8 Quantum & Photonics Startups VCs Are Watching in 2026
The biggest checks in deep tech are flowing into light-based computing. Here are 8 private, independent quantum and photonics startups worth tracking now.
Key Takeaways
- Capital is exploding: Quantum computing companies raised $3.77 billion in the first nine months of 2025, nearly triple the $1.3 billion raised in all of 2024, according to the McKinsey Quantum Technology Monitor 2026.
- Two niches, one substrate: Both quantum computing and AI data-center interconnects are converging on photonics, which uses light instead of electrons to move and process information.
- Private money is institutional now: BlackRock, Temasek, Nvidia, AMD, and Sequoia are leading rounds that used to be the domain of pure deep-tech specialists.
- Public exits are thinning the field: Xanadu, Quantinuum, and Pasqal have all gone public or signed go-public deals, leaving a smaller pool of truly private bets.
- Hardware moats are real: Every company here owns a chip, a fab relationship, or a physics breakthrough that competitors cannot copy with software.
As of June 2026, based on public reporting. Funding figures and valuations sourced from company press releases, TechCrunch, SiliconANGLE, The Quantum Insider, Crunchbase, and McKinsey.
Most investors still file quantum computing under "interesting, but a decade out." They are missing the real story. The capital is not waiting a decade. It is being deployed right now, at scale, by the most disciplined institutional investors on earth.
Quantum computing companies raised $3.77 billion in the first nine months of 2025 alone, nearly tripling the $1.3 billion raised across all of 2024, according to the McKinsey Quantum Technology Monitor 2026. The same report projects quantum technologies could generate up to $97 billion in annual revenue by 2035 and unlock $2.7 trillion in economic value.
And there is a second, quieter boom riding alongside it: photonics for AI. As GPU clusters hit the physical limits of copper wiring, light is becoming the only way to keep scaling. These two waves share the same physics and increasingly the same cap tables.
Here are eight private, independent startups betting that the future of computing runs on light. Every one was verified as still private and independent as of June 2026. We dropped the public names (IonQ, Rigetti, D-Wave) and the recently exited ones (Xanadu, Quantinuum, Pasqal) on purpose.
What Is the Difference Between Quantum and Photonic Computing?
A qubit is the basic unit of quantum information. Unlike a classical bit that is either 0 or 1, a qubit can exist in a combination of both states at once, which is what gives a quantum computer its theoretical edge on certain problems.
Photonic computing is the use of photons, the particles of light, instead of electrons to carry and manipulate data. It shows up in two distinct places: as a way to build qubits (photonic quantum computing) and as a way to move data between AI chips at the speed of light (optical interconnects).
The companies below split across that spectrum. Some are racing to build a useful quantum computer. Others are solving the far more immediate problem of feeding tomorrow's AI superclusters. Both are deep-tech bets with hardware moats, the kind of business that the framework VCs use to evaluate AI startups applies to in spades.
PsiQuantumMillion-qubit photonic quantum computers
PsiQuantum is building fault-tolerant, utility-scale quantum computers using silicon photonics, meaning its qubits are encoded in single particles of light and its chips are manufactured on standard semiconductor lines at GlobalFoundries.
The company was founded in 2016 by Jeremy O'Brien, Terry Rudolph, Mark Thompson, and Pete Shadbolt, a team that spun out of the University of Bristol with deep credentials in photonic quantum optics.
In September 2025, PsiQuantum raised a $1 billion Series E led by BlackRock at a $7 billion valuation, with Temasek, Baillie Gifford, and Nvidia's NVentures joining. That brought total funding above $2.3 billion.
Why it's spicy: PsiQuantum skipped the small-qubit demo phase entirely and bet on manufacturing at scale from day one. It is breaking ground on utility-scale sites in Brisbane and Chicago. If the photonic bet pays off, it leapfrogs everyone still wiring up a few hundred qubits in a lab.
LightmatterPhotonic interconnects for AI supercomputers
Lightmatter builds photonic chips and optical interconnects that move data between AI processors using light, attacking the bandwidth wall that copper wiring creates inside data centers. Its Passage platform delivers co-packaged optics straight into the chip package.
Founded in 2017 as an MIT spinout by Nicholas Harris, Darius Bunandar, and Thomas Graham, the company has turned a PhD thesis into one of the most valuable photonics firms in the world.
In October 2024, Lightmatter raised a $400 million Series D led by T. Rowe Price at a $4.4 billion valuation, quadrupling its prior mark and bringing total capital to roughly $850 million. Fidelity and GV also participated.
Why it's spicy: Lightmatter frames the problem perfectly: model parameters grew 240x in three years, but interconnect bandwidth only improved 2x. That gap is the bottleneck on every frontier AI cluster, and Lightmatter sells the fix.
