01Context
This is the first 90-day Quantum Litmus Readiness Over Time review.
The weekly Reality Check asks whether a new development changes the commercial outlook. The 90-day review asks a different question: after three months of accumulated evidence, what do we understand better about the path from technical progress to useful quantum computing?
That requires more than counting announcements. A result can be important even when it addresses only one part of the overall path to commercial readiness.
Quantum Litmus therefore looks at each major result through five questions:
- What was actually computed or tested?
- What level of evidence does it represent — component, system, algorithm, or operational?
- Was there a meaningful comparison with a classical method?
- What does the result tell us?
- What is the next validation step?
The purpose is not to discount technical milestones. It is to place each milestone correctly on the path toward practical readiness.
02Summary
The short answer
The last 90 days produced real progress.
Researchers improved protection against errors. Quantum processors ran some experiments faster. Software platforms took over more specialist setup work. Industrial manufacturing routes advanced. Independent evaluation of supplier roadmaps became more demanding.
The broad commercial-readiness estimate remains 2030–2033.
Most of the strongest evidence during the period still sits at the component and system levels. The next major validation step for the industry is to show those improvements carrying through into complete algorithms and then into operational workloads where reliability, scale, integration, time and cost all matter together.
The six-pillar scorecard
| Pillar | Rating | 90-day reading in plain English |
|---|---|---|
| Hardware Readiness | Early | Error protection, throughput and integration improved. The next step is dependable execution of larger complete workloads. |
| Software & Algorithms | Developing | Tooling improved and more of the workflow is becoming automated. The next step is repeatable application-level performance. |
| Validated Advantage | Not yet proven broadly | Several technical results are strong. The next step is a useful workload that compares favourably with the best practical classical method end-to-end. |
| Commercial Adoption | Early / Limited | Real deployments and institutional use continue. The next step is repeated customer outcomes with measurable value. |
| Economic Viability | Unproven | Cost visibility remains incomplete. The next step is a credible full-system cost and return case. |
| Ecosystem Readiness | Developing | Manufacturing, suppliers, access and external benchmarking improved. The next step is reliable delivery at scale. |
The executive reading
Prepare selectively. Test specific problems. Increase commitment as evidence moves from technical capability to complete system performance and repeatable business outcomes.
03What was actually computed?
This is the most important interpretive question in the review.
Quantum announcements often use similar language — “breakthrough”, “record”, “advantage”, “commercial deployment” — for very different kinds of evidence.
A useful way to read the field is to separate four levels.
| Evidence level | What is being tested? | Typical comparison | What it tells us |
|---|---|---|---|
| Component | A building block such as a qubit, gate, memory or decoder | Earlier device or internal baseline | Whether an underlying capability is improving |
| System | A processor or complete quantum subsystem | Earlier hardware, another platform or benchmark requirement | How much reliable computation the system can sustain |
| Algorithm | A defined computational problem | Ideally the best practical classical method | Whether the quantum method produces a useful result |
| Operational | A real workflow in use | Full end-to-end alternatives on quality, time and cost | Whether the result creates practical value |
Over the last 90 days, much of the strongest evidence was concentrated in the first two categories.
That is not surprising. Component and system improvements are what make more useful algorithms possible.
The important question now is whether those improvements begin to carry through into complete workloads, with results that can be compared on accuracy, runtime and cost with the best practical conventional alternative.
04Better error protection
One of the more important developments came from Quantinuum’s Helios error-correction work in October.
The work was primarily an error-correction experiment. The machine stored logical quantum information while varying the amount of protection around it. The reported error rate describes the reliability of the logical memory during an error-correction cycle. The work also included logical gate tests.
What was actually computed?
This was not a customer algorithm such as a chemistry calculation, portfolio optimisation or logistics problem.
The experiment addressed a more fundamental question: can quantum information be protected more effectively as additional error-correction resources are applied?
The reported result indicates that it can.
Why this matters
Useful quantum algorithms will require information to survive many operations without errors overwhelming the calculation. Better logical protection therefore strengthens confidence in one of the core engineering requirements for fault-tolerant computing.
Was there a classical comparison?
Not in the sense relevant to a business workload. There was no head-to-head comparison in which a quantum algorithm and a classical algorithm solved the same useful problem and their outputs, runtime and cost were compared. That is not what the experiment was designed to test.
The next validation step
The next step is to show that this level of reliability carries through into complete useful calculations, and continues to hold as those calculations become larger and more demanding.
That would move the evidence from component and system capability toward application-level readiness.
05Faster experiments
IBM’s Nighthawk r2 work addressed a different question.
