PsiQuantum Corp.
Capital-intensive deep-tech hardware: builds and operates utility-scale fault-tolerant quantum computers; revenue model is pre-commercial today (government R&D contracts, deployment partnerships, future QCaaS / system sales), funded primarily by venture equity + sovereign-government capital (equity/grants/loans).
Quantum-computing pre-revenue; only 3 priced points have a disclosed valuation. Series C (Nov 2019) and seed valuations were never disclosed (omitted); a ~$10.5B May-2026 mark appears on aggregators (Sacra/Tracxn) but is uncorroborated by primary news, so omitted. Mar-2025 $6B is Reuters 'sources say' pre-announcement.
Earnings, margins, COGS & capex
Revenue trend
Margins
Bull & bear
If photonic fusion-based architecture scales, PsiQuantum is the only pure-play taking a credible, manufacturing-first run straight at utility-scale fault tolerance (~1M qubits) — backed by a tier-1 fab (GlobalFoundries), Nature-validated chips, and sovereign balance sheets in two countries.
- Manufacturing-first edge: the Omega chipset (Nature, Feb 2025) is fabricated on GlobalFoundries' industrial 200mm/300mm silicon-photonics line — the only quantum modality riding an existing tier-1 semiconductor process rather than bespoke lab fab, de-risking the hardest part (volume manufacturing of millions of components).
- Architectural leapfrog: room-temperature-ish photonic qubits with telecom-fiber networking give a natural, long-range chip-to-chip interconnect — the scaling bottleneck that gates superconducting (Google/IBM/Rigetti) and ion-trap (IonQ/Quantinuum) roadmaps. Reported 0.02% single-qubit / 0.8% two-qubit error rates support the loss-tolerant, fusion-based error-correction thesis.
- Sovereign de-risking: Australia/Queensland (~A$940M / $620M USD equity+grant+loan), Illinois (~$700M), and the US CHIPS program ($100M + a federal equity stake, May 2026) underwrite the capex AND signal national-strategic-asset status — a moat competitors can't easily match.
- Blue-chip cap table validates the bet: BlackRock-led Series E with Temasek, Baillie Gifford, NVIDIA's NVentures, Qatar Investment Authority, Macquarie, Morgan Stanley Counterpoint Global, T. Rowe Price — patient, deep-pocketed capital aligned to a multi-year buildout.
- Winner-take-most optionality: fault-tolerant quantum unlocks chemistry, materials, optimization and cryptography TAMs measured in tens of billions; first credible utility-scale machine (targeted ~2027) could capture an outsized share if the roadmap holds.
A pre-revenue, multi-billion-dollar science bet whose entire value rests on an unproven architecture hitting an aggressive timeline; photonics has never demonstrated a fault-tolerant logical qubit at scale, timelines are already slipping, and the ~$7B mark prices in success that is years and many physics risks away.
- No utility-scale machine exists yet — and no demonstrated fault-tolerant logical qubit. The whole thesis is forward-looking; PsiQuantum skipped the NISQ-era hardware its public rivals ship, so there is no commercial product or revenue to underwrite the valuation.
- Timeline slippage already visible: the Australian (Brisbane → Moreton Bay) utility-scale site, originally end-2027, has reportedly pushed toward 2029, and press notes the Brisbane build 'running very late.' Deep-tech timelines on this scale routinely slip further.
- Photonic scaling risk is real and specific: optical loss, single-photon source/detector efficiency, and the sheer component count for ~1M physical qubits are unforgiven by photonics; a fusion-based architecture that works on a chipset paper has not been shown to error-correct at system scale.
- Capital intensity = perpetual dilution / dependence: multi-site fab+datacenter buildouts burn billions; reliance on continued sovereign and venture capital makes the equity vulnerable to a funding-winter or a shift in government priorities (and a US federal equity stake adds political-overhang risk).
- Crowded, well-funded field with more-mature rivals: superconducting (IBM, Google, Rigetti) and trapped-ion (IonQ, Quantinuum at a ~$10B mark) are further along on demonstrated logical qubits; if a rival reaches useful fault tolerance first, PsiQuantum's manufacturing lead is moot. Public quant pure-plays already trade with extreme valuation risk (analysts flag the sector's stretched multiples).
What it is worth
Last-priced-round anchor + qualitative scenario framing (no public comps multiple applies to a pre-revenue private — venture/strategic-option valuation). Cross-checked against listed quantum pure-play market caps and Quantinuum's ~$10B private mark.
~$2–4B (or lower)
if the photonic architecture stalls, timelines slip materially (the 2029 signal worsens), or the sector de-rates in a funding-winter, the mark compresses toward capital-raised / liquidation-option value; a failed DARPA QBI stage or a rival reaching fault tolerance first could impair it severely.
