password to continue.

Confidential · WS13 STAR-PHASER / QUASAR Chip Lineup

QUASAR Framework

The unifying theoretical scaffold for QLT's photonic dimension ladder — from STAR-PHASER linear-circuit chips (Q < 0.2) through QUASAR hybrid and field-dominant machines (Q ≥ 0.2). Coupled state Ψ = (A, Φ, Ξ), the Q-metric, boundary-layer fab, five scaling laws, and a mandatory never-state register — consolidated from internal memos S81–S85.

Ψ = (A, Φ, Ξ) Q-metric Regions I–IV 7-layer fab stack Five scaling laws GALAXY ★ prototype Never-state register
§1 · Brand & Regime Split

STAR-PHASER → QUASAR

Same physics, different operating coordinate — the Q-metric separates linear circuits from field computers.

Tap any tile to expand. GEMINI ships today; all other chips are roadmap, research, or speculative-framework unless labeled.

I

STAR-PHASER

Region I · Q < 0.2 · linear QFP dominates
GEMToday

GEMINI — d=2, path qubits

v1 shipping product. Dual-rail path encoding on Si₃N₄; Q ≪ 0.05. OPC hygiene only. LDV dominates the action (S81).

d=2 · Region I · Q ≪ 0.05 · TODAY

SOLℤ₁₀

SOLAR — d=10 composite qudit

Structured composite — not GF(10). Q ≈ 0.05–0.10. Region I.

Never-state NS-03: never call SOLAR "GF(10)."

GAL★ Proto

GALAXY — QUASAR-transition prototype

GALAXY is not QUASAR. Rehearses Region II hardware in Region I operation (Q ≈ 0.08–0.15). See §6.

d=16 · GF(16) · Q < 0.2 · roadmap

TETBoundary

TETRIS — last pure STAR-PHASER

d=32, GF(32), Q ≈ 0.12–0.18. Final gate-locked product before QUASAR NOVA at d=64. Law IV AWG stress begins (S84).

II+

QUASAR

Region II+ · Q ≥ 0.2 · hybrid & field-dominant
NOVFirst

NOVA — first QUASAR chip

d=64, GF(64), Q ≈ 0.25–0.45. Distributed OPC mesh, 8×8 octet tiling. ~4× OPC margin at Δf=50 GHz.

SUPCrisis

SUPER — paradigm break

d=128, Q ≈ 0.35–0.65. 16k phase pairs, ~2.7× OPC margin. OPC + squeeze mandatory.

THΘQ≈1

THETA — field balance

d=256, Q ≈ 0.85–1.05. Continuous OPC mesh, Δf=25 GHz. Speculative-framework

LOTApex

LOTUS — Region IV horizon

d=512, Q ≈ 1.2–1.6. Split-band OPC, CV-primary. Research horizon only (NS-08).

Canonical brand rule

STAR-PHASER  ⟺  Region I only  (Q < 0.2)   →  GEMINI · SOLAR · GALAXY · TETRIS
QUASAR       ⟺  Region II+     (Q ≥ 0.2)   →  NOVA · SUPER · THETA · LOTUS
§2 · Field Theory · S81

Lagrangian: Ψ = (A, Φ, Ξ)

Coupled three-sector state with decomposed action — design scaffold, not measured Hamiltonian.

Speculative-framework Entire §2 carries this label. Euler–Lagrange equations recover STAR-PHASER limits at Q ≪ 1 and field-computing limits at Q ≳ 1 — pending THETA/LOTUS telemetry.

A — Algebraic

GF(2^m) control

Symbol labels on frequency bins. QFP sandwiches, FAU kernels (GFADD/GFMUL). Slow manifold updated at gate clock. Maps to fab L3–L4 (TFLN EOM + QFP mesh).

Φ — Field

Spectral manifold

Continuous optical state ψ(f,t). Nonlinear dynamics (FWM, OPC) act primarily on Φ. Maps to L2 SiN bus + comb.

Ξ — Stabilizer

Syndrome manifold

Witness observables — pairwise phases, homodyne quadratures, ChiL scalars, RS syndromes. Maps to L7 readout + digital ECC.

S

Action decomposition

Five Lagrangian sectors
DVQFP

LDV — discrete-variable kinetic

L_DV = Σ_k (iℏ/2)(â_k† ∂_t â_k − ∂_t â_k† â_k) − V_QFP(A, Φ)

Standard photonic mode kinetic energy + QFP gate potential. Region I: V_QFP ≫ nonlinear terms.