IBM measured how quickly the processor could execute quantum circuits and reported more than 100,000 circuits per second — about 25 times the rate of the earlier Heron system for the tested setup, driven largely by a much faster reset path.
What was actually computed?
The machine ran quantum circuits — sequences of quantum operations designed to exercise the processor.
These tests measure how much quantum work the hardware can perform and how quickly researchers can repeat experiments. They are not equivalent to saying that a customer workload ran the same multiple faster.
Why this matters
Many quantum calculations require repeated circuit executions. Higher throughput can shorten experiment cycles, improve developer productivity and eventually reduce the time required to obtain statistically useful results.
Was there a classical comparison?
The principal comparison was with earlier IBM quantum hardware, rather than with a classical computer solving the same business problem.
The next validation step
The useful next question is end-to-end: when this higher circuit throughput is applied to a complete workload, how much does the total solution time improve after preparation, repeated runs, supporting computation and result checking are included?
That is where faster hardware begins to translate into application evidence.
06From technical progress to useful workloads
Better developer tooling
Quantinuum’s Nexus roadmap is primarily software-stack and workflow evidence.
The platform is intended to apply supported error-protection methods, test programs in simulation and run supported programs on current hardware. The significance is practical: tasks that once required specialist manual work are increasingly being built into the platform.
The next validation step is to show that these tools continue to reduce development effort as workloads become larger and hardware becomes more capable.
Industrial manufacturing
Quobly reported one- and two-qubit operations and readout on a silicon chip produced through an industrial semiconductor process. This is manufacturing evidence rather than application evidence.
The result matters because a useful quantum computer will eventually have to be manufactured consistently and at scale. A credible path through an established semiconductor process could become an important advantage.
The next validation step is detailed evidence on performance, yield, connectivity and scaling.
Cross-platform benchmarking
The QUOPS work used standardised circuits to measure the computational capability of several quantum systems and compared those measured capabilities with the estimated requirements of selected challenge problems.
This is closer to a system-level readiness measure than simply counting qubits.
It is not a direct quantum-versus-classical contest on a business problem. Instead, it asks how much reliable quantum computation current systems can sustain relative to what selected useful quantum problems are expected to require.
The next validation step is to see whether measures like QUOPS continue to track actual application performance as systems scale.
07Real-world use: what operational evidence tells us
Operational evidence deserves special attention because it sits closest to a business decision.
On 18 August 2026, NTT DOCOMO and D-Wave announced a second production application using D-Wave’s hybrid quantum technology in mobile-network planning. This was a new operational deployment during the 90-day review window, not a repeat of the earlier February application.
What was actually computed?
According to D-Wave’s October technical account, DOCOMO used D-Wave’s constrained quadratic model hybrid solver, combining classical computing with the Advantage2 annealing quantum computer, to optimise two competing network objectives at the same time: location-registration signals and paging signals.
For a representative problem spanning 333 base stations, three TA-Lists and nine tracking areas, the optimisation completed in about five minutes. DOCOMO reported a 65.3% reduction in peak location-registration signals and a 7.0% reduction in paging signals, with the results incorporated into live network operations.
Why this matters
This is meaningful operational evidence because DOCOMO moved from one production use case to a second, broader and more complex network-optimisation application. It shows repeat use of the technology in an operating environment rather than a one-off laboratory or proof-of-concept result.
Was there a classical comparison?
D-Wave reports that DOCOMO’s previous classical approaches could not solve the broader optimisation within a production-relevant time. The deployed solution, however, is a hybrid system: classical computation and the annealing processor work together.
For a buyer, the performance of the complete system is what matters operationally. For a readiness assessment, the next useful comparison is to quantify how much of the result comes specifically from the annealing component versus the surrounding classical optimisation.
The next validation step
The next validation step is a matched end-to-end comparison with the strongest practical classical alternative, including result quality, total runtime, integration effort, operating cost and repeatability, while also isolating the incremental contribution of the quantum component.
That would strengthen the case from successful hybrid operational use toward evidence of quantum-specific economic advantage.
Correction, 11 November 2026: An earlier version treated the August announcement as a repeat of the earlier application. It announced a distinct second production deployment and belongs within this review’s evidence window. The broad commercial-readiness estimate remains 2030–2033.
08Outside scrutiny, manufacturing and ecosystem progress
DARPA’s Quantum Benchmarking Initiative
DARPA advanced Atom Computing, Diraq, IBM and IonQ into Stage C of its Quantum Benchmarking Initiative in October.
This is not itself a computing result. It matters because it subjects supplier roadmaps and engineering assumptions to outside scrutiny.