Ayar LabsCo-packaged optics for AI scale-up
Ayar Labs makes optical interconnects, specifically co-packaged optics that replace the copper links between GPUs with light, letting thousands of chips behave like one giant accelerator. Its TeraPHY optical engine and SuperNova light source are the core products.
The company was founded in 2015 by CEO Mark Wade and CTO Vladimir Stojanovic, building on optical I/O research from MIT, Berkeley, and Colorado.
In March 2026, Ayar Labs closed a $500 million Series E led by Neuberger Berman at a $3.75 billion valuation, pushing total funding to $870 million. The cap table is a who's-who of silicon: Nvidia, AMD, MediaTek, Qatar Investment Authority, and Sequoia Global Equities.
Why it's spicy: When both Nvidia and AMD write checks into the same interconnect startup, that is not a hedge. That is the chip industry agreeing the future of scale-up is optical, and Ayar is the merchant supplier positioned to ride it.
QuEra ComputingNeutral-atom quantum computing at scale
QuEra builds quantum computers from neutral atoms, an approach that traps individual atoms with lasers and uses them as qubits, prized for scalability and long coherence times. Its Aquila machine is already available on Amazon Braket.
QuEra was incorporated in 2018 out of pioneering research at Harvard and MIT, including the labs of Mikhail Lukin and Markus Greiner.
In September 2025, QuEra closed a $230 million Series B (first unveiled in February and expanded by Nvidia's NVentures) that pushed it to unicorn status at a $1 billion valuation. SoftBank Vision Fund 2 and Google Quantum AI are among the backers.
Why it's spicy: QuEra is chasing error correction and logical qubits rather than vanity qubit counts, and it just landed an AWS partnership to deploy its next-generation "Libra" system. Google and SoftBank rarely co-invest in the same physics by accident.
Atom Computing1,200-qubit neutral-atom quantum systems
Atom Computing is the other serious neutral-atom contender, building gate-based quantum computers from optically trapped atoms. Its AC1000 system advertises more than 1,200 fully connected qubits, among the highest counts publicly claimed.
The company was founded in 2018 by Ben Bloom, a physicist who previously worked on the world's most precise atomic clocks.
In 2025, Atom Computing announced more than $300 million in combined commitments, anchored by a $100 million Series C led by Third Point Ventures (with DCVC and Cisco Investments) plus a $100 million Letter of Intent from the U.S. Department of Commerce.
Why it's spicy: Atom Computing is co-developing quantum supercomputers with Microsoft and sold its first on-premises machine to a Nordic initiative backed by the Novo Nordisk Foundation. Government and hyperscaler validation in the same year is a strong signal.
ORCA ComputingRoom-temperature photonic quantum computers
ORCA Computing builds photonic quantum computers that run at room temperature using time-bin encoded photons stored in quantum memory loops, sidestepping the cryogenic cooling most rivals require. Its rack-mounted PT Series targets quantum-accelerated generative AI and optimization.
ORCA was spun out of the University of Oxford in 2019 by Professor Ian Walmsley, Richard Murray, and Josh Nunn, combining three decades of photonics research with national quantum-program leadership.
In November 2025, ORCA raised a $37 million Series B with a notable investor list including Nvidia's NVentures and Qatar Investment Authority. In 2025 it also acquired Oxford Quantum Circuits in a rare cross-modality consolidation.
Why it's spicy: Room-temperature, air-cooled, and data-center-ready is the holy grail for deployment economics. ORCA is small relative to PsiQuantum, but its hardware slots into existing racks, which is exactly what enterprise buyers want.
Q.ANTPhotonic processors for AI and HPC
Q.ANT builds photonic processors that perform computation with light rather than transistors, targeting the energy-hungry nonlinear math at the heart of AI and scientific simulation. The company claims one optical element can replace 100 to 1,000 transistors for the same function.
Q.ANT was founded in 2018 by Michael Förtsch, a physicist and former Max Planck researcher, as a spin-out from German laser giant TRUMPF.
In July 2025, Q.ANT raised a €62 million Series A co-led by Cherry Ventures, UVC Partners, and imec.xpand, the largest photonic-computing Series A in Europe. A second tranche pushed total funding past $80 million.
Why it's spicy: Q.ANT already deployed its Native Processing Server at Germany's Leibniz Supercomputing Centre. Europe's largest photonic-computing raise, real hardware in a real HPC center, and a 30x energy-efficiency pitch make it the continent's clearest light-based-AI bet.
QuandelaPhotonic quantum computers from single photons
Quandela is France's photonic quantum champion, building full-stack quantum computers from quantum-dot single-photon sources, plus open-source software like its Perceval framework. Machines include Belenos (12 qubits) and the planned Canopus (24 qubits).
