Quantum Compute General
General-purpose quantum computing is a hardware count problem with a known threshold; the only numbers that matter are error-corrected logical qubits and logical error rate, and nothing a client can buy today clears either.
Clustered only. No lab work behind it. Cannot be cited.
Confidence
55%unresearchedExpiry
77duntil review · 19 Nov 2026Lead time
−14moopened after mainstream — recorded honestlyOwnership
Unownedcandidate — a named human electsWhere it is
The field is unusually measurable for something this far out: the roadmap is public, the physical-to-logical ratio is published, and every vendor announcement can be checked against the same two numbers. In 2026 the leading platforms demonstrated a few dozen logical qubits at error rates that make small error-corrected circuits real. Cryptographically relevant machines need on the order of a thousand logical qubits at rates two orders better. The commercial applications short of that — chemistry, optimisation — have not beaten classical methods on any problem a client has. We keep this open because Quantum Encryption depends on it and because an executive will ask.
Why a Quantium decision hinges on it
Two reasons, neither of them 'do quantum'. First, the PQC migration question in banking hinges on the date this field crosses threshold, and that date is the most-asked crystal-ball question in the vertical. Second, the firm gets pitched quantum-optimisation for logistics and pricing every quarter and needs a fast, evidenced no. Being able to say 'here are the two numbers, here is where they are' is the entire value of holding the field.
Field attributes
Position
What is demonstrated, what is hype, what would have to be true.
The shape every position request answers. Signal-tier fields carry a draft; assessed and tested fields carry a validated one.
- 01Below-threshold error correction: logical error rate falls as code distance grows, on at least two hardware platforms.
- 02Tens of logical qubits running small error-corrected circuits for hours, with published error rates.
- 03Quantum-optimisation vendors have not beaten a tuned classical solver on any logistics or portfolio problem a client has brought us.
- 01Physical qubit counts. A thousand noisy physical qubits is not a thousand qubits of anything.
- 02'Quantum advantage' on sampling problems nobody wanted solved.
- 03Quantum-inspired optimisation, which is classical annealing with a marketing budget.
- 01A public demonstration of ≥100 logical qubits at logical error ≤1e-6, on a roadmap that credibly reaches ~1,000.
- 02A chemistry or optimisation result on an industrially relevant problem that a classical method could not reproduce within a week on a cluster.
- 03Cloud access with a pricing model an AU client could put in a budget, under AU data-residency terms.
- 01Maintain the two-number tracker and answer the timeline question from it (sa-pqc-timeline lives under Quantum Encryption).
- 02Decline every optimisation pitch that does not come with a classical baseline; log each as a signal.
- 03If the logical-qubit trigger fires: re-run the PQC timeline position within 48 hours, because that is the field that moves.
Signals · 10 in this cluster
What the cluster is made of.
Every item carries its source, tier and sightings. Detector-found signal sits beside human drops; downstream they are indistinguishable except by provenance.
Leading platform reports 48 logical qubits at 1e-4 logical error, sustained for hours
Largest error-corrected demonstration to date; numbers were independently checked by a university group within the month. This is the datapoint on the tracker. Still 20× short on count and 100× short on error rate for cryptographic relevance.
extracted claimTens of logical qubits at 1e-4 error are now demonstrated, not projected.
'When do we actually have to move our crypto?'
Asked by a CISO's office; third time this year in banking. Routed to sa-pqc-timeline. Logged here because the answer is derived from this field's two numbers.
Error-Corrected Simulation of a Catalyst Active Site Beyond Classical Reach — Partially
Claims a chemistry result classical methods approximate poorly. A classical group matched it within three weeks using a new tensor-network method. Not the crossing, but the closest so far, and the pattern — quantum claim, classical rebuttal — is the one to watch.
Pharma major hiring 'Quantum Chemistry Applications Lead' in Melbourne
First AU-based quantum-applications role we have seen outside a university. Argus inference: at least one pharma company treats chemistry-grade advantage as near enough to staff for.
