Why I am leading Ireland Quantum 100
I am Michael English. I have spent twenty years building production systems inside Tesco, Dunnes, and Oracle — the kind of work where a release that goes wrong at 02:00 wakes a country, not a stand-up. I founded IMPT in 2024 because the climate-offset stack needed engineering rigour rather than narrative. Ireland Quantum 100 is the next layer of the same problem: the chemistry that decides whether direct-air-capture sorbents, perovskite photovoltaics, and next-generation battery cathodes are viable cannot be solved on classical hardware in any reasonable timeframe. It needs a sovereign machine, on Irish soil, run by people who treat uptime as a first-class engineering discipline.
This page is the engineering position behind the project — why a Clonmel-based founder with a platform-engineering background is the right person to deliver a 100-physical-qubit superconducting transmon system, and what that actually means in hardware and software terms.
The architecture I am committing to
Ireland Quantum 100 is a superconducting transmon machine. That decision is not aesthetic. Transmons — Josephson-junction-based charge qubits operated in the regime where E_J/E_C is large enough (typically >50) to suppress charge-noise dephasing — are the only modality with a credible path from current NISQ-class devices to surface-code-protected logical qubits inside a five-to-seven-year window. Trapped ions have superior coherence but poor gate-clock economics for the chemistry workloads we care about. Photonic and neutral-atom platforms are interesting; neither yet has the gate fidelity envelope for variational chemistry at 100-qubit scale.
The physical stack:
- Dilution refrigerator operating at the mixing-chamber stage below 15 mK, with the qubit plane targeted at 10–12 mK to keep
k_B Twell under the transmon level spacing (typically 4–6 GHz, soℏω/k_B≈ 200 mK). - Heavy-hex coupling topology — degree-3 connectivity that trades raw two-qubit reach for reduced frequency-collision density and a cleaner path to surface-code and heavy-hex code embeddings.
- Fixed-frequency transmons with cross-resonance two-qubit gates, the conservative choice. Tunable couplers are on the roadmap for the second-generation chip; they are not on the day-one die.
- Multiplexed dispersive readout via Purcell-filtered resonators, with TWPA-class quantum-limited amplifiers at the 4 K stage.
- Room-temperature control via an FPGA-based AWG/digitiser stack with sub-ns timing resolution, exposed to users through OpenQASM 3, Qiskit, PennyLane, and Cirq.
Nothing in that list is exotic. It is the conservative, instrumented, surface-code-compatible path. That is deliberate. Ireland's first sovereign quantum machine should not be a science experiment — it should be a workload-bearing instrument from first-light.
Why a platform engineer, not a physicist, runs the programme
A 100-qubit superconducting machine is, in operational terms, a distributed real-time system with a 10-millikelvin endpoint. The hard problems on the path to useful output are not, today, the physics — the physics is increasingly de-risked by a decade of public literature on transmon design, randomised benchmarking, and surface-code thresholds (the >99% two-qubit gate fidelity bar is now routinely cleared on small devices). The hard problems are:
- Calibration drift management — qubit frequencies, readout discriminators, and cross-resonance amplitudes drift on hour-to-day timescales. This is a continuous-deployment problem, not a physics problem.
- Compiler and transpiler discipline — mapping a chemistry ansatz onto heavy-hex with sensible SWAP overhead is a graph-problem and a software-supply-chain problem.
- Pulse-level scheduling under crosstalk constraints, expressed as a real-time scheduling problem on the control FPGAs.
- Observability — every shot, every calibration, every fridge telemetry channel has to be queryable. This is exactly the kind of system I built and broke and rebuilt at Oracle.
The Ireland quantum founder question is, at heart, whether you want this machine led by someone who has run national-scale production estates with hard SLAs, or by someone whose instinct is to publish first and stabilise later. I have made my choice. The Clonmel quantum founder position is engineer-first.
Why climate workloads first
The first cohort on Ireland Quantum 100 is climate science. That is a deliberate constraint, not a marketing line. Variational quantum eigensolver (VQE) and quantum phase estimation (QPE) workloads on 50–100 logical-equivalent qubits are exactly the regime where chemistry problems start to matter:
- Amine and metal-organic-framework sorbents for direct-air-capture, where binding-energy accuracy at chemical-accuracy thresholds (~1 kcal/mol) determines whether a candidate is worth synthesising.
- Perovskite and tandem photovoltaic cells, where defect-state energetics drive degradation pathways.
- Lithium-sulphur and solid-state battery cathodes, where polysulphide intermediates are notoriously hard for classical DFT.
- Nitrogenase and rubisco-class enzyme active sites relevant to fertiliser and atmospheric-CO₂ chemistry.
These are problems where the IMPT offset-stack and the Ireland Quantum 100 chemistry pipeline genuinely connect. A better sorbent screened on the quantum machine is, downstream, a better candidate for the IMPT supplier register. That is the only integration claim I will make on this page, because it is the only one I can defend in code.
The twelve-month delivery position
The schedule is public and I will be held to it:
- Q3 2026 — site fit-out at the Tipperary facility: vibration isolation, EMI shielding, helium recovery, three-phase power conditioning.
- Q4 2026 — cryostat installation, cooldown
Research collaboration or early access
Direct with Michael. No charge for the call.
Book a research call →