The Landscape in 2026

Picture the quantum-computing ecosystem in 2026 as a single city seen from the air. You would see three broad districts — industry along one axis, academia along another, national programmes as the roads that connect them — and four big capitals spread across continents, each with its own accent on what matters. Zoom in and you find a hundred smaller neighbourhoods: startups in Bangalore, in Boston, in Paris; research groups in Caltech, Delft, USTC, IIT Madras; corporate labs in Yorktown Heights, Mountain View, and Oxford.

You have come through the curriculum and by now you know what a qubit is, what NISQ means, what a logical qubit is in practice, and the platform-by-platform state of the art in early 2026. What you deserve next is the map — not the hardware numbers, but the ecosystem. Who is working on this, who is paying for it, where they sit, and what each of them is actually trying to accomplish. That is this chapter.

The answer, in one sentence: as of April 2026 the field has about ten major hardware companies, dozens of serious academic groups, four national programmes with combined public funding near \sim30B across their respective decades, and more than a hundred startups — enough structure to call it a real industry, not yet enough useful output to call it a mature one.

The three layers

Three layers, but they are not isolated. The major companies have co-authored papers with academic groups (Google and Caltech, Quantinuum and Microsoft, IBM and dozens of IIT partnerships). The national programmes fund both industry and academia — NQII grants flow to both IBM and MIT, the NQM funds both QpiAI and IIT Madras. A realistic mental model is three overlapping circles, not a stack.

The ten major hardware companies

This is where the money is and where the largest machines are being built. Eight of the ten are American or US-UK; one is Canadian (Xanadu); three are European (Pasqal, IQM Finland, and parts of Quantinuum UK-US). China's hardware programme sits inside state labs rather than independent companies, which is discussed separately below.

IBM — the broadest commercial footprint

IBM Quantum runs the largest public cloud for real quantum hardware (IBM Quantum Platform). Condor is the 1121-qubit qubit-count flagship (2023); Heron is the 133-qubit workhorse chip with dynamic circuits — mid-circuit measurement with classical feedback, the prerequisite for error correction — that customers actually run algorithms on in 2026. The declared roadmap targets the Starling machine with 200 logical qubits by 2029.

IBM's commercial model is access-as-a-service: hundreds of Indian, European, Japanese, and US universities access Heron through the IBM Quantum Network; IIT Madras, IIT Kanpur, TIFR, and IISc are all members. This is, in 2026, the single largest on-ramp for working scientists to write and run actual quantum code on real hardware.

Google — the first below-threshold qubit

Willow (105 superconducting qubits, December 2024) crossed the surface-code threshold and produced the first on-hardware demonstration of a logical qubit that got better as you gave it more physical qubits. See the state-of-the-art chapter for the protocol in detail. Google publishes less and demonstrates more than IBM; there is no equivalent of IBM Quantum Platform for Google hardware (access is via collaboration or Google Cloud's limited Cirq backend).

Quantinuum — trapped ions and logical gates

Quantinuum — UK-US company from the 2021 merger of Honeywell's ion-trap group with Cambridge Quantum's software arm — runs the H-series of trapped-ion machines. H2 holds 56-64 ytterbium or barium ions with two-qubit gate error around 5 \times 10^{-4}, ten times better than the best superconducting platforms. In 2024, jointly with Microsoft, they demonstrated four logical qubits on a colour code with fault-tolerant transversal Clifford gates — the first end-to-end logical-qubit computation with a non-trivial gate count.

IonQ — the other trapped-ion commercial player

IonQ runs Forte (35 algorithmic qubits) and Tempo (64-qubit-target, rolling out 2024-25). Its commercial model is AWS Braket and Azure Quantum cloud access — on the menu alongside Rigetti and OQC — plus on-premises systems for national laboratories. IonQ's Algorithmic Qubits benchmark (a quantum-volume variant) is its main marketing metric.