ASD ISM update: PQC migration guidance with indicative 2030 milestones
Guidance, not mandate, but the first AU document with dates attached. Argus reads it as the precursor to a hard deadline; APRA typically follows ASD on crypto by 12–24 months.
Resource Estimates for RSA-2048 Under Current Overheads, Revised
Revises the physical-qubit requirement for breaking RSA-2048 downward by ~4× on newer codes. The number to watch: it has fallen an order of magnitude in eighteen months, which is the red team's whole argument.
extracted claimPhysical-qubit requirements for cryptographic relevance keep falling faster than roadmaps assume.
Logged from Claude Code: vendor QUBO demo on our route-optimisation case lost to OR-Tools
Reproduced a quantum-optimisation vendor's demo on a delivery-routing instance from a retail engagement. OR-Tools found a better solution in 40 seconds on a laptop. Tried tier; the standing answer to every optimisation pitch since.
'Quantum advantage for portfolio optimisation is here'
Vendor post claiming advantage on a 50-asset portfolio problem. No classical baseline reported. Kept as the pitch pattern; contradicted by our own reproduction.
Below-Threshold Surface Code Memory on a Second Hardware Platform
Logical error rate halves per code-distance step on a trapped-ion platform, replicating the superconducting result. Two platforms below threshold means error correction is a property of the theory, not one vendor's chip.
extracted claimError correction below threshold is demonstrated on more than one hardware platform.
Analyst timeline: cryptographically relevant quantum computer 'most likely 2032–2035'
Consensus-style estimate. Consistent with our own reading but the red team notes every such estimate since 2019 has been revised earlier, not later.
Claims · 4 supporting, 1 refuting
The atoms.
A document cannot go stale; an assertion can. Claims are immutable and stamped with the extractor that produced them, so staleness, diffs and the graveyard operate at claim level.
Logical error rate now falls with code distance on at least two hardware platforms; error correction is demonstrated, not theoretical.
A cryptographically relevant machine needs roughly 20–50× the demonstrated logical qubit count at 100× better error rate; on published roadmaps that is not before 2031.
AU regulators will set a PQC deadline before the compute threshold is crossed, so the client-facing timeline is regulatory, not physical.
Quantum chemistry is the first application to cross classical parity, and it will do so before optimisation.
Quantum optimisation delivers advantage on logistics and portfolio problems today.
Position history · the diff is the product
2 validation runs against a fixed brief. Confidence 50% → 55%.
Logical-qubit counts up, error rates down, both on schedule. The client-facing date is regulatory, not physical — a regulator will set a PQC deadline first. Overhead reductions are the risk to our estimate.
- Quantum chemistry is the first application to cross classical parity, and it will do so before optimisation.
- AU regulators will set a PQC deadline before the compute threshold is crossed, so the client-facing timeline is regulatory, not physical.
- c-quantum-compute-2 ↓ 0.72 → 0.66
Scoring · ordinal bands
Agents propose. A named human commits.
Uncommitted scores are visibly marked and never leave the building. Bands, not point estimates — false precision is the tell that a number was generated rather than derived.
Impact
committed · AWCrossing threshold breaks the cryptography every client runs on. Nothing else in the graph has that shape.
Timeline
committed · PRTwo numbers, both public, both two orders of magnitude short of cryptographic relevance.
TAM
agent-estimatedAgent-estimated for the consultancy-addressable slice (readiness, migration). Hardware TAM is not ours.
Cost
committed · PRTracking two numbers quarterly is an afternoon.
Cost of being wrong
committed · LFBeing late on the threshold date means clients migrate crypto late. Harvest-now-decrypt-later makes late unrecoverable.
Demand
committed · CDAsked as a timeline question in banking; nobody asks for quantum compute itself. Optimisation pitches arrive from vendors, not clients.
Workforce readiness
agent-estimatedOne engineer can read the papers; nobody can run a circuit. Agent-estimated.
Relevance · per vertical
Why it matters here, or explicitly does not.