CVSqueeze

LCV — continuous-variable squeezing

L_CV = ∫ df (|∂_t q(f)|² + |∂_t p(f)|²)/2 − V_sqz(q, p; r(f))

Mandatory from SUPER upward. Reduces effective σ_gate in whitepaper §2.3.

OPCConjugate

LOPC — phase conjugation

L_OPC = − κ_OPC Σ_k (â_k† â_k* + h.c.) · η_OPC(k)

χ⁽³⁾ FWM with undepleted pump [demonstrated theory]. THETA+: distributed along bus z.

INTCoupling

LINT — A–Φ–Ξ interaction

EOM drives algebra onto field. ChiL witness loop reads Φ, writes Ξ. Digital QEC maps Ξ → A — OPC ≠ QEC compliance.

FWMχ⁽³⁾

LFWM — four-wave mixing

Supplies H_NL bucket. GALAXY: first chip with non-zero L_FWM in production stack. Region II+: duty rises until nonlinear ≈ linear.

Coupling picture
Gate instruction  →  δA  →  drives  δΦ  via EOM/QFP
Noise / drift       →  δΦ  →  observed by  δΞ
OPC + squeeze       →  feedback on Φ, witness update on Ξ
Digital QEC         →  discrete jump on A, Ξ (syndrome decode)
Regime limits
LimitConditionChips
Linear circuitQ ≪ 0.2GEMINI–TETRIS
Hybrid0.2 ≤ Q < 1NOVA, SUPER
Field balanceQ ≈ 1THETA
Field-dominantQ > 1LOTUS
§3 · Q-Metric · S82

Regions I–IV

When does a photonic processor stop behaving like a linear interferometer with nonlinear patches?

Q ≡ (H_NL + H_OPC + H_CV) / H_linear
Region I

Q < 0.2

STAR-PHASER. Linear QFP dominates. GEMINI → TETRIS.

Region II

0.2 ≤ Q < 1

QUASAR hybrid. Distributed OPC + squeeze. NOVA, SUPER.

Region III

Q ≈ 1 ± 0.1

Field balance. THETA. Speculative-framework

Region IV

Q > 1

Field-dominant. LOTUS. Research horizon only.

ChipdQ targetH_linearH_OPCH_CVH_NLRegion
GEMINI2≪ 0.05~98%~1%~1%~0%I
SOLAR10~0.05–0.10~90%~5%~5%~0%I
GALAXY ★16~0.08–0.15~85%~6%~4%~5%I
TETRIS32~0.12–0.18~82%~8%~5%~5%I
NOVA64~0.25–0.45~60%~15%~12%~13%II
SUPER128~0.35–0.65~50%~18%~17%~15%II
THETA256~0.85–1.05~45%~28%~22%~20%III
LOTUS512~1.2–1.6~35%~30%~25%~25%IV
η_FWM ≈ (γ_eff · P_pump · L_eff)²     (low-gain, Δk ≈ 0)
H_OPC ∝ N_OPC · η_FWM · P_pump
H_NL  ∝ N_FWM_pathways · (γ P L)²
η_eff = 1 − (1 − η_FWM)^K   (K recirculating passes)
§4 · Fab Stack · S83

Boundary-layer fab & 7-layer cross-section

Nonlinearity as distributed boundary layer co-equal with the SiN backbone.

Tap any layer to expand. L5 boundary NL is the QUASAR differentiator — rehearsed on GALAXY, continuous on LOTUS.

L1–7

Seven-layer vertical stack

Bottom → top · canonical fab diagram
L1Sub

Substrate — Si handle (725 µm)

Thermal sink. All QUASAR chips share this foundation.

L2Bus

Backbone — stoichiometric Si₃N₄ (400 nm)

Damascene bus + rings. 0.05–0.1 dB/cm class. Linear routing stays pure SiN.

L3EO

EO mix — TFLN heterobond TW-EOM (300 nm)

Gate drives on A sector. Maps to LDV.

L4Shape

Shaper — SiN microring SPU bank

16 rings (GALAXY) scaling to 512 rings (LOTUS). Per-bin spectral shaping.

L5★ NL

Boundary NL — AlGaAs / As₂S₃ χ⁽³⁾ overlay

200–500 nm overlay where mode overlap exceeds threshold. OPC + FWM execute here. GALAXY: 2–3 cells (~15–20 mm period). LOTUS: continuous mesh (~2–3 mm). BEOL deposition <250 °C.

