CUSTOM QUANTUM CIRCUITS · ibm_fez 156q · CO-PUBLICATION
Custom Quantum Circuits
You bring the research question. We design the Qiskit circuit, execute it on real IBM Quantum hardware, and enter the result in a public ledger your reviewers can re-run themselves. Every claim cites a job ID. No simulations presented as results.
THE EXECUTION STANDARD
A claim without a job ID is a simulation.
Most quantum research you can buy is simulated and presented as if it ran on hardware. We hold every result to four properties — and if a circuit cannot satisfy all four, it does not go in your report.
This is the same standard we hold our own research to — and it works on us. We ran full Shor-ECDLP circuits on ibm_fez with public job IDs; when we re-analysed the raw data the quantum signal didn't hold, so we corrected the claim ourselves. You get the method and the honest verdict, not the marketing.
CIRCUIT FAMILIES
Four families. One execution discipline.
Three of these are already in the public ledger with hardware results. The fourth is yours — scoped against your research question and executed to the same standard.
Shor / ECDLP attack circuits
Full Shor-ECDLP circuits executed on real ibm_fez hardware across 10 prime-order curves (4–21 bit), every job ID public. ⚠ The prior 'broke 17 curves' framing is RETRACTED — independent re-analysis found the key recovery was classical, not quantum (correct key at the noise floor). We run the real circuit; we do not claim a quantum break.
Quantum Phase Estimation surveys
QPE and coupling-sweep circuits executed on real ibm_fez hardware. ⚠ The prior 'α=1/137 detected, SNR 25.4×' claim is RETRACTED — re-analysis showed a DC noise pile-up against the wrong null, not a detection. The E8 coupling sweep (20 circuits / 10 angles) recorded an honest null. We report what the data shows, including 'no'.
Multi-qubit entanglement circuits
16-qubit entanglement circuits executed on real IBM hardware. ⚠ The prior 'full Bell certification / fidelity 1.0000' claim is RETRACTED — logs/quantum.log records the hardware CHSH test REJECTED (S≈0; a violation requires S>2). Real-hardware execution, no Bell certification claimed.
Custom circuit families, your problem
Circuit design against your research question — transpiled, error-mitigated, executed on ibm_fez, and entered into the open ledger under your engagement. Scoped per problem.
WHO COMMISSIONS THIS
When a simulation is not good enough.
The common thread: a result that has to survive an adversarial reviewer, a federal auditor, or a co-author who will re-run it. Simulated data does not survive that. Hardware-anchored, ledger-cited data does.
THE EXECUTION STACK
Eight stages. Specification to signed evidence.
No hardware time is spent on a circuit that fails simulation. No claim leaves the engagement without a job ID and a reproducible evidence hash.
Qiskit Design
Circuit specification, ansatz construction, parameter binding
Transpilation
Hardware-aware routing for ibm_fez coupling map, gate-set lowering
Dry-Run
Noiseless + noisy simulation gate before any hardware time is spent
ibm_fez Execution
156-qubit Heron r2 — real superconducting hardware, IBM Quantum
Error Mitigation
Measurement-error mitigation, dynamical decoupling, readout calibration
Statevector Export
Published statevector / counts so reviewers re-derive the result
Open Ledger Entry
Job ID cross-referenced in the public execution log
SATOR Evidence
Signed SQLite WAL with a reproducible hash — auditor-verifiable
ENGAGEMENT
From research question to co-published result.
Scoping
Research question, target circuit family, publication or coordinated-disclosure terms agreed and put in writing.
Circuit design
Qiskit circuit specified against the question. Ansatz, depth budget, and qubit count fixed to the ibm_fez topology.
Transpile + dry-run
Hardware-aware transpilation, then a noiseless and noisy simulation gate. No hardware time spent on a circuit that fails simulation.
Hardware execution
Executed on ibm_fez (156q). Every job ID recorded. Error mitigation and readout calibration applied.
Ledger entry
Result cross-referenced in the public open ledger. Statevector / counts published so the claim is independently re-runnable.
Co-publication
Joint paper or coordinated disclosure on the shared timeline agreed at scoping. Evidence file SATOR-signed and handed over.
ALREADY IN THE LEDGER
The standard, applied to our own research.
These are not brochures. Each one is a circuit family with hardware results and a public job trail you can open right now.
Bring the question.
We'll bring the hardware.
Scoped per circuit family and publication terms. Co-publication or coordinated disclosure on a shared timeline, agreed in writing before any hardware time is spent.
Every deliverable cites its IBM Quantum job ID, ships a published statevector, and carries a SATOR-signed reproducible evidence file. Acceptable to adversarial reviewers and federal auditors.