~$7B
the Series E (Sep 2025) post-money mark; appropriate as the carrying value absent a new round or material milestone.
~$15–25B+
if a fault-tolerant logical-qubit demonstration lands and the ~2027 utility-scale roadmap holds, PsiQuantum re-rates toward/above the Quantinuum mark and an IPO premium, pricing in winner-take-most option value on a multi-tens-of-billions TAM.
Valuation is option-like and binary-ish: it is a discounted call on an unproven physics+engineering outcome, not a multiple on cash flows. The ~$7B is asserted by the last round, not earned by financials — treat as speculative. Not financial advice.
SWOT
Strengths
- Only pure-play pursuing utility-scale fault tolerance via an industrial tier-1 fab (GlobalFoundries), not bespoke lab fabrication
- Nature-validated Omega silicon-photonic chipset with reported best-in-class component error rates (0.02% 1Q / 0.8% 2Q)
- Natural telecom-fiber networking / long-range chip-to-chip interconnect — a structural scaling advantage
- Sovereign-backed balance sheet across two countries; blue-chip Series E cap table
Weaknesses
- Pre-revenue at scale; no audited financials, no commercial fault-tolerant product yet
- Extreme cash burn and capex intensity → ongoing dilution / external-capital dependence
- Skipped NISQ hardware — no installed base or near-term commercial traction to bridge to fault tolerance
- Photonic modality unproven at logical-qubit / system scale relative to peers' demonstrations
Opportunities
- First credible utility-scale, fault-tolerant machine (~2027 target) could capture an outsized share of chemistry/materials/optimization TAM
- Government as anchor customer + co-investor (US CHIPS, Australia, Illinois) and DARPA QBI final-stage evaluation
- QCaaS / cloud-access and per-problem monetization once systems are online
- NVIDIA partnership (NVentures + likely CUDA-Q / hybrid HPC integration) bridges to the AI/HPC datacenter stack
Threats
- A rival modality (superconducting / trapped-ion / neutral-atom) reaching useful fault tolerance first
- Further timeline slippage eroding investor patience and the valuation mark
- Funding-winter / sovereign-priority shift cutting off the capital the buildout requires
- Quantum-sector valuation correction (public pure-plays flagged as overvalued) repricing private marks
- Political overhang from US-government equity ownership; geopolitical export-control exposure
Moats, dependencies & bottlenecks
Moats
Producing quantum chips on an industrial tier-1 semiconductor line is a genuine differentiator vs lab-fab rivals — but it is a partnership, not owned fab, and the underlying architecture still must prove out.
loss-tolerant photonic error correction; telecom-fiber interconnect) Nature-published, patent-protected approach with a structural networking advantage; durability hinges entirely on whether it scales to fault tolerance — unproven at system level.
Two-country government backing (US CHIPS + Illinois, Australia/Queensland) is hard for rivals to replicate and funds the capex — but is exposed to political-priority shifts.
~$2.3B+ raised gives runway, but capital is a moat only while the funding environment stays open and the roadmap stays credible.
Dependencies
supplier / manufacturing Single anchor foundry for the Omega chipset; any process, capacity, or relationship disruption would directly gate the roadmap.
Pre-revenue, multi-billion buildout requires ongoing external capital; a funding-winter or government-priority shift is existential.
superconducting single-photon detectors, cryogenics) supplier / materials Specialized, low-volume supply chain for novel materials at scale; PsiQuantum scaling BTO production itself partly mitigates.
Architecture proving out at scale (loss-tolerant fault tolerance) technology / physics The core unproven assumption — no demonstrated fault-tolerant logical qubit yet; the dependency the entire valuation rests on.
CHIPS, Australia/Illinois) regulatory / customer Anchor funding + first-customer roles tied to political programs that can be re-scoped or cut.