How we do it: Adiabatic taper cell Lcell = 2–4 mm; comb-derived pump rotation per G45.

L6CV

CV inject — squeeze coupler + OPO shunt

Evanescently couples to SiN bus at octet boundaries. L0 boundary (GALAXY optional 3–6 dB) → L3 field (THETA/LOTUS ≥8–10 dB). Co-fabricated windows with L5.

L7RO

Readout — WDM + SPAD / PD array

Top BEOL metal. Homodyne witness per octet for ChiL. Maps to Ξ sector.

Activation zones Z0–Z4
ZoneLocationDutyFunction
Z0 passiveI/O routing0% NLLow-loss transport
Z1 refreshBetween QFP stagesLow–medOPC conjugation
Z2 computeDense QFP blocks0% χ⁽³⁾SU(d) gates
Z3 CVOctet boundariesMedium3–10 dB squeeze
Z4 pump busParallel waveguideHigh when activePump routing
Spectral coupling grid
ChipdΔfBcombNFWM
GALAXY1650 GHz750 GHz~3.8k
NOVA6450 GHz3.15 THz~262k
THETA25625 GHz6.38 THz~16.8M
LOTUS51220 GHz10.2 THz~134M

Grid: fk = fref + k·Δf. LOTUS requires split-band factorization (Band A ∥ Band B).

§5 · First Prototype · S85 §1.5

GALAXY — QUASAR-transition prototype

Trials Region II hardware in Region I operation. GALAXY is not QUASAR.

Build card
  • SiN microcomb 16-line · Δf=50 GHz · Bcomb=750 GHz
  • TFLN EOM · dual-octet shapers (2×8 genesis)
  • 2–3 AlGaAs distributed OPC cells (~15–20 mm period)
  • GF(16) FAU — GFADD/GFMUL/GFINV kernels
  • Optional 3–6 dB on-chip squeeze
  • Q target: 0.08–0.15 < 0.2 (Region I)
Prototype vs full QUASAR
ElementGALAXYNOVA+
L5 boundary layer2–3 AlGaAs cells8–100+ mesh nodes
Mesh period~15–20 mm sparse2–5 mm distributed
CV injection3–6 dB optionalMandatory ≥6 dB
Octet tiling2×8 genesis8×8 full scale
Q target0.08–0.15≥0.25
SignoffT-GALAXY matrixT-NOVA+

T-GALAXY milestone gating chain

T-GALAXY-01   GF(16) gate F_ρ > 0.95              [to-be-tested]
T-GALAXY-OPC  AlGaAs distributed OPC η            [to-be-tested] ──gates──► NOVA OPC claims
T-GALAXY-CAL  240-pair calibration σ_φ ≤ 0.05 rad  [to-be-tested]
T-GALAXY-Q    Pump-power vs phase fidelity Q < 0.2  [to-be-tested]
T-GALAXY-SQZ  3–6 dB on-chip squeeze              [roadmap]
§6 · Phase Diagram · S85

Stability phase diagram

Q vs d — canonical diligence artifact for the STAR-PHASER / QUASAR lineup.

Q ↑ 1.8 ┤ ╭ LOTUS (1.4) ··· Region IV · research horizon 1.6 ┤ ·····╯ 1.4 ┤ ····· 1.2 ┤ ····· 1.0 ┤ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ THETA (1.0) ─ ─ Region III · field balance ─ ─ ─ ─ ─ 0.8 ┤ ··· 0.6 ┤ SUPER (0.50) ★ crisis zone 0.4 ┤ NOVA (0.35) ← first QUASAR 0.2 ┤ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ 0.15┤ TETRIS (0.15) 0.12┤ GALAXY ★ (0.12) ← QUASAR-transition prototype 0.08┤ SOLAR (0.08) 0.05┤ GEM (≪0.05) TODAY └────┬────┬────┬────┬────┬────┬────┬────┬────→ log₂(d) 1 4 8 16 32 64 128 256 512 ├────────── Region I (STAR-PHASER) ──────────┤├──── Region II ────┤├III┤├IV┤ ↑ d=64 brand boundary
Contour I

OPC margin = 2×

Crosses SUPER, THETA. Amber curve vs d. Tightens as Bcomb → BOPC.

Contour II

Npair = 10⁴

SUPER crisis at 16k pairs. ChiL++ mandatory. Linear + periodic OPC breaks.

Contour III–V

FWM · AWG · Qmax

NFWM=10⁶ crosses SUPER–THETA. AWG limit at d≈32–64. Red zone Q>1.2 at LOTUS.

