QPC has proved a complex quantum computation job classic computation cannot do

Twenty-one NGFS climate-risk logics computed together in one run on IBM Heron r2 (ibm_kingston, 156 qubits). QPC has proved a complex quantum computation job that classic computation cannot do.

Published in Computational Sciences

QPC has proved a complex quantum computation job classic computation cannot do

Share this post

Choose a social network to share with, or copy the URL to share elsewhere

This is a representation of how your post may appear on social media. The actual post will vary between social networks

Dr. Ilan Kreitmann

Quantum Polycontextural Computation (QPC) · quantumpolycontextural.ai · readytogo@quantumpolycontextural.ai

Abstract

QPC has proved a complex quantum computation job that classic computation cannot do. 

Complex computation means that all named parts of a job evolve together in one computation run. Classic computation is linear: it runs separate models one after another, and each passes its outputs to the next. The job is a climate-risk decision published by the Network for Greening the Financial System (NGFS), which central banks use for supervision. NGFS computes it as a chain of models. Quantum Polycontextural Computation (QPC) computed it in one run. Twenty-one published Phase V logics, covering carbon price, energy mix, temperature, crop failure, wildfire and flood damage, sat together as co-resident contextures on IBM Heron r2 (ibm_kingston, 156 qubits). Transjunctions coupled them during the run, and the register was read out once. The expectation written before submission, Net Zero 2050 over Current Policies, passed on raw counts: 0.651 against 0.284. The coupled register carried a cross-context correlation of 0.197; the same-depth separable control returned 0.002. Public IBM job dbssf4vq0h0s73bk2lr0 is the record. Classic arithmetic on the same frozen table reaches the same pair ranking by its own sequential route, within 0.003. The answer is shared. The job, all twenty-one logics evolving together in one run, is executed only by a quantum computer.

Keywords: quantum computation architecture; polycontextural logic; co-resident contextures; transjunctions; NGFS climate scenarios; IBM Heron r2; pre-registered hardware experiment

1. The goal and the result

The goal was to show the quantum computation community a real-world complex job that runs only on a quantum computer, and to run it. The job chosen is one the source institution itself computes as a chain: the NGFS ranking of named climate futures across transition, climate and physical-risk factors. QPC ran that job as one computation on a public IBM processor. Every logic was present on the processor at the same time, the logics acted on each other during the run, and the ranking came from a single readout. The goal is achieved, and the job identifier dbssf4vq0h0s73bk2lr0 is the public record.

2. What quantum computation has shown, and what it has not

2026 produced the field's strongest evidence yet. On 30 July, IBM and its partners announced three preprints and declared a quantum advantage era. The community Quantum Advantage Tracker lists them as active candidates that still require further benchmarking [1]. Google's Quantum Echoes experiment measured a second-order out-of-time-order correlator on 65 qubits and estimated 3.2 years for the same data on the Frontier supercomputer [2]. Q-CTRL simulated the dynamics of a 60-site Fermi-Hubbard chain on 120 qubits of an IBM Heron processor [3].

Every one of these results is a physics model. Classical simulation keeps answering them. IBM's 2023 utility experiment was reproduced more accurately by belief-propagation tensor networks [4]. Within a week of the Q-CTRL announcement, Algorithmiq reproduced single-site data from the 120-qubit run in two and a half minutes on standard hardware [3]. The contest between quantum hardware and classical simulation is real, and it is being fought on physics.

Industry demonstrations follow a different template. In the HSBC and IBM bond-trading trial, a quantum feature map ran offline on Heron processors, and classical machine-learning models were then trained on the transformed data [5]. That is the industrial pattern: a classical workflow in control, a quantum circuit as one step inside it. A 2026 survey of the field describes hybrid classical-quantum systems as the de facto operational model of the NISQ era, with classical resources orchestrating the workflow and performing the iterative optimisation [6].

One question remains open. Can a real-world decision made of many independently published logics be computed as one coupled quantum state and read out once, with an expectation registered before submission, a separable control at the same depth, and a public job identifier? We found no published industrial demonstration that does this. This paper reports one.

3. Linear and complex computation of one decision

We use one word in a fixed sense. Linear means sequential: separate models run one after another, each passing its outputs to the next. A chain of models that are each mathematically non-linear is linear in this sense. Complex computation means all named parts evolve together in one computation run.