The next validation step is the evidence that emerges from those evaluations: measured system performance, engineering feasibility and economic assumptions.
More hardware and infrastructure
Research institutions continue to buy, install and operate quantum systems. Procurement, installation, maintenance, user access and support are real parts of building an industry.
A supplier can build a meaningful infrastructure business before broad quantum advantage is established.
The next validation step is what those installed systems actually enable users to compute, how reliably they operate and whether the resulting work improves scientific or commercial outcomes.
No architecture has won
The 90-day evidence does not justify declaring one hardware approach the winner.
- Trapped ions lead in several published accuracy measures.
- Superconducting systems lead in supplier depth and operating ecosystem.
- Neutral atoms lead in array size and flexibility.
- Silicon spins offer compatibility with semiconductor manufacturing.
- Photonics offers networking and integration possibilities.
These are different kinds of progress. The relevant question is not which company has the largest headline number, but which architecture can eventually combine scale, accuracy, speed, manufacturability and economics in the same system.
09Readiness judgment and what would change it
The 90-day evidence strengthens confidence that several engineering problems are becoming more tractable.
It does not require a change to the broad commercial-readiness estimate.
| Industry stage | Early |
| Broad commercial-readiness estimate | 2030–2033 |
| Change from August 10 | None |
What improved
- Error protection became more convincing.
- Quantum circuit throughput improved.
- Developer workflows became more practical.
- Industrial manufacturing evidence strengthened.
- Cross-platform benchmarking became more informative.
- External scrutiny of supplier roadmaps increased.
- Operational hybrid use remains visible.
What this means
The industry is progressing through the expected sequence:
better components → stronger systems → useful algorithms → repeatable operational value
The evidence has moved forward within that sequence. The next phase of validation is to show that the technical gains survive when they are combined into complete workloads.
What would move the outlook earlier?
- A useful workload that compares favourably with the best practical classical method from start to finish.
- Independent confirmation of that result.
- Error-corrected operation sustained at realistic scale.
- Repeated evidence that quantum results improve real user decisions.
- A credible full-system cost and return case.
What could move it later?
- Persistent scaling difficulties.
- Excessive error-correction overhead.
- Poor operating reliability.
- Unexpectedly high costs.
- Rapid improvements in conventional methods that raise the performance threshold quantum systems must clear.
The same evidence-first standard applies in both directions.
The business conclusion
For executives, the right stance is neither to dismiss quantum computing nor to assume that technical progress automatically translates into near-term business value.
Prepare selectively, benchmark against strong conventional alternatives, and increase commitment as the evidence progresses from technical capability to system performance and repeatable business outcomes.
One strong result can justify a focused experiment. Moving the broad industry window requires several independent results showing that useful performance, dependable delivery and economics are coming together.
A Evidence framework and source notes
Quantum Litmus evidence progression
Component evidence
Does a building block work better? Typical examples: qubit fidelity, logical memory, gate accuracy, decoder performance and fabrication yield.
Next validation step: show that the improvement survives integration into the full system.
System evidence
Can the machine sustain useful amounts of reliable quantum computation? Typical examples: circuit depth, throughput, logical operations and cross-platform benchmarks.
Next validation step: run a defined useful algorithm with reliable end-to-end performance.
Algorithm evidence
Does a quantum method solve a defined problem well? Typical examples: chemistry, optimisation, simulation or machine-learning workloads.
Next validation step: compare against the best practical classical method on realistic data and operating constraints.
Operational evidence
Does the result create repeatable value in practice? Typical examples: production workflows, repeat customer use and measured business or scientific outcomes.
Next validation step: replication across customers, workloads or environments with full cost and performance data.
Source notes
Sources reviewed for this draft include Quantum Litmus Readiness Over Time and Q3 2026; Quantinuum’s October Helios error-correction account; IBM’s Nighthawk r2 throughput evidence; Quantinuum’s Nexus roadmap; DARPA’s October QBI Stage C update; Quobly’s silicon manufacturing evidence; the QUOPS benchmarking work; NTT DOCOMO’s production network-optimisation disclosure; D-Wave’s later account of that deployment; and Quantum Litmus weekly evidence through 6 November 2026.
Supplier releases are treated as primary accounts of supplier claims, not independent validation. Repeated coverage of the same result is not counted as independent confirmation.
The full task remains the unit of comparison. A faster quantum step may matter, but business readiness depends on the complete workflow: preparation, quantum execution, supporting classical computation, checking, integration and cost.
Staging note: this draft remains subject to final source reconciliation and editorial approval before production publication.