Atom Computing — the biggest single array

Atom Computing trapped 1180 caesium atoms in a single optical-tweezer array in November 2024, as of early 2026 the largest programmable-qubit array on any platform. The two-qubit error is around 5 \times 10^{-3} — competitive but not yet in the Quantinuum regime. Mid-circuit measurement on neutral atoms was the genuinely hard thing they solved; it is the prerequisite for dynamic-circuit error correction on this platform.

PsiQuantum — the silicon-photonic bet

PsiQuantum is not building NISQ machines. Their explicit strategy is to skip NISQ and build a million-photonic-qubit fault-tolerant machine by the early 2030s. Omega — the silicon-photonic chip they taped out at GlobalFoundries Fab 10 in late 2024 — is the building block: a wafer-scale integration of single-photon sources, beamsplitters, phase shifters, and superconducting-nanowire single-photon detectors. No interim NISQ demos are planned.

Xanadu — continuous-variable photonics

Xanadu (Toronto) pursues a different photonic route: continuous-variable (CV) encoding in the quadratures of squeezed light. Their 2022 Borealis machine demonstrated Gaussian boson sampling with 216 modes. Their software stack (Strawberry Fields, PennyLane) is one of the most-used toolchains for variational quantum machine learning, regardless of which hardware the algorithm eventually runs on.

Microsoft — Azure Quantum and the topological bet

Microsoft is simultaneously a cloud aggregator (Azure Quantum lets you access IonQ, Quantinuum, Rigetti, Pasqal, and others) and a hardware contender with the topological-qubit Majorana 1 prototype announced in February 2025. The physics claims of Majorana 1 remain scientifically contested in early 2026; if they hold up, topological qubits could dramatically shorten the road to fault tolerance; if they don't, Microsoft retains the Azure-Quantum software-and-aggregation business and its partnerships (especially with Quantinuum).

QuEra and Pasqal — the neutral-atom duopoly

QuEra (Boston) and Pasqal (Paris) both run rubidium neutral-atom arrays. Their distinctive feature is analog-plus-digital operation: the same machine can run continuous-time Rydberg-atom Hamiltonian evolution (ideal for combinatorial-optimisation ansätze and condensed-matter simulation) or discrete-gate circuits. QuEra Aquila holds 256 atoms; Pasqal pushed past 300 in mid-2025. Pasqal is the strongest European hardware player; EU Quantum Flagship money has been a material part of its story.

The academic leaders

Dozens of academic groups do serious quantum-computing research. The following are the most load-bearing in the sense that a paper out of their lab materially changes the field.

Caltech — Preskill and the theory capital

John Preskill coined the NISQ label in 2018. The Caltech Institute for Quantum Information and Matter (IQIM) is the theory capital of the field — quantum error correction, quantum complexity, quantum gravity. If you read one set of free lecture notes on quantum information, read Preskill's. They have been the curriculum for a generation of quantum-information students, including many of the authors of this wiki.

Harvard-MIT — the neutral-atom powerhouse

Mikhail Lukin at Harvard and Vladan Vuletic and Ike Chuang at MIT have built the academic neutral-atom programme that QuEra spun out of. A second spinout — dozens of labs worldwide now work on neutral-atom quantum information because of this lineage.

Stanford — photonics and quantum information theory

Monroe (previously at Maryland, now partly at Duke), Kimble (retired, Caltech), and Vuckovic at Stanford lie at the heart of a long-running US academic photonics programme. Stanford also hosts the Q-NEXT National QIS Research Center under the NQII.

USTC — China's academic flagship

The University of Science and Technology of China in Hefei, led for quantum by Pan Jianwei, has produced the Chinese Academy of Sciences' flagship results: Micius (the 2016 QKD satellite), Jiuzhang (the photonic boson-sampler that in 2020-2023 repeatedly raised the supremacy bar), and Zuchongzhi (superconducting-qubit platforms). USTC is a state laboratory in function; its budget is bundled into CAS and the broader Chinese quantum programme.