NGFS computes climate risk linearly. Integrated assessment models such as REMIND-MAgPIE produce the transition pathways. MAGICC produces the climate response. The Climate Impact Explorer produces physical impacts [7]. The short-term scenarios link GEM-E3, a computable general equilibrium model, to the macro-financial model EIRIN and the credit model CLIMACRED, each handing its results to the next [8]. Each model solves its own equations jointly. The stack runs in sequence. A supervisor then weighs the finished tables against one another to rank the named worlds.

QPC computes the same ranking as one state. Each published logic becomes a contexture: a register of the processor carrying its own quantum logic. Contextures act on each other through transjunctions, junction qubits that couple the registers during the run. The relation between the logics lives in the quantum state, not in a classical step that joins finished parts. No classical optimiser or solver is in control, and the reported result is a measured property of the quantum state. A classical computer can describe that state. It does not execute it. The two computations reach the decision by different routes, and this paper is about the route.

4. The desk

The units are six of the seven NGFS Phase V worlds: Current Policies, Net Zero 2050, Below 2°C, Delayed Transition, Fragmented World and Low Demand [7]. Nationally Determined Contributions was excluded at the freeze [author: state the reason]. The locked pair is Net Zero 2050 against Current Policies.

Transition columns come from the NGFS IIASA Scenario Explorer, model REMIND-MAgPIE 3.3-4.8, region World. Physical columns come from the Climate Impact Explorer, United States, median. Horizons are 2030, 2050 and 2100. The Kotz-Wenz physical-GDP variables stay out of the panel, following the NGFS notice on the critiques of their underlying paper [9].

A pairwise Spearman screen on the public columns kept twenty-one logics against a minimum of fifteen: MAGICC temperature, CO2, carbon price, consumption, energy investment, coal, nuclear, biomass, crop failure, wildfire, soil moisture, maize yield and flood damage across the three horizons. The geometry is 6 × 21 + 3 = 129 architecture qubits on the 156-qubit Kingston register. Expectation NGFS-P5-E1 was locked before simulator and hardware submission: under the co-resident logics, Net Zero 2050 outranks Current Policies.

5. The run

The job ran as SamplerV2 on ibm_kingston, IBM Heron r2, 156 qubits, with 8,192 shots [10]. The coupled sitting and a same-depth separable control ran in one job family. Dynamical decoupling and measurement twirling were applied at submission. Readout mitigation was computed after retrieval as a diagnostic only. Every claim in this paper stands on raw counts. The public job identifier is dbssf4vq0h0s73bk2lr0 [11].

6. What the chip returned

NGFS-P5-E1 passed on raw counts. Net Zero 2050 scored 0.651 and Current Policies 0.284, a margin of 0.366. The full coupled ranking reads Net Zero 2050 0.651, Low Demand 0.648, Below 2°C 0.551, Delayed Transition 0.509, Fragmented World 0.335, Current Policies 0.284. Net Zero 2050 and Low Demand sit within 0.003 of each other, inside the shot noise of 8,192 samples, so they share first place. The two deep-decarbonisation worlds lead together, and the hot-house world comes last.

The architecture metric is the junction gap. The coupled register carried a mean absolute cross-context correlation of 0.197. The same-depth separable control returned 0.002. The gap of 0.195 is roughly a hundredfold separation: the joint structure lives in the coupled state and vanishes when the coupling is removed.

The classical equal-weight ordering of the same frozen panel puts Net Zero 2050 over Current Policies with a gap of 0.369. The quantum computation returned 0.366. A coupled state of twenty-one logics, read once on raw counts, landed on the industrial answer within 0.003.

7. Why classic computation cannot do this job

A classic computer executes one step, stores the result and starts the next. It can solve each NGFS model. It can add and weigh the finished columns. It reaches a ranking by that sequence. What it cannot do is hold twenty-one logics present together and acting on each other in one run, because its every intermediate state is a single determinate value. That joint run is the complex job. QPC executed it on ibm_kingston, and the same-depth control shows that the joint structure was real: 0.197 coupled against 0.002 separable, a hundredfold separation that disappears when the coupling is removed.