ChipdQ (design)RegionBrandAnnotation
GEMINI2≪ 0.05ISTAR-PHASERTODAY — v1 ships
SOLAR10~0.08ISTAR-PHASERℤ₁₀ composite
GALAXY ★16~0.12ISTAR-PHASERQUASAR-transition prototype
TETRIS32~0.15ISTAR-PHASERLast pure STAR-PHASER
— — — BOUNDARY Q = 0.2 — — —
NOVA64~0.35IIQUASARFirst QUASAR · WS12 synthesis
SUPER128~0.50IIQUASARCrisis zone · OPC+squeeze mandatory
THETA256~1.0IIIQUASARField-computing experimental
LOTUS512~1.4IVQUASARApex · speculative-framework
Mat

Lineup maturity table

Demonstrated vs roadmap vs speculative
DemToday

Demonstrated / v1 today

GEMINI: Path qubits, SFWM class — v1 ships. External: Kues D=10 qudit, Lu 64×64 FB unitaries, Fülöp octave microcomb.

RoadBuild

Roadmap / designed-target

SOLAR d=10 stack · GALAXY 16-bin tile + T-GALAXY · TETRIS 32-tooth AWG stress · NOVA G46 tile + ChiL · SUPER T-SUPER OPC+squeeze.

SpecHorizon

Speculative-framework / experimental

THETA Q≈1 field computing · LOTUS Region IV Q>1 · continuous LOTUS mesh · field-QEC at d=512. Campaign blockers: single-photon OPC [to-be-tested]; OPC ≠ QEC [never-state].

§7 · Scaling Laws · S84

Five failure mechanisms

Coupled power laws in d — no single knob fixes all. Δf table is primary mitigation authority.

I–V

Five scaling laws

Tap to expand each law
IOPC

Law I — OPC bandwidth vs comb span

B_comb = (d − 1) · Δf   must satisfy   B_comb ≤ B_OPC / m_margin
m_margin = B_OPC / B_comb     (B_OPC ≈ 12 THz, m_margin ≥ 1.2 minimum)

First warn: SUPER (~2.7×). THETA @ Δf=50 GHz explicitly FAILS (12.75 THz > 12 THz). Mitigation: reduce Δf; split-band OPC (LOTUS).

IIPhase

Law II — Pairwise phase budget

N_pair = d · (d − 1)     (directed pairwise phase relationships)

ChiL stress begins TETRIS (992). SUPER crisis (16,384). THETA (65,280) FAIL without continuous OPC + global φ₀. Mitigation: octet factorization (8× reduction).

IIIFWM

Law III — FWM pathway explosion

N_FWM ≈ d² · (d − 1)     scales ~O(d³)

NOVA ~262k pathways (WARN). LOTUS ~134M (FAIL). Failure: gain competition, thermal runaway, crosstalk floors. Mitigation: pump rotation, pathway nulling, reduce γ·P·L in Z1.

IVAWG

Law IV — AWG / routing resolution

Δλ_bin ≈ λ² · Δf / c     @ λ ≈ 1550 nm

TETRIS monolithic AWG flagged (WARN). LOTUS 512-ch at limit (FAIL). Lower Δf helps Law I but worsens Law IV — coupled optimization. THETA Δf=25 GHz is Pareto point.

VQ

Law V — Nonlinear duty / Q ceiling

Q(d) = (H_NL + H_OPC + H_CV) / H_linear  →  Q > Q_max(d) ⇒ instability

Running GALAXY at NOVA pump duty → Q > 0.2 → thermal crosstalk catastrophe. Region III–IV stability is Speculative-framework. Mitigation: duty caps per zone, SBS guard, per-chip Q setpoint.

Canonical Δf recommendation table (WS13-MANIFEST authority)

ChipdΔfBcombOPC margin @12 THzPrimary lawStatus
GEMINI2N/Av1 today
SOLAR1050 GHz450 GHz≫ (~27×)PASS
GALAXY1650 GHz750 GHz≫ (~16×)Law II rehearsalPASS
TETRIS3250 GHz1.55 THz~7.7×Law IV WARNPASS
NOVA6450 GHz3.15 THz~3.8×Law III WARNPASS
SUPER12835 GHz4.45 THz~2.7×Laws II+IIIWARN
THETA25625 GHz6.38 THz~1.9×Law I WARNWARN
LOTUS51220 GHz10.2 THz~1.2×Law I + split-bandWARN*

* LOTUS ~1.2× assumes split-band OPC — single-band operation FAILS Law I. Bcomb = (d−1)·Δf.