TU Delft — spin qubits and quantum networks

QuTech at TU Delft (co-located with TNO) leads on silicon-and-diamond-spin-qubit platforms and on the early Quantum Internet stack. Stephanie Wehner's programme on quantum-network protocols is the most-cited academic work on the networking side.

IQC Waterloo — the Canadian hub

The Institute for Quantum Computing at the University of Waterloo is the oldest large-scale academic quantum-information programme in North America. Faculty there have spun out Xanadu; Canadian quantum computing runs largely through IQC.

Oxford, ETH Zurich, and the European academic map

Oxford's ion-trap and photonics groups, ETH Zurich's superconducting programme, and a dozen Max-Planck and German university groups make up the European academic spine. The EU Quantum Flagship distributes funding across them.

India — IITs, IISc, TIFR, RRI

IIT Madras leads superconducting hardware (in partnership with TCS under the NQM). IISc Bangalore and TIFR Mumbai lead trapped-ion work. The Raman Research Institute in Bangalore leads quantum optics. IIT Delhi, IIT Bombay, and IIT Kanpur have strong theory and software programmes. The NQM's Quantum Computing T-Hub coordinates across these.

The four national programmes

USA — the National Quantum Initiative Act

The National Quantum Initiative Act was signed into US law in December 2018. It authorised a ten-year coordinated federal programme across the Department of Energy, National Science Foundation, NIST, and DARPA, with an initial public budget of 1.275 billion over the first five years and annual reauthorisations since. By 2025 US federal quantum spending exceeded **1.2 billion per year**, with cumulative funding since 2018 near $8 billion. The programme structure is:

• Five QIS Research Centers funded by DOE: Q-NEXT (Argonne), QSC (Oak Ridge), C2QA (Brookhaven), SQMS (Fermilab), and QSA (Lawrence Berkeley). Each runs at $25-30M per year and coordinates a consortium of universities and national laboratories.
• NSF Quantum Leap Challenge Institutes at Illinois, Colorado, Maryland, Berkeley, and elsewhere — academic-heavy centres focused on training, algorithms, and materials.
• NIST runs the post-quantum-cryptography standardisation programme (NIST PQC, which produced CRYSTALS-Kyber, Dilithium, and SPHINCS+ in 2024). This is arguably the NQII's single most commercially consequential output so far.
• DARPA runs a set of programmes (Underexplored Systems for Utility-Scale Quantum Computing, Quantum Benchmarking) that specifically probe whether current quantum-computing claims are real and useful.

Industry contracts with NQII money flow to IBM, Google, Quantinuum, IonQ, Rigetti, Atom Computing, QuEra, and PsiQuantum. Willow was partly enabled by a long-running Google-Caltech collaboration that sits inside the broader NQII academic ecosystem.

China — the CAS quantum programme and Hefei Laboratory

China's quantum programme is embedded in the Chinese Academy of Sciences, centred at the University of Science and Technology of China (USTC) in Hefei and at the Hefei National Laboratory for Physical Sciences at the Microscale. Public disclosure is lower than the other programmes; estimates of total funding vary from 10B to25B across 2016-2030 depending on what is counted (pure-science, hardware, communications infrastructure). The headline outputs:

• Micius (Quantum Experiments at Space Scale satellite, launched 2016): the first space-to-ground QKD demonstration. Landmark result in quantum communication.
• Jiuzhang series of photonic boson-samplers (2020, 2021, 2023): repeated supremacy-style benchmarks that pushed the classical-simulation frontier hard.
• Zuchongzhi series of superconducting processors (2021-2024): China's superconducting-hardware line, comparable in qubit count and fidelity to the US mid-tier platforms.
• Beijing-Shanghai and Beijing-Hefei QKD backbones: longest deployed fibre QKD networks in the world as of 2023-24.