The two computations reached the same answer on the locked pair, 0.369 by classic equal-weight arithmetic and 0.366 by the quantum run. That agreement is a strength. It proves the complex quantum computation correct at scale: twenty-one coupled logics, 129 architecture qubits, raw counts, no decoder, no post-selection, and the known industrial answer. A complex quantum computation that is right where the answer is known is the foundation for every case where its answer will differ from the chain.

8. The next run: where the two computations part

NGFS states that its long-term scenarios do not represent tipping points and do not capture compound effects [7]. Its short-term scenarios, released in May 2025, add compound extreme events and cross-regional transmission of shocks through a chain of models [8]. In the Disasters and Policy Stagnation scenario, regional GDP losses range from 6% in Asia to 12.5% in Africa [8].

That pair of published products defines the next pre-registered test. The long-term factors, which contain no compound effects, go into a regional desk as coupled contextures and as a separable control. The transjunction rules are frozen from Phase V data alone and published before anyone reads the short-term outputs. The expectation: the coupled ranking moves toward the compound-risk ordering of the short-term model, and the separable control does not. A pass means the coupling carried compound structure that was absent from its inputs. The job identifier will be published whatever the result.

9. Conclusion

The quantum computing field has shown physics beyond the reach of tested classical methods, and it has shown industrial workflows in which a quantum circuit serves a classical pipeline. QPC has now proved a complex quantum computation job classic computation cannot do: a real-world supervisory decision, computed with all its logics together in one run. Six named NGFS worlds, twenty-one published logics, one Sampler job on IBM Heron r2. The expectation passed on raw counts. The coupling carried structure that the control did not. The answer matched the industrial arithmetic, and the job that produced it ran as one complex quantum computation that the classic chain executes only in sequence. Job dbssf4vq0h0s73bk2lr0 is the record.

References

[1] IBM Quantum (2026). Researchers demonstrate quantum advantage through trusted quantum computation. IBM Quantum blog, 30 July 2026. Quantum Advantage Tracker: https://quantum-advantage-tracker.github.io

[2] Google Quantum AI and collaborators (2025). Observation of constructive interference at the edge of quantum ergodicity. Nature.

[3] Q-CTRL (2026). Fast, accurate, high-resolution simulation of large-scale Fermi-Hubbard models on a digital quantum processor. Analysis and the Algorithmiq Majorana Propagation reproduction: PostQuantum, May 2026.

[4] Tindall, J., Fishman, M., Stoudenmire, E.M., and Sels, D. (2024). Efficient tensor network simulation of IBM's Eagle kicked Ising experiment. PRX Quantum, 5, 010308.

[5] HSBC and IBM (2025). HSBC demonstrates world's first-known quantum-enabled algorithmic trading with IBM. Press release, 25 September 2025.

[6] Amiri, S.M.H., and Zahan, A. (2026). Hybrid Classical-Quantum Systems: The Unsung Heroes of the NISQ Era. Preprints.org, doi:10.20944/preprints202608.1044.v1.

[7] Network for Greening the Financial System (2024). NGFS Climate Scenarios, Phase V, and High-level overview. https://www.ngfs.net/ngfs-scenarios-portal/

[8] Network for Greening the Financial System (2025). NGFS short-term scenarios: main takeaways; and NGFS (2026), Occasional Paper: Two Lenses on Climate Risk.

[9] Network for Greening the Financial System (2025). Note on key assumptions of NGFS long-term scenarios Phase V, notice on physical risk estimates.

[10] IBM Quantum. Processor types: Heron r2. https://quantum.cloud.ibm.com/docs/en/guides/processor-types

[11] Kreitmann, I. (2026). QPC NGFS Phase V desk, expectation NGFS-P5-E1. IBM Quantum job dbssf4vq0h0s73bk2lr0 on ibm_kingston. quantumpolycontextural.ai

[12] Günther, G. (1976-1980). Beiträge zur Grundlegung einer operationsfähigen Dialektik. Felix Meiner.

[13] Preskill, J. (2018). Quantum Computing in the NISQ era and beyond. Quantum, 2, 79.

Follow the Topic

Quantum Computing
Mathematics and Computing > Computer Science > Theory of Computation > Models of Computation > Quantum Computing