The parade example of what pure polycontextural quantum architecture can do — and why every other quantum computer on Earth cannot do the same thing.
Frustrated quantum magnets — materials where competing magnetic interactions create no single stable ground state — are central open problems in condensed matter physics. Understanding their phase transitions connects directly to some of the most consequential material science challenges of our time:
These are not theoretical curiosities. They are billion-dollar research frontiers where progress is blocked by the inability to computationally represent what a frustrated quantum system actually does — which is to exist in multiple competing phases simultaneously, with no preferred ordering. Standard computers, both classical and quantum, cannot properly model this. QPC can.
A frustrated spin system has multiple simultaneously competing ground states. The ferromagnetic, antiferromagnetic, and spin-liquid phases do not take turns — they coexist as genuine simultaneous physical alternatives. Any computation that forces a choice between them destroys the very thing being studied.
Standard quantum computers using Boolean logic must collapse the system to a single context. They force a choice between competing phases, losing the rich structure of the phase diagram entirely.
To study multiple phases, they run separate circuits for each and compare results classically afterward. This is not simultaneous multi-phase computation — it is sequential single-phase computation with classical post-processing.
The frustrated phase specifically — the spin-liquid — cannot be represented at all. Boolean evaluation forces a preferred ordering that the frustrated Hamiltonian does not have. The phase is not misrepresented. It is erased.
QPC's morphogrammatic encoding assigns each phase to a distinct logical contexture. All competing phases coexist simultaneously as separate contextures within a single quantum circuit.
Transjunctional operations allow these contextures to interact with each other without forcing collapse to a single answer. The output is not one ground state — it is the full polycontextural phase structure.
The spin-liquid phase is encoded as a genuine third contexture with its own internal structure — incommensurate phases, suppressed correlations, frustration signatures — all preserved and measurable simultaneously with the FM and AFM contextures.
IBM researchers themselves acknowledged that "quantum computers cannot prove executing complex real-world tasks" beyond narrow demonstrations — this is precisely because they lack multi-context logical architecture. The gate count is not the constraint. The logical architecture is.
Network World, 2025 — reporting on IBM Quantum research findings
The constraint is not hardware. IBM's Heron processor has 156 qubits and executes circuits with thousands of gates. Google's Willow demonstrated error suppression at scale. IonQ runs trapped-ion systems with high fidelity. The hardware has outpaced the architecture.
IBM's own roadmap targets "accurate execution of a quantum circuit at a scale beyond exact classical simulation — 5,000 gates on 156 qubits." Even at that scale, IBM remains in single-context Boolean logic. More gates do not solve the multi-context representation problem. They make the single-context computation larger — not richer.
QPC demonstrates something different: that the number of gates is not the constraint — the logical architecture is. A 104-gate QPC circuit on 128 qubits does something that a 5,000-gate standard quantum circuit cannot do at any qubit count, because the capability is architectural, not computational.
QPC operates as a universal quantum computation layer above the hardware. It does not modify or replace IBM's qubits, gates, or error correction. It augments the logical interpretation of quantum computation itself — replacing Boolean evaluation with polycontextural logic that can sustain multiple coexisting logical contextures.
The result, demonstrated here on real IBM Heron hardware across three qubit scales, is a computation that has never been performed before: three physically distinct quantum phases — ferromagnetic, antiferromagnetic, and spin-liquid — simultaneously active, simultaneously evolving, and simultaneously measurable in a single quantum circuit.
Detects the spin-liquid phase that Boolean logic forces to random noise. QPC ctx2 shows measurably suppressed ZZ correlations — the physical signature of genuine frustration — while the standard baseline cannot distinguish frustration from disorder.
Measures transjunctional structure between contextures. This metric has no standard quantum equivalent — single-context logic has no inter-context boundary to measure. ICC confirms that QPC's contextures are interacting in a structured, architecturally defined way that survives hardware noise.
ctx0, ctx1, and ctx2 show physically distinct ZZ correlator profiles simultaneously — ferromagnetic ordering, antiferromagnetic structure, and spin-liquid suppression all present in a single measurement record. Standard quantum produces one undifferentiated profile.
Both columns run on IBM Quantum hardware. The difference is entirely architectural — same qubits, same gates, different logical organization.
| Criterion | Standard IBM Quantum (single-context baseline) | QPC on IBM Heron |
|---|---|---|
| 100% quantum (no hybrid) | No — needs classical preprocessing to compare phases | Yes — pure QPC encoding, all phases in one circuit |
| Multi-context logic | No — Boolean evaluation, single context only | Yes — morphogrammatic blocks + transjunctional operations |
| Parallel computation | Partial — one context per circuit run | Full — all three phases computed simultaneously |
| 128–156 qubit usage | Inefficient — qubits wasted on context switching overhead | Efficient — each qubit carries structured polycontextural state |
| Frustrated phase representation | Erased — Boolean forces preferred ordering that doesn't exist | Preserved — ctx2 encodes incommensurability structurally |
| Complexity representation | Low — forced collapse to single ground state | High — structured coexistence of competing phases |
| NISQ compatible | Yes | Yes — max depth 104 gates across 128 qubits |
The theoretical argument is above. What follows is what IBM Heron hardware actually produced when QPC executed the Polycontextural Frustrated Quantum Magnet at 27, 64, and 128 qubits. Every number below is a raw hardware result — no simulation, no post-selection, no classical preprocessing of circuit outputs.