EU — the Quantum Flagship

The EU Quantum Flagship was launched in 2018 as a ten-year, €1B programme (matched by member-state contributions for an effective total near €2-3B). It funds four pillars — quantum computing, quantum simulation, quantum communication, quantum sensing — through consortium projects across universities, research institutes, and industry. The Flagship is the reason Pasqal (France), IQM (Finland), and the Quantum Internet Alliance (Netherlands-led) exist at their current scale. The programme's successor is already being drafted for 2028-onwards.

India — the National Quantum Mission

The National Quantum Mission (NQM) was approved by the Indian Union Cabinet in April 2023 with a budget of ₹6003 crore over 2023-2031 — about 720 million at launch exchange rates, roughly800 million at mid-2026 rates. Four Technology Hubs (T-Hubs), each focused on one vertical:

• T-Hub 1 — Quantum Computing: IIT Madras (lead), with TIFR Mumbai, IISc Bangalore, IIT Delhi, IIT Bombay as partners. Target for Phase 1 (2023-27): 50-100 physical qubits across superconducting, trapped-ion, photonic, and neutral-atom platforms. Target for Phase 2 (2027-31): 500-1000 physical qubits and the first error-correction demonstrations on Indian-built hardware.
• T-Hub 2 — Quantum Communication: ISRO (lead), IIT Delhi, Raman Research Institute, IIT Kanpur. Ground-based and satellite QKD.
• T-Hub 3 — Quantum Sensing and Metrology: IIT Bombay (lead), TIFR, BARC, PRL Ahmedabad. Atomic clocks, magnetometers, gravimeters.
• T-Hub 4 — Quantum Materials and Devices: IISc Bangalore (lead), IIT Delhi, IIT Kanpur, various CSIR labs. Topological materials, single-photon sources, semiconductor-spin-qubit fabrication.

India is not trying to out-spend the US or China on hardware; the NQM's sovereign-capability thesis is to build domestic hardware at moderate scale, domestic software across the stack, and domestic expertise in parallel with cloud access to global platforms. By 2031 the target is 1000-qubit-class superconducting hardware at IIT Madras, 100-qubit-class domestic platforms on trapped ion and photonics, deployed QKD networks in government and strategic sectors, and a trained workforce of several thousand quantum-specialist scientists and engineers.

The startup ecosystem

On top of the ten major companies and the big academic groups sits a layer of more than a hundred startups. A complete catalogue would be out of date by the time you finished reading it; what matters is the categories.

• Hardware startups (non-major): Rigetti (superconducting, US), SandboxAQ (US, spin-out of Alphabet), Quantum Machines (Israel, control electronics), Zurich Instruments (Switzerland, control electronics), Silicon Quantum Computing (Australia), QpiAI (India — see below).
• Software and algorithms: Zapata (US), Classiq (Israel, quantum-software compilation), Multiverse Computing (Spain, quantum-inspired classical + quantum finance), Horizon Quantum (Singapore).
• Cryptography and security: IDQ (Switzerland, QKD hardware), Toshiba Quantum (Japan, QKD), QNu Labs (India).
• Applications: Menten AI (drug discovery), Aqemia (pharma), ProteinQure (biotech), Quantinuum's applications division.

The Indian startup trio

Three Indian quantum startups sit at the centre of the NQM commercialisation story.

QpiAI (Bangalore, founded 2019) is building superconducting-qubit hardware domestically while running a commercial classical-ML-plus-quantum-optimisation platform. They raised Series A funding in 2024 and partner with IISc and IIT Bangalore on fabrication. QpiAI's ambition is to be India's first end-to-end domestic hardware company.

BosonQ PSI (Bhilai, Chhattisgarh, founded 2020) focuses on quantum-optimisation for engineering simulation — computational fluid dynamics, structural mechanics, topology optimisation — using hybrid classical-quantum ansätze on NISQ hardware. Their customers are aerospace and heavy-industry engineering firms who want to evaluate whether quantum-enhanced CFD is faster or more accurate than pure classical pipelines.