Quandela was founded in 2017 out of the Institut d'Optique by Niccolo Somaschi, Pascale Senellart, and Jean Senellart, pairing world-class photon-source physics with software depth.
The company has raised more than €107 million to date, anchored by its 2023 Series B and an additional €50 million expansion round backed by France's France 2030 program, with former OVHcloud CEO Michel Paulin now chairing the board.
Why it's spicy: Quandela is delivering real machines to real customers, including a system installed at France's CEA under the EuroHPC program. It was also selected for PROQCIMA, France's €500 million sovereign quantum initiative, which functions as a state-funded vote of confidence.
How to Evaluate a Deep-Tech Bet Like This
Quantum and photonics break the usual SaaS playbook. There is no $1M-ARR-to-product-market-fit story here. The signal is different: physics that works, a fab relationship that scales, a customer who is a government or a hyperscaler, and a cap table of investors who can underwrite a decade.
The market backdrop is real. The global quantum computing market is projected to grow from $3.52 billion in 2025 to $20.20 billion by 2030 at a 31.3% CAGR, according to MarketsandMarkets. On the photonics side, the silicon photonics market is forecast to climb from roughly $2.8 billion in 2025 to $9.6 billion by 2030, per market research from Wissen Research, driven almost entirely by AI data-center demand. Sizing that opportunity correctly is its own discipline, which is why getting TAM right matters so much for hardware bets.
This is exactly the kind of analysis where a data-driven scoring tool earns its keep. Unicorn Screener is a research-backed startup evaluation platform that runs structured assessments across founder quality, market dynamics, traction signals, and competitive positioning, so you can compare a photonic-quantum moonshot against an optical-interconnect cash machine on the same scorecard. Check the top-ranked startups on our leaderboard to see how today's most-watched deep-tech names stack up.
No scoring model can guarantee outcomes, and deep tech carries binary technical risk that consumer software never does. But a consistent framework beats gut feel, especially in a field where the jargon is designed to dazzle.
What These Eight Have in Common
Look across the list and the pattern is sharp.
Hardware moats, not feature moats. Every company owns a chip, a fab relationship, or a physics breakthrough. None of them can be cloned by a competitor shipping a faster web app.
Institutional capital, not just deep-tech specialists. BlackRock, T. Rowe Price, Temasek, Nvidia, AMD, and Sequoia are leading these rounds. That is patient money that understands long horizons, the same power-law math that drives venture returns.
Two timelines, one substrate. Optical interconnects (Lightmatter, Ayar Labs, Q.ANT) monetize now, riding the AI buildout. Quantum (PsiQuantum, QuEra, Atom, ORCA, Quandela) is the longer bet. Photonics ties them together.
| Company | Niche | Latest Round | Lead / Date |
|---|---|---|---|
| PsiQuantum | Photonic quantum | $1B Series E ($7B) | BlackRock, Sep 2025 |
| Lightmatter | Photonic interconnects | $400M Series D ($4.4B) | T. Rowe Price, Oct 2024 |
| Ayar Labs | Optical interconnects | $500M Series E ($3.75B) | Neuberger Berman, Mar 2026 |
| QuEra | Neutral-atom quantum | $230M Series B ($1B) | NVentures, Sep 2025 |
| Atom Computing | Neutral-atom quantum | $100M Series C (+$200M LOI/grants) | Third Point, 2025 |
| ORCA Computing | Photonic quantum | $37M Series B | NVentures/QIA, Nov 2025 |
| Q.ANT | Photonic computing | €62M Series A | Cherry/UVC/imec, Jul 2025 |
| Quandela | Photonic quantum | €50M expansion (+2023 Series B) | France 2030, 2026 |
What This Means for You
-
Separate the two waves. Optical interconnects have revenue today. Quantum is a 5-to-10-year horizon. Price them differently and do not let "quantum" hype inflate an interconnect multiple or vice versa.
-
Follow the fab. A relationship with GlobalFoundries, TSMC, or imec is a harder moat than any pitch deck. Ask who actually manufactures the chip.
-
Watch the exits. Xanadu went public, Quantinuum IPO'd, Celestial AI got acquired by Marvell, and Pasqal signed a SPAC deal. The private pool is shrinking, which means the remaining independents get scarcer and more valuable. Learn what actually makes a startup a unicorn before you chase the next one.
-
Score before you wire. These eight all clear a high bar on team, market, and capital validation. Try Unicorn Screener to run the same structured evaluation on the next light-based startup that lands in your inbox.
Want to screen startups like a top-tier VC? Score any startup for free with our research-backed evaluation model.