Advantages
- Only pure-play attacking utility-scale fault tolerance through an existing industrial semiconductor process (GlobalFoundries) rather than bespoke fabrication
- Photonic qubits' natural telecom-fiber networking gives a long-range interconnect — the scaling bottleneck that constrains superconducting and ion-trap rivals
- Peer-reviewed (Nature) evidence of working chip-scale components with best-in-class reported error rates
- Two-country sovereign capital + national-strategic-asset positioning competitors can't easily match
- Blue-chip, patient cap table (BlackRock, Temasek, NVIDIA, QIA, Baillie Gifford) aligned to a long buildout
Weaknesses
- Pre-revenue at scale; no commercial fault-tolerant product, no audited financials
- Extreme, perpetual cash burn → dilution and external-capital dependence
- Skipped NISQ hardware — no installed base or near-term commercial bridge
- Photonic modality unproven at logical-qubit/system scale vs peers' demonstrations
- Single-foundry concentration and novel-materials supply-chain fragility
- Timeline already slipping (Australia site toward 2029)
Bottlenecks
- Demonstrating a fault-tolerant logical qubit, then a system-scale error-corrected machine — the gating physics milestone, not yet achieved by any photonic platform
- Optical loss and single-photon source/detector efficiency at the component counts (~millions) required for ~1M physical qubits
- Construction/commissioning execution risk on first-of-kind fab + cryogenic-datacenter sites (Chicago IQMP, Moreton Bay) — already showing schedule slippage
- Scaling Barium-Titanate optical-switch production to volume
- Sustaining multi-billion-dollar capital inflows through a multi-year, pre-revenue buildout
Top signals & trends
Top signals
Deep, patient, blue-chip capital validates the multi-year thesis and funds the buildout
National-strategic-asset status and non-dilutive-ish capital; but a government cap-table seat adds political overhang
First sign the aggressive roadmap is slipping — the central risk to the whole valuation
Peer-reviewed evidence the manufacturing-first approach produces working components at fab scale
Independent government technical validation; a QBI failure would be a major negative catalyst
Capital secured and ground broken, but execution/construction risk on a first-of-kind facility
The entire thesis remains forward-looking and physics-gated
Liquidity for early holders is via secondary markets (EquityZen / Nasdaq Private Market) at the ~$7B mark
Trends
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Sole/anchor foundry — fabricates the Omega silicon-photonic chipset at its NY fab (Fab 8, Malta NY); the single most critical supply-chain dependency.
Investor (NVentures) + ecosystem partner — hybrid HPC/AI classical co-processing (CUDA-Q stack) and likely classical control/compute for the quantum datacenter.
Cryogenics / dilution-refrigeration & photonics-component suppliers Superconducting single-photon detectors require cryogenic cooling (PsiQuantum's PsiCube datacenter-style cooling); Barium-Titanate (BTO) optical-switch material supply chain. Specific vendors not disclosed.
Analysis: telecom optical-fiber interconnect — PsiQuantum's networking relies on standard telecom fiber; Corning is a representative fiber/photonics supplier (not a disclosed contract).
Australian Commonwealth & Queensland Governments Anchor 'customer'/co-investor — funding and hosting the world's-first utility-scale site at Moreton Bay (A$940M / ~$620M USD package).
Dept. of Commerce/CHIPS) DARPA QBI final-stage evaluation + Quantum Proving Ground; CHIPS $100M award with a federal equity stake — government as both funder and prospective user.
Largest funder (~$700M) of the Chicago IQMP campus where the first US million-qubit-scale machine is to be deployed.
Pharma, chemicals, materials, energy, finance — the chemistry/optimization TAM for a fault-tolerant machine; no named commercial customers disclosed (pre-commercial).
Largest pure-play by revenue (~$130M 2025, 2026 guide $225–245M); trapped-ion, acquired Oxford Ionics/Lightsynq/Capella — further along on demonstrated logical qubits and commercial traction, though different modality.
Highest-fidelity trapped-ion systems (H-series, Helios); ~$10B 2025 round — the sector's largest valuation. Honeywell-majority-owned; the most credible fault-tolerance rival.
Superconducting full-stack with own fab; public but small revenue (~$7M FY25, declining). Direct fault-tolerant ambition, less capitalized.
Quantum annealing + emerging gate-model; FY25 revenue ~$24.6M (+179%), 83% gross margin. Different problem class (optimization) but competes for the same capital/attention.
The closest public photonic peer — thin-film lithium-niobate photonic chips, owns Fab 1 (Arizona); much smaller and earlier, but a direct modality comparable.
Superconducting leader with a public fault-tolerance roadmap (Starling) and the deepest enterprise/cloud channel; a CHIPS quantum-program peer.
Willow chip / below-threshold error correction milestone (2024) — the most advanced demonstrated superconducting error correction; could reach useful fault tolerance first.
Topological qubit bet (Majorana 1) — earliest-stage but, like photonics, a long-term scaling argument; also the Azure Quantum cloud aggregator.
Canadian photonic peer (squeezed-light/continuous-variable); >$287M raised, announced plans to go public in early 2026 — the most direct photonic competitor by approach.
European photonic players (single-photon-source, time-domain-multiplexing, SiN PIC respectively); smaller scale/funding, validate the photonic modality but not yet utility-scale threats.