QNu Labs (Bangalore, founded 2016) builds and ships quantum key distribution hardware and post-quantum-cryptography products for Indian enterprise and defence customers. Their QKD devices have been deployed in banking and defence pilots in 2024-25. QNu Labs is probably the most commercially mature of the three because QKD is a product that works today — the quantum advantage for QKD is information-theoretic (no assumed hardness), not speed.

Alongside these three sit perhaps 20-30 smaller Indian quantum-software and quantum-consulting firms, mostly based in Bangalore, Hyderabad, Pune, and the NCR.

The worked examples

Common confusions

• "The ecosystem is just the ten hardware companies." It is not. The software stack (Qiskit, Cirq, PennyLane, Strawberry Fields, Classiq), the algorithms researchers (academic + industrial), the cloud platforms (IBM Quantum, Azure Quantum, AWS Braket), the cryogenics and control-electronics vendors (Bluefors, Oxford Instruments, Quantum Machines), and the hundred-plus startups all matter. A machine without a stack and a community is inert.
• "National programmes are competing." At the level of prestige and soft power, perhaps. At the level of science, collaboration is routine. Indian students work at Caltech; Chinese students publish in Nature with European co-authors; US researchers attend Chinese conferences; Indian industrial partners buy IBM and Quantinuum machines. The wiring is global even when the headlines are national.
• "The NQM is insufficient compared to the NQII." On raw dollar terms, yes. On sovereign-capability terms, the NQM is the right scale for its stated mission. Indian quantum computing is not trying to build bigger machines than Google; it is trying to build serious machines domestically, train a workforce, and position Indian startups for the 2030s commercialisation wave.
• "The hundred startups are all unicorns." Most are not. Many are sub-Series-A research shops; some are defence-contractor subsidiaries; a handful have raised serious venture capital (PsiQuantum's $650M in 2021 is the record). The mortality rate in the startup layer will be high over the 2026-2030 window; expect consolidation.
• "If you don't work at one of the ten companies, you can't contribute." False. Academic groups produce most of the theoretical and algorithmic progress; national-laboratory teams run some of the most interesting hardware experiments; undergraduate and graduate students at IITs and IISc have published first-author quantum-information papers. The hardware gets built by a handful of companies; the field is much bigger than the builders.
• "Quantum communications is separate from quantum computing." Organisationally, often yes — the NQM has separate T-Hubs, the Chinese programme has separate Micius and Jiuzhang lines. Technologically, they converge at the hardware level (single-photon sources and detectors serve both; quantum memories are needed for both; error correction applies to both). A complete picture of the 2026 ecosystem must include QKD, the quantum-internet roadmaps, and the post-quantum-cryptography migration — all of which are covered in adjacent chapters of this curriculum.

Going deeper

Cross-border collaborations

The Willow team included Caltech (Preskill, Alex Kubica), Harvard, and MIT academics alongside Google engineers. The Quantinuum-Microsoft colour-code demonstration pooled Quantinuum's hardware, Microsoft's Azure Quantum software stack, and academic theorists from multiple institutions. The Micius QKD satellite collaborated with Austrian groups (Zeilinger's team) for international QKD experiments. The NQM formally coordinates with NQII and EU Flagship via academic exchanges and joint workshops.

The pattern: academia is the glue. A hardware company cannot hire the theory all-stars it needs full-time; an academic group cannot afford hardware at scale. The natural structure is long-running industry-academia collaborations with joint papers.

How the money actually moves

National programme → academia: direct grants, typically 3-5 year competitive awards. Academic PIs hire PhD students and postdocs; a fraction of the grant pays for hardware time on industry platforms.

National programme → industry: contracts (DOE SBIR/STTR, DARPA programme money, NQM hardware-build awards), tax credits, and co-funded consortiums. In India, the NQM funds hardware-build projects at IIT Madras that subcontract fabrication to QpiAI and other domestic partners.

