Companies — vendor landscape

Companies — vendor landscape

A quantum-networking workspace that doesn't track vendors loses touch with what can actually be procured. This page has two layers: a deep-dive dossier per vendor — milestones, roadmap, capability metrics, all cited to primary sources — and a static landscape snapshot for vendors where the dossier hasn't been written yet Ezratty 2025 Meddeb 2025 .

Vendors split naturally into three buckets: computers (build quantum processors), networking & QKD (build the network fabric or the QKD endpoints), and components (build the memories, sources, and other infrastructure that the other two consume).

Vendor dossiers

Each dossier below links to a per-vendor page with milestones, roadmap, and modality-specific capability details. Compiled from primary sources (papers, vendor technology pages); see each page for source links. Use the search box and category chips to narrow the grid.

Category
Country

Quantum computers

QKD

Quantum memory

Networking

Sources & detectors

Software & middleware

Enabling supply chain

QC systems and roadmaps

A snapshot of what's actually shipping today and what the leading vendors have committed to publicly. Logical-qubit numbers are 0 for NISQ machines; positive integers reflect publicly-demonstrated error-corrected logical qubits from a vendor announcement or a peer-reviewed paper. Roadmap rows carry the vendor's stated target year — these slip routinely, so treat them as direction rather than schedule.

Currently shipping or publicly demonstrated

Vendor Machine Year Physical Logical Modality
QuEra Aquila 2022 256 0 neutral-atom (analog)
Atom Computing Phoenix 2023 1180 0 neutral-atom
IonQ Forte Enterprise 2024 36 0 trapped-ion
Quantinuum H2 2024 56 0 trapped-ion
Rigetti Ankaa-3 2024 84 0 transmon
Google Willow 2024 105 1 transmon (research)
IBM Heron R2 2024 156 0 transmon
Origin Quantum Wukong 2024 72 0 transmon
CAS / QuantumCTek Tianyan-504 (Xiaohong) 2024 504 0 transmon
Microsoft Majorana 1 2025 8 0 topological (claims contested)
Xanadu Aurora 2025 12 0 photonic, networked
PsiQuantum Omega (chipset) 2025 — (chipset) 0 silicon-photonic (FBQC)
IonQ Tempo 2025 64 (AQ) 0 trapped-ion (Ba⁺, networked-cell)
Pasqal Orion Gamma 2025 140 0 neutral-atom
IBM Nighthawk 2025 120 0 transmon
USTC Zuchongzhi 3.0 2025 105 0 transmon (research)
Origin Quantum Wukong-180 2026 180 (+ 251 coupling) 0 transmon
Quantinuum Helios 2025 98 48 trapped-ion (Ba-137)

Public roadmaps

Vendor Machine Target Physical Logical Modality
IonQ Roadmap rung 2026 100–256 12 trapped-ion (Ba⁺)
IBM Kookaburra 2026 4 158 qLDPC memory demo transmon
QuEra Gemini-class 2026 ~3 000 30 neutral-atom
IonQ Roadmap rung 2027 10 000 800 trapped-ion (Ba⁺)
PsiQuantum Brisbane QCC 2027 utility-scale early FTQC silicon-photonic (FBQC)
Pasqal Vela 2027 200+ 20 neutral-atom
Quantinuum Sol 2027 192 tens trapped-ion
IonQ Roadmap rung 2028 20 000 1 600 trapped-ion (Ba⁺, photonic interconnect)
Pasqal Centaurus 2028 ~10 000 early FTQC neutral-atom
IBM Starling 2029 200 transmon (qLDPC)
Quantinuum Apollo 2029 thousands hundreds trapped-ion
IonQ Roadmap rung 2029 200 000 8 000 trapped-ion (Ba⁺)
PsiQuantum Million-qubit FTQC 2029 ~1 000 000 hundreds silicon-photonic (FBQC)
IonQ Roadmap rung 2030 2 000 000 80 000 trapped-ion (Ba⁺)
IBM Blue Jay 2033 2 000 transmon (qLDPC)

Each row is sourced from a vendor press release, peer-reviewed paper, or investor disclosure. Two patterns stand out: trapped ions lead on demonstrated logical-qubit count today (Quantinuum Helios is the only shipping product with publicly-verified error-corrected qubits), and the inflection to ~200 logical qubits at fault-tolerance is uniformly targeted for ~2029 across superconducting, trapped-ion, neutral-atom, and silicon-photonic roadmaps. IonQ's roadmap rungs and PsiQuantum's million-qubit Brisbane target are the most aggressive published; treat the implied delivery dates with the same caution as any vendor schedule.

Notes on this page

  • The landscape snapshot is a starting point. Names + platforms + one-line positions, drawn from survey literature. Goes stale within months.
  • Dossiers are the long-term tracking surface. Milestones, roadmap, capability metrics, and citations. Fields the dossier couldn't verify to a primary source are shown as unverified rather than filled in with a guess.
  • Source priority for dossiers: peer-reviewed paper > arXiv preprint > vendor blog or technology page > press release > industry tracker. Wikipedia is acceptable only for company-history facts, never for technical numbers.
  • Staleness. Each dossier records when it was last verified. Entries older than three months get a stale badge on the dossier page — a prompt to re-check the primary sources.

For survey-level reference on the per-platform vendor landscape, see Ezratty UQT Pt 2 Ezratty 2025 and Kumar 2025 IJNDC Kumar et al. 2025 .