Ranking is per vertical, not global. Sector owners commit notes against agent drafts.
The only banking relevance is the PQC migration date, which this field's two numbers determine.
Mechanism · Logical-qubit trigger fires → sa-pqc-timeline re-issued → CPS 234 crypto inventory work brought forward.
ASD's ISM already carries PQC guidance; agencies will need a migration plan regardless of the hardware date.
Mechanism · Field supplies the evidence base for 'how urgent'; the work itself sits under Quantum Encryption.
Every quantum-optimisation pitch for range or logistics has lost to a classical solver. No retail mechanism until chemistry-grade advantage arrives, which is not retail.
Mechanism · None.
Quantum chemistry is the first plausible commercial application; relevant only if the firm enters pharma, which is a sponsor question.
Mechanism · Molecular simulation as a service to drug-development clients.
Red team · the strongest case against
The strongest case against: our two-number framing assumes the surface code and its overheads. Newer codes and hardware-specific tricks have cut physical-to-logical ratios by an order of magnitude in eighteen months, and a further cut would collapse our 2031 estimate to 2028. We are tracking the right numbers with the wrong conversion factor.
- —Physical-to-logical overhead has fallen faster than any roadmap predicted; extrapolating today's ratio is the same mistake every quantum-sceptic timeline has made.
- —Harvest-now-decrypt-later means the client-relevant date is not the threshold date but several years before it; our timeline band may be correct and still too late to act on.
- —Chemistry-grade advantage does not need a cryptographically relevant machine; the first commercial crossing may arrive with 200 logical qubits and we are not positioned in the vertical that benefits.
Source diversity
- Quantum hardware research40%
- Vendor / analyst20%
- Regulatory15%
- Internal / Engel25%
A field supported by one epistemic community is a flag, not a finding.
Cross-pollination · typed joins
Connected, not merely similar.
Enabling, compounding, substituting, blocking. A satisfied dependency trigger is a far stronger signal than semantic proximity.
Trigger · APRA, ASD or the RBA publishes a hard PQC migration deadline for AU ADIs or critical infrastructure. That is the moment clients ask about the compute side rather than the crypto side, and this candidate goes to election.
When the trigger fires, this field is resurfaced automatically. Watchable rather than parked.
Trigger · A peer-reviewed molecular-simulation result on an industrially relevant molecule that a classical method cannot reproduce within a week on a commodity cluster, with the classical baseline published; Argus flags it from any of the three leading hardware platforms.
When the trigger fires, this field is resurfaced automatically. Watchable rather than parked.
The PQC field is gated on this one; its timeline is derived from our two numbers.
Harvest-now-decrypt-later is a state-actor behaviour; the threat model is theirs, the date is ours.
Quantum chemistry is the first plausible commercial crossing and pharma is where it lands.
Share graph
Provenance running forward.
Discovery, not accountability. No counts, no rankings, no rollups to managers.
Convergence · who else is here
- PRPriya Raman · Research engineer · inference3 drops
- CDClaire Dubois · Sector owner · Banking2 drops
- LFLena Fischer · Red team & assurance1 drop
- HNHana Novak · Sector owner · Health1 drop
- ?Anonymous · Anonymous drop1 drop
- ABAisha Bello · Sector owner · Government1 drop
Several people’s drops meet here. An informal working group already exists and probably does not know it.
Lineage
What this field produced, and what it killed.
Experiments, recommendations and graveyard entries stay attached. The reasoning that killed a claim is the reusable asset.
No experiments, recommendations or graveyard entries yet. That is what a candidate looks like.
Open questions · return to the pile
Every run leaves a record. Separately, its question either closes or returns to the pile with notes — which is what the next person proposing the same thing will see.
- 01What conversion factor from physical to logical qubits should the tracker use, given it has fallen 10× in eighteen months?
- 02Does harvest-now-decrypt-later mean the client action date is 2027 regardless of when the machine exists?
- 03Is there any optimisation problem in a served vertical where a quantum method has a credible path, or should the standing 'no' be permanent?