Private capital → startups: venture rounds. PsiQuantum has raised 1B+; Atom Computing and Quantinuum each300M+; QuEra, Pasqal, IonQ (now public), Rigetti (public) in the 100-500M range. Indian startups have raised smaller Series A rounds so far (single-digit to ~50M), consistent with their stage.

Customer revenue → companies: IBM Quantum Network subscriptions from universities and enterprises; Azure Quantum, AWS Braket, and Google Cloud compute-time charges; QKD hardware sales (QNu Labs, IDQ, Toshiba); consulting and custom-algorithm work. Total global 2025 revenue for the field is estimated at 500M-1B — small compared to the funding.

The 2030 projection

Extrapolating the 2024-2026 rate forward plausibly to 2030:

• Hardware consolidation: expect 5-8 major hardware companies by 2030, down from 10-12. The smaller NISQ players without a path to fault tolerance will merge or exit.
• First useful fault-tolerant demos: 2028-2030 on small-molecule chemistry (H_2, LiH, BeH_2) and small lattice-gauge-theory simulations. 10-100 logical qubits, running 10^6-10^9 logical gates.
• Post-quantum-cryptography migration: near-complete for new-deployment government and banking systems globally, partial for legacy infrastructure. Indian Aadhaar, UPI, and banking sectors on CRYSTALS-Kyber and Dilithium for new systems by ~2028-2030.
• NQM phase transition: India's NQM enters Phase 2 in 2027; by 2030-2031 the programme should have 500-1000 domestic physical qubits on superconducting hardware and 100-qubit-class trapped-ion and photonic platforms operating.
• Chinese programme: Micius II (a proposed successor QKD satellite) likely operational; Jiuzhang-class photonic samplers pushed past 200 photons; superconducting-qubit platforms in the 1000-5000 physical qubit range.
• Startups: expect a handful of acquisitions (major companies buying specialised capability) and a handful of IPOs (for the companies with mature commercial revenue).

For the Indian reader considering this area

If you are a 15-year-old reading this in India in 2026, here is the calibrated advice. The NQM is real. The IIT and IISc groups are hiring, training, and producing first-rate students. Indian quantum-computing PhDs are employable globally and at home. The three domestic startups — QpiAI, BosonQ PSI, QNu Labs — will plausibly employ hundreds to thousands of engineers each by 2030. The global companies (IBM, Google, Quantinuum, IonQ) have active Indian offices and collaborations.

The route in is: strong linear algebra and quantum mechanics in undergraduate study; internships at IIT/IISc/TIFR groups during undergrad; a masters or PhD in quantum information or quantum engineering; then either academic track (IITs/RRIs/TIFR faculty, postdocs abroad and back) or industrial track (NQM-funded startup or global company's India team). The field will hire across both tracks for the next two decades. The commercialisation wave is coming in 2028-2035; the training takes 6-10 years; if you start now, you graduate into the right decade.

Where this leads next

• The Logical Qubit Era — the next chapter, framing the NISQ-to-fault-tolerant transition as a phase change rather than a gradual transition.
• Utility-Scale Quantum Computing — what "utility" actually means in 2026 and how to distinguish it from advantage.
• State of the Art 2026 — the platform-by-platform, number-by-number companion to this chapter.
• What NISQ Means — the regime most 2026 hardware is still in.
• Logical Qubits in Practice — the protocol that made Willow a milestone.

References

1. Government of India, National Quantum Mission (DST) — dst.gov.in/national-quantum-mission.
2. US Congress, National Quantum Initiative Act (2018, reauthorised 2024) — quantum.gov.
3. European Commission, Quantum Flagship — qt.eu.
4. John Preskill, Quantum Computing in the NISQ era and beyond (2018) — arXiv:1801.00862.
5. Wikipedia, List of companies involved in quantum computing or communication.
6. Google Quantum AI, Quantum error correction below the surface code threshold (2024) — arXiv:2408.13687.