The defining test of a genuine architectural advantage versus a noise artifact: does the effect behave consistently and physically correctly as scale increases? QPC passes this test on all primary metrics across three independent hardware runs on IBM Heron.
What this chart proves. ctx2 (spin-liquid) remains the lowest correlator at all three qubit scales — 0.202 at 27Q, 0.076 at 64Q, 0.101 at 128Q. ctx0 (ferromagnetic) remains the highest throughout. The baseline decays faster than QPC structure as qubit count increases. If these results were noise artifacts, all contexts would converge toward the same value as scale increases. Instead they maintain distinct physical identities across three independent hardware runs.
The theta sweep scans the system across its full phase diagram. At each angle, three contextures evolve independently under their morphogrammatic encoding, interact via transjunctions, and are measured simultaneously on one 128-qubit circuit running on IBM Heron.
The most striking single data point: at θ = 0, ctx0 shows 75.5% ZZ correlation (ferromagnetic ordering), ctx1 shows 68.1% (antiferromagnetic), and ctx2 shows just 7.0% (spin-liquid frustration). One 128-qubit circuit. Three physically distinct quantum phases active and measurable simultaneously. The single-context baseline at the same angle produces one undifferentiated ordered state at 85.7% — it cannot distinguish between the phases at all.
Reproducible across all three qubit scales, at θ ≈ 2.693 rad the spin-liquid contexture (ctx2) shows elevated rather than suppressed correlations. This inversion appears at the same angle at 27Q, 64Q, and 128Q — it is not noise. It is a property of the QPC architecture.
θ = 2.693 rad: ctx2 |ZZ| = 0.294 while ctx0 |ZZ| = 0.069 and baseline = 0.053. The spin-liquid contexture is more ordered than the ferromagnetic contexture at this angle. The anomaly partially resolves at 128Q, suggesting a finite-size effect that evolves with contexture register size.
The most likely explanation is a secondary ordering transition near θ = 2π/3 × (1 + 1/√2), arising from the irrational phase formula in ctx2's morphogrammatic encoding. At this specific angle, constructive interference across the frustrated contexture produces an emergent ordered state that has no counterpart in either the FM or AFM contextures — and is completely invisible to single-context quantum computation. This is an open research question in QPC architecture.
All runs executed on IBM Quantum backend ibm_fez (Heron class, 156Q). Circuits compiled with Qiskit IBM Runtime SamplerV2 at optimization level 3. No readout error mitigation — raw hardware counts only. All 48 job IDs listed on the results page for independent verification.
Each QPC PFQM circuit consists of two morphogrammatic passes (one at θ, one at θ/2) separated by three directed transjunctions (ctx0→ctx1, ctx1→ctx2, ctx0→ctx2). The spin-liquid contexture uses Hadamard initialization with irrational-multiple RZ phases to encode incommensurability. No Toffoli gates. No classical feedback. Purely quantum execution.
| Parameter | 27Q run | 64Q run | 128Q run |
|---|---|---|---|
| Qubits per contexture | 9 | 21 | 42 |
| Bridge pairs (transjunction) | 3 | 5 | 8 |
| Max transpiled depth | 63 | 81 | 104 |
| Shots per circuit | 4096 | 4096 | 4096 |
| Mean ICC | 0.259 | 0.206 | 0.187 |
| ctx2 mean |ZZ| | 0.202 | 0.076 | 0.101 |
| FSP suppression | +0.193 | +0.235 | +0.182 |
| QPC advantage confirmed | Yes | Yes | Yes |
ICC (Inter-Context Correlation) — differential ZZ correlator asymmetry between even and odd bridge qubit pairs across the transjunction boundary. Values above 0.05 confirm structured cross-context correlations surviving hardware noise. No equivalent metric exists in standard quantum computation.
FSP (Frustration Signature Presence) — mean absolute ZZ correlator suppression in ctx2 relative to the single-context baseline. Positive values confirm ctx2 encodes frustration more correctly than Boolean logic can represent.
ZZ Correlator ⟨ZᵢZⱼ⟩ — computed directly from shot counts, range [−1, +1]. Full statistical power at 4096 shots regardless of qubit count. Directly interpretable as a physical observable in condensed matter physics.