LawFailure modeScales asFirst warnFirst fail
IOPC band overflowBcomb = (d−1)·ΔfSUPERTHETA @ Δf=50G
IIPhase calibrationNpair = d(d−1)TETRISSUPER (16k)
IIIFWM crosstalkNFWM ≈ d²(d−1)NOVALOTUS (134M)
IVAWG/demux resolutionΔλ ≈ λ²·Δf/cTETRISLOTUS
VQ / thermal runawayQ(d, duty)NOVAmis-run GALAXY
§8 · Mandatory Register · S85 §3

Never-state list

Prohibited public language — content review gate on all chip and framework pages.

NS

Never-state register NS-01–NS-15

Tap to expand · correct framing included
01–05Product

Shipping · encoding · QEC · squeeze

  • NS-01: ~~"GALAXY/NOVA/SUPER ships today"~~ → only GEMINI is v1; others are roadmap/research
  • NS-02: ~~"GF(d) = base-d decimal"~~ → use GF(2^m) notation
  • NS-03: ~~"SOLAR is GF(10)"~~ → structured composite qudit (ℤ₁₀ ring)
  • NS-04: ~~"OPC replaces QEC"~~ → OPC pre-layer + digital QEC
  • NS-05: ~~"Squeezing optional from SUPER up"~~ → OPC + squeeze mandatory
06–10Brand

STAR-PHASER · THETA · photon count · FT

  • NS-06: ~~"STAR-PHASER handles d≥64"~~ → QUASAR from NOVA
  • NS-07: ~~"THETA/LOTUS demonstrated"~~ → speculative-framework / experimental
  • NS-08: ~~"Region IV product"~~ → long-term research horizon
  • NS-09: ~~"16/64/512 bins = 16/64/512 photons"~~ → single-photon qudit
  • NS-10: ~~"Fault-tolerant qudit processor today"~~ → roadmap
11–15Claims

Gates · bases · AlGaAs · OPC band · extrapolation

  • NS-11: ~~"Native GF quantum gates on silicon"~~ → GF algebraic control layer
  • NS-12: ~~"QUASAR = different base"~~ → dimension d qudit system
  • NS-13: ~~"AlGaAs OPC demonstrated at QLT"~~ → [to-be-tested] T-GALAXY-OPC
  • NS-14: ~~"Single-band OPC covers LOTUS"~~ → split-band OPC
  • NS-15: ~~"Linear circuit extrapolates to THETA"~~ → field-computing architecture
Source memos
  • S81 — QUASAR Lagrangian Ψ=(A,Φ,Ξ)
  • S82 — Q-metric Regions I–IV
  • S83 — Boundary-layer fab 7-layer stack
  • S84 — Five scaling laws + Δf table
  • S85 — Stability diagram + never-states

Research corpus: 7-Research-Campaign/WS13-StarPhaser-Chip-Lineup/S81–S85. Promotion to canonical docs requires CAMPAIGN-MANIFEST ¶7 human approval.

Figures · Art direction

Framework visuals for asset production

QUASAR Ψ=(A,Φ,Ξ) three-sector Lagrangian coupling diagram with L_DV, L_CV, L_OPC, L_INT, L_FWM termsFig Q1
Fig Q1. Ψ=(A,Φ,Ξ) sector couplings [speculative-framework].
QUASAR stability phase diagram Q vs log₂(d) with Regions I–IV, eight chip pins, OPC margin and N_pair contoursFig Q2
Fig Q2. Stability phase diagram Q vs d, Regions I–IV.
Seven-layer boundary fabrication stack — Si₃N₄ bus, TFLN EOM, AlGaAs OPC overlay, optional squeezeFig Q3
Fig Q3. Seven-layer boundary fabrication stack.
Five scaling-law failure mechanisms Laws I–V with PASS/WARN/FAIL severity bars per chip dimensionFig Q4
Fig Q4. Five scaling-law failure mechanisms per chip d.
GALAXY QUASAR-transition prototype recipe and NOVA handoff gating chainFig Q5
Fig Q5. GALAXY QUASAR-transition prototype recipe.
Distributed OPC mesh cell architecture from GALAXY sparse to THETA/LOTUS continuous field-regimeFig Q6
Fig Q6. Distributed OPC mesh cells every 2–5 mm.
Related

From framework to chips

QUASAR framework consolidates WS13 diligence. Explore the GALAXY prototype, today's GEMINI product, and the GF(64) encoding roadmap.